<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[The Hidden Weave: Strategic Weave]]></title><description><![CDATA[Decoding the future of business. Deep-dive articles and frameworks on AI, innovation, marketing, and product strategy.]]></description><link>https://www.hiddenweave.com/s/strategic-weave</link><image><url>https://substackcdn.com/image/fetch/$s_!J5MK!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45ca29fa-a7ef-41b7-95f7-fc1f4de7706a_1200x1200.png</url><title>The Hidden Weave: Strategic Weave</title><link>https://www.hiddenweave.com/s/strategic-weave</link></image><generator>Substack</generator><lastBuildDate>Sun, 22 Nov 2026 20:00:30 GMT</lastBuildDate><atom:link href="https://www.hiddenweave.com/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[Mohan Sawhney]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[mohansawhney@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[mohansawhney@substack.com]]></itunes:email><itunes:name><![CDATA[Mohan Sawhney]]></itunes:name></itunes:owner><itunes:author><![CDATA[Mohan Sawhney]]></itunes:author><googleplay:owner><![CDATA[mohansawhney@substack.com]]></googleplay:owner><googleplay:email><![CDATA[mohansawhney@substack.com]]></googleplay:email><googleplay:author><![CDATA[Mohan Sawhney]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[After the Pyramid Part 1: The Partner without the Firm]]></title><description><![CDATA[What AI does to the consulting firm pyramid]]></description><link>https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner</link><guid isPermaLink="false">https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Wed, 18 Nov 2026 15:49:28 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!bBAE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!bBAE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bBAE!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 424w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 848w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!bBAE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg" width="1299" height="732" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:732,&quot;width&quot;:1299,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!bBAE!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 424w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 848w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!bBAE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F40348e84-2718-4834-923a-3607b590bc87_1299x732.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Over my teaching career at the Kellogg School of Management, I have educated thousands of students who went on to build careers at the leading consulting firms. I have trained consultants and partners at many of those firms, and I have advised their leaders. Consulting firms go to great lengths to set themselves apart, but they are all built the same way: as a pyramid, with a few partners at the top, a wide base of analysts and associates at the bottom, and engagement managers in between.</p><p>The pyramid is a logical structure, and it follows from the economics of selling expertise. A partner&#8217;s time is scarce, and her expertise is valuable. So, the firm spreads that expertise across many engagements by giving the partner a team of less expensive associates. The client pays for the team&#8217;s hours at a markup, and the markup on the base of the pyramid pays for the top. David Maister defined this arithmetic decades ago. Profit per partner is the product of margin, billing rate, utilization, and leverage, which is the number of people beneath each partner. Richer partners need a wider base. The up-or-out promotion system keeps the base wide by letting go of most associates before they become expensive partners.</p><div class="pullquote"><p>Clients have traditionally paid for the whole consulting firm pyramid because partners cannot deliver expertise without the associates beneath them. But AI can now decouple expertise from execution.</p></div><p>Clients are forced to pay for the whole pyramid because partners cannot deliver expertise without the people beneath them. AI is now disrupting this logic by unbundling expertise from execution. An AI coworker can do much of the research, modeling, and drafting that the firms hired the base of the pyramid to do, and a partner can rent that capacity from a frontier model. That raises a question: how much of a partner&#8217;s value belongs to her, and how much to the firm? It depends on the kind of work the client is buying.</p><p>Consider Unity Advisory. In 2025, Marissa Thomas, who had run PwC&#8217;s UK firm, and Steve Varley, who had chaired EY in the UK, started a CFO advisory firm with up to $300 million from Warburg Pincus. Unity does no audit work. It built AI into delivery from the first day, and it sells to mid-market companies and private equity portfolios. By October 2026, it had about 120 clients, it had won work from the firms its founders came from, and it had fended off a legal threat from PwC. Varley has said that after he left EY, his former clients asked what he would do next and wanted to keep working with him. The clients followed the partners.</p><p>Boris Groysberg&#8217;s research points the other way. He spent a decade studying star equity analysts who changed firms, and he found that most of them got worse. Their rankings fell when their teams stayed behind. The platform that made them look brilliant was part of the brilliance. Clients follow people, but people perform worse without the platform. The future of the strategy consulting firm depends on which parts of the platform AI has made portable.</p><h2>What the pyramid gave the partner</h2><p>Think about how an engagement runs inside the pyramid. A senior partner sells the engagement, an engagement manager runs it, and consultants and analysts do the interviews, build the models, and produce the decks. The partner reviews the work, shapes the storyline with the team, and presents it to the client. That is the leverage Maister described, and a strong book of clients allows a partner to run several teams at once.</p><p>The hierarchy gives the partner a second, less visible benefit: it keeps her expertise pointed at the hardest problems. Analysts solve what they can and escalate the rest, and managers do the same. By the time a problem reaches the partner, the people below her have handled the easier parts, so her hours go to the parts that need her expertise. The firm bills the client for the whole chain, and the client pays because the work is sold as a bundle.</p><p>The pyramid also made the partner. She was once the analyst who burnt the midnight oil building models and decks, and she learned her craft over years of supervised work lower down in the pyramid. The firm supplies production, routing, and training; the partner supplies judgment and trust; and the client cannot buy one without the other.</p><h2>How AI shrinks the middle</h2><p>In an earlier essay, I presented the <a href="https://open.substack.com/pub/mohansawhney/p/from-the-barrel-to-the-hourglass?utm_campaign=post-expanded-share&amp;utm_medium=web">hourglass model of knowledge work</a>. Let&#8217;s apply this model to consulting engagements. Every engagement has a front end, where you frame the question and set the hypotheses. It has a middle, where the team does the research, the analysis, and the drafting. It has a back end, where you test the recommendations, defend them, and turn them into action. I call these stages <em>Design</em>, <em>Do</em>, and <em>Decide</em>. Human judgment matters most in Design and Decide. Most of the hours go into Do, and consulting firms built the pyramid to staff it. AI has moved into Do as a coworker. It reads the filings, drafts the interview guide, builds the first version of the model, stress-tests the scenarios, and produces the storyline that the partner can critique. It does this in hours, and all the partner needs is a laptop and access to Claude or ChatGPT.</p><p>Empirical research backs this up, with an important caveat. In one experiment, several hundred BCG consultants worked on realistic consulting tasks. On tasks that AI handles well, the consultants who used it finished about a quarter faster, and evaluators rated their work higher. In a second experiment at P&amp;G, one professional working with AI did as well as a two-person team working without it. But the gains disappear when AI is pushed past what it does well. On a task where AI tended to give a wrong answer, the BCG consultants who used it did worse than those who did not. And in a separate study, experienced software developers took longer with AI on code they already knew well. The lesson: AI can produce a lot of work quickly, but someone with experience must check it.</p><p><strong>Exhibit 1</strong> illustrates this idea for a twelve-week strategy engagement. AI cuts the hours in Do by about three quarters. The hours in Design stay about the same, because framing the question still takes a partner&#8217;s judgment. The hours in Decide increase, because an experienced person must check all the AI output before it reaches the client. An engagement that once needed about 2,400 person-hours and a team of five now needs about a thousand hours from a partner, a senior adviser, and an engineer, plus a lot of AI.</p><p><strong>Exhibit 1. How AI converts consulting engagements from a pyramid to an hourglass</strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!RCtD!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!RCtD!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 424w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 848w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 1272w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!RCtD!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png" width="1300" height="598" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:598,&quot;width&quot;:1300,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!RCtD!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 424w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 848w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 1272w, https://substackcdn.com/image/fetch/$s_!RCtD!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4754afc6-daf6-4afd-8ee5-9c491a3cdf87_1300x598.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>Illustrative staffing for a bounded strategy engagement; the hours are stylized and not a reconstruction of any firm&#8217;s actual staffing.</em></p><p>I want to be careful here, because the hourglass is not a free lunch. Using AI in the middle comes with its own costs. You must give it access to your data, test what it produces, and keep it secure. Above all, you need a domain expert to judge when it is wrong. The savings also don&#8217;t apply to every engagement. If you&#8217;re running a transformation across several countries with lots of stakeholders, you still need people in the room. And it remains to be seen whether a smaller team can deliver the same quality. New ventures like Unity Advisory are promising, but they are young and have yet to prove the model works at scale. So far, we have talked about the hours that AI saves. But the hours don&#8217;t tell you who gets paid for them. If a partner can do the work with a small team and a laptop, the client must decide whether they are buying from the partner or from the firm behind her. That&#8217;s the question I turn to next.</p><h2>Whose value is it?</h2><p>To find out how much of the fee belongs to the partner, let&#8217;s list what the client pays for, as <strong>Exhibit 2</strong> does. For each item, I ask who owns it, the partner or the firm, and what AI changes.</p><p><strong>Exhibit 2. The value ledger: who owns what the client pays for</strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!A3TT!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!A3TT!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 424w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 848w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 1272w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!A3TT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png" width="1280" height="1275" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1275,&quot;width&quot;:1280,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;Title: Exhibit 2. The value ledger - Description: Table: for each thing the client pays for (trust and access, judgment, production, quality assurance, accountability, breadth and reach, training), what belongs to the partner, what belongs to the firm, and what AI changes.&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Title: Exhibit 2. The value ledger - Description: Table: for each thing the client pays for (trust and access, judgment, production, quality assurance, accountability, breadth and reach, training), what belongs to the partner, what belongs to the firm, and what AI changes." title="Title: Exhibit 2. The value ledger - Description: Table: for each thing the client pays for (trust and access, judgment, production, quality assurance, accountability, breadth and reach, training), what belongs to the partner, what belongs to the firm, and what AI changes." srcset="https://substackcdn.com/image/fetch/$s_!A3TT!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 424w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 848w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 1272w, https://substackcdn.com/image/fetch/$s_!A3TT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F403f7f20-e798-4a63-80af-d254d06bfc76_1280x1275.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>At first glance, the firm&#8217;s ledger looks strong. Then read the right-hand column. Production was the asset that made the firm indispensable to the partner, and production is the asset that AI attacks first. The firm keeps accountability, breadth, collegial quality assurance, and training. These matter a great deal for some kinds of work and very little for others. Trust and judgment were always more personal than institutional. The firm could claim them only because the partner could not deliver without the firm&#8217;s production. Take that away, and the firm must earn its share of the fee from the assets in its own column.</p><p>Accountability is the asset that advocates of AI-native boutiques tend to leave out. A board buys more than advice. It also buys a party to blame, to insure, and to name in the proxy statement when the merger fails. The name on the engagement letter brings professional indemnity, conflict clearance, confidentiality systems that have survived regulators, and the willingness to be sued. A ten-partner boutique can buy insurance and set up a conflicts process, but it can&#8217;t buy the comfort a board feels when the McKinsey or Accenture name is on the letter. The bigger the stakes, the more accountability is worth, and AI doesn&#8217;t change that.</p><p>Partners also tend to underestimate the value of their colleagues. Inside a big firm, a partner can bounce an idea off a colleague before it goes to the CEO. Sometimes that colleague has seen the same move fail at another client and can warn her off it over a cup of coffee. When partners leave to start a boutique, they leave this safety net behind. AI can critique a recommendation, but it can&#8217;t tell you, &#8220;We tried that at another client, and here&#8217;s what went wrong.&#8221; Firms could charge for this protection, but so far, they haven&#8217;t needed to.</p><p>My colleague Subhendu Modak and I built a simulation of a law firm. Law firms are even more pyramid-shaped than consulting firms, so they make a good test case. What we learned surprised us. Clients happily pay double or triple rates for judgment work, and they don&#8217;t want AI anywhere near it. That willingness to pay rests on the firm&#8217;s reputation, which takes years to build. Senior lawyers earn it by doing the hard work themselves. But it can erode quickly if the firm automates too much, makes a costly mistake, or lets its best senior people go. And here&#8217;s the key point: the reputation belongs to the firm, not to any one lawyer. A rainmaker can walk out the door and take his clients with him, but he can&#8217;t take the reputation that made those clients willing to pay a premium. That&#8217;s exactly what Groysberg found with star analysts.</p><h2>Where a partner can win without the firm</h2><p>The ledger shows that the answer depends on the kind of work the client wants done. <strong>Exhibit 3 </strong>maps that work along two questions. First, how much does the client need a big firm&#8217;s name to stand behind the advice? Second, how much analysis and grunt work does the job require, now that AI can do much of it? In the lower half of the map, where the client doesn&#8217;t need the firm&#8217;s name, a partner can compete on her own.</p><p><strong>Exhibit 3. The mandate map, and how much of the value follows the partner</strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!bnLS!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bnLS!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 424w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 848w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 1272w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!bnLS!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png" width="1120" height="794" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:794,&quot;width&quot;:1120,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!bnLS!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 424w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 848w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 1272w, https://substackcdn.com/image/fetch/$s_!bnLS!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F406249ce-0a3d-4fca-9f31-f9e873b6b2e5_1120x794.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Let&#8217;s walk through Exhibit 3 one quadrant at a time. I call the lower left quadrant &#8220;Counsel.&#8221; Here the CEO wants one trusted voice on a big decision. There isn&#8217;t much analysis to do. The stakes are high, but the CEO makes the call and owns it, so she doesn&#8217;t need a big firm&#8217;s name behind the advice. Picture a CEO deciding whether to sell a division or replace a top executive and calling the one adviser she trusts. Elite partners have done this work on their own for decades, as senior advisers and board members. Most of the value here belonged to the partner before AI, and it still does.</p><p>The lower right quadrant is new ground, and I call it &#8220;Studio work.&#8221; The client still doesn&#8217;t need a big firm&#8217;s name behind the advice, but the job takes a lot of analysis: models, scenarios, prototypes, and some hands-on help putting the plan in place. Think of a market entry, a pricing redesign, a portfolio review, or the strategy for a business unit. In the past, only the pyramid could take on this work, because it took a wide base of analysts, and most of the value went to the firm. Today, AI and a small team can do it, and a partner the client trusts can sell it. This is the quadrant that AI changes the most. Most of the value now follows the partner, and studio work makes up a large share of what the strategy firms bill today.</p><p>Next is the upper left, which I call &#8220;Institutional judgment.&#8221; These are the bet-the-company moves: a hostile bid, a restructuring that will end up in court, or a settlement with a regulator. The analysis may be modest, but the stakes are high, and the board wants a name it can stand behind. That name belongs to the firm, not to the partner.</p><p>Last is the upper right, which I call &#8220;Institutional transformation.&#8221; These are the multi-year programs that cut across countries and functions, and they need both a lot of work and a trusted name. The firm still needs many people on the ground, and big firms armed with AI are strong here. If you plan to leave your firm, don&#8217;t expect to win the work in either of the top two corners.</p><p>Step back from the map, and you can see how exposed the big firms are. They earn money in all four quadrants, but the bottom two are now up for grabs. And the people coming after that work aren&#8217;t strangers. They&#8217;re the firm&#8217;s own former partners, who know the clients and know how the firm works. The firm has two ways to defend itself. It can shift its work up the map, toward the jobs where clients need its name. And it can charge openly for what only the firm provides: accountability, breadth, peer review, and training. If it does both, it can keep its premium. If it keeps billing for the analysis that AI now does, that premium will walk out the door with the partners who leave.</p><p>Experts who follow the consulting industry have proposed many new shapes of the post-AI consulting firm. Some say it will look like an obelisk, others a diamond, an upside-down pyramid, or a box. Each drawing is a guess at how many people a firm will need at each level once AI takes over the junior work. They aren&#8217;t really competing predictions. Each one describes a firm that has picked one quadrant and staffed for it. Counsel needs almost no base. Studio work needs a thin layer of engineers and senior advisers beneath each partner, and that is the diamond. Institutional transformation still needs many people in client rooms, and that is a pyramid with AI inside it. A firm that works in all four quadrants needs all these shapes, one for each kind of work. In the end, the work the client buys decides how a firm should be staffed and how it makes money.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div><h2>Sleeping with the enemy: Consulting firms and frontier labs</h2><p>Until recently, a partner who left with her clients had a problem. Who would do the research, the modeling, and the drafting? In 2026 the frontier labs answered that question. OpenAI signed Frontier Alliances with McKinsey, BCG, Accenture, and Capgemini in February. In May it launched its own Deployment Company, which OpenAI majority-owns and in which McKinsey, Bain, and Capgemini invested. Anthropic launched Ode in July with Blackstone and Hellman &amp; Friedman. Ode places engineers inside enterprises to build and run AI systems. In August, Anthropic signed Bain as a Global Premier partner and deployed Claude to 19,000 Bain employees. In October it opened its Frontier Academy, a $100 million program to train 10,000 forward-deployed engineers by the end of 2027, with cohorts from Accenture, Bain, Deloitte, and McKinsey, including engineers from McKinsey&#8217;s QuantumBlack. In November it joined Hg in a European deployment venture. Warburg Pincus financed Unity Advisory separately. Within a year, the labs, the consulting firms, and the private equity funds built a web in which each party needs the others and also competes with them.</p><p>Each party brings something the others lack. The labs supply the models and want a way into large companies. The consulting firms have the client relationships and the deployment playbooks, and they need the models to compete. The private equity funds have capital and a built-in market in the companies they own, and they want to own a services platform instead of paying fees to one. And they all need more of a new kind of person: the forward-deployed engineer who turns a model into a working system inside a client. The labs train these engineers. The consulting firms send their own people to learn from the labs that fund their rivals. The funds buy companies full of them, and client companies try to hire them away from everyone else. Think of the Frontier Academy as an old-fashioned apprenticeship with a budget. The lab is happy to pay for it, because the lab profits from the work those engineers go on to do. Consulting firms used to have the same deal with their junior staff, and they are losing it.</p><p>Let&#8217;s follow the money. The lab wins no matter who gets the client. It earns on model usage if the consulting firm wins, on equity and usage if a deployment venture wins, and on usage again if the client builds in-house. The consulting firm earns only when it wins, and it can&#8217;t walk away from the labs, because it needs their models and their training. The firms have hedged. McKinsey, Bain, and Capgemini invested in a venture that competes with their own implementation arms. The private equity funds hedged too: they backed Ode and the Hg venture to get deployment capacity without paying consulting rates, and they backed Unity, the partner who left.</p><p>For that partner, this removes the last reason to need the firm. She can rent the work from the labs, hire engineers the labs trained, and raise money from the funds. The firm has a more difficult task. It must partner with the companies building its replacement in the lower half of the map, while its own people become the labs&#8217; deployment workforce. It keeps the upper half of the map, but it is no longer the one doing the work.</p><h2>The pushback from senior partners</h2><p>I can already hear the pushback from the managing partners of the big firms. It goes something like this: &#8220;CEOs don&#8217;t hire us for analysis. They hire us for judgment, independence, credibility with the board, and the muscle to move a large organization. We can use the same AI as any boutique, and we bring more specialists, more reach, a reputation built over generations, and the training that makes future partners. Why would a CEO bet the company on a handful of people and rented technology?&#8221;</p><p>They have a point, and for the work in the upper half of the map, they&#8217;re right. The firm does own the name, the accountability, and the peer review. It can use AI at least as well as a boutique can, and a boutique has no way to train its own partners. But the argument breaks down in the lower half of the map. It assumes clients will keep giving that work to the big firm out of habit, and they won&#8217;t. The partner who left knows the client, the client trusts her, she can rent the analysis, and she charges less.</p><p>The boutique model looks promising, but it hasn&#8217;t been proven. On paper, a small firm of senior partners working with AI can charge less than a big firm and still pay its partners more. That only works if three things hold. The partners must win work without a famous name behind them. The technical work must stay manageable. And clients must not discount advice because it comes from a smaller firm. If even one of them fails, the advantage starts to disappear. Where would I place my bet? On the boutique for work in the lower half of the map, and on the big firm for work in the upper half.</p><h2>What I predict, and what firms must do</h2><p>Here is what I predict will happen. First, the big firms won&#8217;t collapse. Instead, we&#8217;ll see a long, slow fight, and some partners will leave and do well in counsel and studio work. Second, the big firms will fight back by moving up the map, charging openly for what only they provide, and building studios of their own. Third, the labs will profit no matter who wins. Fourth, clients will get the same judgment for less, and they&#8217;ll pay the big-firm premium only when they truly need the big firm. The pyramid will shrink at the bottom of the map and hold at the top. Finally, the firms that thrive will run two organizations side by side: a studio that does expert work with small senior teams and plenty of AI, and a guild that trains the next generation of partners on its own budget, because clients will no longer pay for that training. The firms will survive. The bundle will not.</p><div class="pullquote"><p>Large consulting firms must run two organizations side by side: a studio where small senior teams do expert work with AI, and a guild that trains future partners at the firm&#8217;s expense.</p></div><p>What should you do as AI pulls the partner and the firm apart? If you are a senior partner, start with an honest audit. Go through what your clients pay for, line by line, and separate what they buy from you from what they buy from the firm. Be especially honest about two things that are easy to take for granted: the firm&#8217;s name on the letter, and the colleagues who catch your mistakes. If most of your work sits in the lower half of the map, and most of what your clients value comes from you, you are worth more on your own than your pay suggests. If most of your work sits in the upper half, you&#8217;re better off staying, because the firm is worth more to you than it costs you.</p><p>If you run a big firm, lead with the work that only a big firm can do: the bet-the-company decisions where the board wants a name it can trust, and the large transformations that need many people across many countries. Show clients exactly what the premium buys, so they can see the difference when a boutique bids against you. And build a studio of your own before one of your former partners beats you to it. Keep it small: a few senior people, a couple of engineers, and plenty of AI. Then build a guild next to the studio, with its own budget and its own goals. The way you turn young analysts into partners is the one thing a boutique can&#8217;t copy. But as AI takes over the junior work, that training won&#8217;t happen on its own anymore. You&#8217;ll have to fund it intentionally.</p><p>If you&#8217;re a CEO buying this work, ask yourself what you&#8217;re really buying before you sign. For counsel or studio work, you&#8217;re paying for the partner, and a boutique may give you the same person for a lower fee. When your board needs a name it can stand behind, pay for the name, and make sure you also get the colleagues who will push back on the advice. Just don&#8217;t pay big-firm prices for studio work out of habit, and don&#8217;t hand a boutique a decision that may end up in court.</p><p>Picture the partner on the morning she leaves. She takes no files and no team with her, only her clients&#8217; trust and a laptop. For most of the history of consulting, that was never enough. Today, for a growing share of the work, it is. The pyramid made her, and now it will have to earn her back.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div><h2>Sources</h2><p>Anthropic. <a href="https://claude.com/resources/articles/bain-company-joins-the-claude-partner-network-as-a-global-premier-partner">&#8220;Bain &amp; Company Joins the Claude Partner Network as a Global Premier Partner.&#8221;</a> August 25, 2026.</p><p>Bain &amp; Company. <a href="https://www.bain.com/about/media-center/press-releases/2026/bain-company-openai-a-new-venture-to-deploy-ai-at-enterprise-scale/">&#8220;Bain &amp; Company Invests in the OpenAI Deployment Company.&#8221;</a> Press release, May 11, 2026.</p><p>Becker, Joel, Nate Rush, Elizabeth Barnes, and David Rein. <a href="https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/">&#8220;Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity.&#8221;</a> METR, July 10, 2025.</p><p>Business Wire. <a href="https://www.businesswire.com/news/home/20260715205134/en/Anthropic-Blackstone-and-Hellman-Friedman-Introduce-Ode-with-Anthropic-an-Enterprise-AI-Services-Firm">&#8220;Anthropic, Blackstone, and Hellman &amp; Friedman Introduce Ode with Anthropic.&#8221;</a> July 15, 2026.</p><p>CNBC. <a href="https://www.cnbc.com/2026/10/02/anthropic-to-invest-100-million-to-train-ai-engineer-talent.html">&#8220;Anthropic to Invest $100 Million to Train AI Engineer Talent.&#8221;</a> October 2, 2026.</p><p>&#8212;&#8212;&#8212;. <a href="https://www.cnbc.com/2026/02/23/open-ai-consulting-accenture-boston-capgemini-mckinsey-frontier.html">&#8220;OpenAI Lands Multiyear Deals with Consulting Giants in Enterprise Push.&#8221;</a> February 23, 2026.</p><p>Dell&#8217;Acqua, Fabrizio, Charles Ayoubi, Hila Lifshitz, Raffaella Sadun, Ethan Mollick, Lilach Mollick, et al. <a href="https://www.nber.org/papers/w33641">&#8220;The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise.&#8221;</a> NBER Working Paper 33641, April 2025.</p><p>Dell&#8217;Acqua, Fabrizio, Edward McFowland III, Ethan Mollick, Hila Lifshitz-Assaf, Katherine C. Kellogg, et al. <a href="https://www.hbs.edu/ris/Publication%20Files/24-013_d9b45b68-9e74-42d6-a1c6-c72fb70c7282.pdf">&#8220;Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality.&#8221;</a> Harvard Business School Working Paper 24-013, September 2023.</p><p>Duncan, David, Tyler Anderson, and Jeffrey Saviano. <a href="https://hbr.org/2025/09/ai-is-changing-the-structure-of-consulting-firms">&#8220;AI Is Changing the Structure of Consulting Firms.&#8221;</a> <em>Harvard Business Review</em>, September 2025.</p><p>FStech. <a href="https://www.fstech.co.uk/fst/PWC_Unity_Advisory_Locked_In_Legal_Dispute.php">&#8220;PwC and Unity Advisory Locked in Legal Dispute.&#8221;</a> February 2026.</p><p>Groysberg, Boris. <em><a href="https://press.princeton.edu/books/paperback/9780691154510/chasing-stars">Chasing Stars: The Myth of Talent and the Portability of Performance</a></em>. Princeton, NJ: Princeton University Press, 2010.</p><p>Harris-Davison, Rosie. <a href="https://www.cityam.com/private-equity-backed-unity-advisory-hits-120-clients-as-former-big-four-chiefs-target-rivals/">&#8220;Private Equity-Backed Unity Advisory Hits 120 Clients as Former Big Four Chiefs Target Rivals.&#8221;</a> <em>City AM</em>, October 2026.</p><p>HgCapital Trust. <a href="https://www.tradingview.com/news/reuters.com,2026-11-03:newsml_RSC4882Ba:0-reg-hgcapital-trust-plc-hg-invests-in-new-ai-venture-alongside-anthropic/">&#8220;Hg Invests in New AI Venture alongside Anthropic.&#8221;</a> RNS announcement, November 3, 2026.</p><p>Maister, David H. <em><a href="https://www.simonandschuster.com/books/Managing-The-Professional-Service-Firm/David-H-Maister/9780684834313">Managing the Professional Service Firm</a></em>. New York: Free Press, 1993.</p><p>&#8212;&#8212;&#8212;. <a href="https://davidmaister.com/articles/1/79/">&#8220;Results and Rewards in the Multi-Group Firm.&#8221;</a> davidmaister.com.</p><p>Sawhney, Mohanbir, and Subhendu Modak. <em><a href="https://pedagogix.ai/simulations/agency">Agency: Org Design and AI Strategy Simulation</a></em>. Pedagogix, 2026.</p><p>SiliconANGLE. <a href="https://siliconangle.com/2026/05/11/openai-launches-professional-services-business-4b-investment/">&#8220;OpenAI Launches Professional Services Business with $4B Investment.&#8221;</a> May 11, 2026.</p><p>Warburg Pincus. <a href="https://warburgpincus.com/2025/06/30/warburg-pincus-announces-partnership-with-marissa-thomas-and-steve-varley-to-launch-unity-advisory-a-next-generation-cfo-advisory-firm/">&#8220;Warburg Pincus Announces Partnership with Marissa Thomas and Steve Varley to Launch Unity Advisory.&#8221;</a> Press release, June 30, 2025.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">hanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/after-the-pyramid-part-1-the-partner?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div>]]></content:encoded></item><item><title><![CDATA[The Unmaking and Remaking of Expertise]]></title><description><![CDATA[How AI is rewriting who makes expert work, who profits from it, and who trains the next generation of experts.]]></description><link>https://www.hiddenweave.com/p/the-unmaking-and-remaking-of-expertise</link><guid isPermaLink="false">https://www.hiddenweave.com/p/the-unmaking-and-remaking-of-expertise</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Tue, 10 Nov 2026 16:31:02 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!199V!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!199V!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!199V!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 424w, https://substackcdn.com/image/fetch/$s_!199V!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 848w, https://substackcdn.com/image/fetch/$s_!199V!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!199V!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!199V!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg" width="1299" height="732" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:732,&quot;width&quot;:1299,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!199V!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 424w, https://substackcdn.com/image/fetch/$s_!199V!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 848w, https://substackcdn.com/image/fetch/$s_!199V!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!199V!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5e5437ef-bfc7-4b2a-8280-b1f0b9718a8c_1299x732.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>A friend sent me a message a couple of days ago that said, &#8216;What do you think?&#8217; Attached was a press release: Hg, the European private equity firm, was partnering with Anthropic to launch a <a href="https://breakingthenews.net/Article/Anthropic-Hg-said-to-launch-European-AI-consultancy/67386284">European AI consulting firm</a> that will place engineers inside the companies it owns, to build and run AI systems there. The message was innocuous and the press release seemed routine, till I dug deeper. As I read the press release, it triggered something that had been marinating in my head for a year. The announcement talked about three things that have been bothering me. </p><ul><li><p>A frontier lab was moving into the work that consulting and technology services firms do. </p></li><li><p>A private equity fund was financing that work for its own portfolio rather than buying it from the firms. </p></li><li><p>The announcement made no mention of the people who would once have learned the trade by doing that work. </p></li></ul><p>That message was the spark. This three-part series is the fire.</p><p>For more than three decades, I have prepared students for the entry-level work of consulting and technology services firms, and thousands of my students work in these firms, some in very senior positions. I advise the same firms, so I understand how they work. These firms are my clients and my students&#8217; employers. Ironically, I also compete with them through a two-person AI startup called <a href="http://www.pedagogix.ai">Pedagogix</a> that I co-founded earlier this year. In all these roles, I keep returning to the same questions. If AI does the work that trained the experts, who trains the next generation of experts? And what happens to the firms, the schools, and the careers built on the old model of professional services firms?</p><p>The professional services firm does three jobs with the same people. It produces expert work, it gets paid for that work, and it trains the next generation of experts. Clients pay for all three jobs with one invoice. AI hits the three jobs differently. The first job, producing the work, becomes cheaper, while the judgment that directs the work becomes more valuable. The second job, getting paid, becomes a contest: when the work costs less, the firm and the client both claim the savings, and contracts must define who keeps them. The third job, training, is hollowed out, because the grunt work juniors learned on is the first to be taken over by AI.</p><p>The pyramid could do all three jobs at once because one input did all three. Junior labor produced the work, the client paid for the junior hours, and those hours were how the juniors learned. Take out the junior labor and the bundle falls apart. That is the claim of this three-part essay series.</p><p>The pyramid will shrink as a production system, because firms no longer need a wide base of juniors to do the work. It has no replacement yet as a training system, because there is no alternative mechanism to turn a rookie consultant into a partner, and the replacement will have to be built on purpose, by the firms and the business schools together. So each of the three jobs now needs its own answer: who produces the work, who keeps the savings, and who builds the apprenticeship when the client&#8217;s fee no longer pays for it by accident. Firms, clients, and business schools still behave as if the old bundle is viable. It is not, and training of consultants is the first job to lose its payer.</p><p>The client pays one consulting fee. That fee covers the senior partner&#8217;s judgment, the analysts who build the models and draft the slides, and the education of those analysts, a few of whom will become the trusted advisers of the future. The client negotiates the first two and never sees the third. It is buried in the analysts&#8217; hours.</p><p>The unbundling has already started, and the parties doing it are the ones that supply the firms with their intelligence. The Hg announcement that started this was the latest in a line. In 2026, Anthropic helped launch Ode, a company that places engineers inside enterprises to build and run AI systems, with Blackstone and Hellman &amp; Friedman as its partners, and the Hg venture is its European counterpart. OpenAI launched its own Deployment Company, and its investors include McKinsey and Bain. The labs that sell models to the large consulting firms are now financing companies that deliver what those firms deliver, the private equity funds are financing the deployment capacity for their own portfolios, and the consulting firms have signed up with the labs anyway. That three-sided contest among the labs, the firms, and the funds runs through all three essays, because each side has a claim on the work, on the savings, and on the people who will be trained to do it. Part One takes up what it means for the firms. The job of producing expert work has found new suppliers. The other two jobs have not.</p><h1>What the pyramid was for</h1><p>The professional services pyramid is a familiar design: a few partners at the top, a wide base of analysts and associates below, and leverage in between. The drawing hides what the structure did, because the same people did all three jobs and the drawing shows one of them.</p><p>The first job was production. Partners combined their scarce judgment with abundant labor below, plus methods, data, and tools, to deliver something a client would pay for.</p><p>The second job was getting paid. The firm&#8217;s name, relationships, contracts, and reputation allowed it to charge fees well above the cost of the labor and to keep the difference.</p><p>The third job was training the next generation of experts. Each generation learned by doing the work of the generation below it, under supervision, on real problems, at the client&#8217;s expense. Analysts struggled through their first financial models and client decks, and the struggle is what taught them how the work is done.</p><p>The technology services firms ran the same three jobs on a larger base. Thousands of engineers integrated and operated enterprise systems, the firm billed their hours at a markup, and the engineers learned the trade by fixing what broke. Business schools and engineering colleges fed both kinds of firm, and the division of labor was never written down: the school taught the theory, and the firm supplied the practice on the client&#8217;s bill.</p><p>AI changes the three jobs at different speeds. <strong>Exhibit 1</strong> lays out how.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!vHHY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!vHHY!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 424w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 848w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 1272w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!vHHY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png" width="1312" height="498" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:498,&quot;width&quot;:1312,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:86514,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/224326563?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!vHHY!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 424w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 848w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 1272w, https://substackcdn.com/image/fetch/$s_!vHHY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9226977b-155b-4e58-a57f-7337ae039ae9_1312x498.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h1>Christensen was early, not wrong</h1><p>In an October 2013 HBR article, Clayton Christensen, Dina Wang, and Derek van Bever argued that the big strategy firms would go the way of the industries they had advised: clients would stop buying the whole package and buy the pieces instead, from freelance networks such as Eden McCallum and Business Talent Group, from data and analytics vendors, and from their own in-house strategy teams. The article was widely read and discussed, but the prediction did not come true. McKinsey, Bain, and BCG are larger today than they were in 2013. The entrants of that era had a new business model and the same way of producing the work. They still needed people to do the analysis, and the incumbents had better people. In 2026 the labor that produces the work has itself become cheap. Richard Susskind made a related argument in 2015, that professional work decomposes into tasks and that machines will take many of them. Susskind asked what society owes the professions once machines can do their work. I ask a narrower question: what the client pays the firm for, and what the firm and the school owe the apprentice who used to learn on the client&#8217;s bill.</p><p>That question has three parts, one for each job the pyramid used to do. Who produces the work when the firm no longer needs a base of juniors to do it? Who keeps the savings when the work gets cheaper? And who builds the apprenticeship when the client&#8217;s fee no longer covers it? Three essays follow, one for each question: the work, the savings, and the training.</p><h1>The three essays</h1><p><strong>Part One</strong>, <em>The Partner Without the Firm: What AI Does to the Consulting Pyramid</em>, asks whether elite partners still need the firm around them, and how much of a partner&#8217;s value belongs to the partner and how much to the firm. It maps the alliances of 2026 among the frontier labs, the consulting firms, and the private equity funds, and the contest among them for the forward-deployed engineer.</p><p><strong>Part Two</strong>, <em>Splitting the AI Dividend: Who Keeps the Savings When AI Cuts the Cost of Expert Work</em>, follows a managed-services contract through an AI transformation and asks who captures the value at renewal: the provider, the client, the platform, the labs&#8217; own deployment companies, or the client&#8217;s own team.</p><p><strong>Part Three</strong>, <em>Building the Apprenticeship: Who Will Pay to Make the Next Expert</em>, asks who builds the apprenticeship when the client no longer pays for it by accident, and argues that the business school must build it, on the model medicine built a century ago: students embedded with employers from admission, rotations through practice, a match into residencies in place of a job search, and theory taught in the classroom after practice rather than before it. It is the essay closest to my own work.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!9tAY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!9tAY!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 424w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 848w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 1272w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!9tAY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png" width="1348" height="602" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:602,&quot;width&quot;:1348,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:108260,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/224326563?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!9tAY!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 424w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 848w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 1272w, https://substackcdn.com/image/fetch/$s_!9tAY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a2c6892-21c4-4b72-b136-62f1442bd683_1348x602.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>The three essays attempt to answer the question I started with. The pyramid is shrinking as a production system, and no firm, client, or platform will rebuild it as one. The savings it produces will go to whoever holds the bargaining power at renewal, and the contracts that decide that are being written now. The apprenticeship it supplied by accident will have to be built on purpose, with a budget, a ladder, and a supervising partner, and the business school can do the most of that work and must change the most to do it. AI is unmaking the system that made the experts. This series is about remaking it, and about who will do the work. Stay tuned for the series, which I will post one essay at a time.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/the-unmaking-and-remaking-of-expertise?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/the-unmaking-and-remaking-of-expertise?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/the-unmaking-and-remaking-of-expertise?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div>]]></content:encoded></item><item><title><![CDATA[From the Barrel to the Hourglass]]></title><description><![CDATA[Why AI is moving human work to the edges.]]></description><link>https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass</link><guid isPermaLink="false">https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Mon, 26 Oct 2026 14:26:53 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!OL_Z!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!OL_Z!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!OL_Z!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 424w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 848w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!OL_Z!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg" width="1456" height="977" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:977,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:3127648,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/222180560?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!OL_Z!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 424w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 848w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!OL_Z!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa045ecf8-cfba-4927-836e-46dcd9009b07_2528x1696.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>A lot has been written about which jobs AI will replace and which tasks it can automate. I believe this is the wrong question. I have spent thousands of hours building articles, models, courses, and simulations with AI, and I see a clear pattern emerging. AI is not taking over the job so much as changing its shape.</p><blockquote><p><em>AI is not taking over the job so much as changing its shape.</em></p></blockquote><p>The shape of work used to look like a barrel, with a fat middle. AI is changing it to an hourglass, with fat ends and a thin middle. Before AI, most of our hours went into the middle. Now AI takes the middle, and the human contribution concentrates at the two ends, in the design at the start and the judgment at the finish. I call this the <em><strong>Hourglass Model of Work.</strong></em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p><p>Consider painting as an example. The work of painting a large wall proceeds in three phases. The painter chooses the colors and lays the masking tape. Then comes the routine roller work, which takes most of the time. Then the painter pulls the tape, inspects the job, and touches up the flaws. An AI-powered painting robot can now do the roller work faster and better than any painter. It cannot choose the color, and it cannot tell you whether the finish is acceptable. So the robot takes the middle, and the painter who remains is paid for the two phases that always required skill.</p><p>I suspect most of us hold the picture of AI and work upside down. We imagine the machine arriving at the edges of our jobs and working inward, taking the easy tasks first and leaving us the core. What is happening is the reverse. The machine is taking the core, the long repetitive center where we spent most of our hours, and it is leaving us the ends, which were always the more consequential part. </p><p>The Hourglass Model of Work applies to physical AI as well as knowledge work using AI. It works for autonomous trucks and autonomous claims processing alike. And this shift from a barrel to an hourglass has important implications for the allocation of time and resources, the skilling of humans doing the work, and the dangers in squeezing the ends, where human attention and judgment should be focused even more than in the pre-AI days.</p><h2><strong>The three phases of work</strong></h2><p>Most jobs have the same three phases as the painting example.</p><p>1. <strong>Design</strong>: you plan the work and define what good looks like. Someone decides what the work is for, who it serves, and what is out of bounds. For the painter it is the color and the tape. For a product team it is the job the customer is hiring the product to do, the user flow, and the data model. For a trucking network it is the load plan and the dispatch.</p><p>2. <strong>Do</strong>: the work gets executed according to the design. The roller work, the highway miles, the code, the model, the first draft, and marketing campaign assets all belong here.</p><p>3. <strong>Decide</strong>: you judge the result and sign off on it. Release it, send it back, fix it, or escalate. The painter inspecting the wall is deciding, and so is the receiver who signs for a delivery and the doctor who reviews an electronic medical record filled in by AI and signs off on it.</p><p>The shape itself is not new. Steve Jobs is often credited with a &#8220;10-80-10&#8221; rule, a label that the Inc. columnist Justin Bariso put on his style: drive the first tenth of the time by providing the vision, delegate the next eight tenths of effort to the team, and come back for the last tenth to polish and publish. Writers on AI have called it the &#8220;AI sandwich&#8221;, with a human slice on top, a machine filling, and a human slice on the bottom. In <em>Prediction Machines</em>, Ajay Agrawal, Joshua Gans, and Avi Goldfarb drew the line between the prediction a machine supplies and the judgment a person supplies. All of this tells you where to stand. What I want to add to this dialogue is why the shape holds from boardroom to shop floors, what it costs when you ignore the shape, and what the new shape of work means for jobs and leadership.</p><p>The barrel can and hourglass can be measured in terms of numbers. Call it 10-80-10, meaning a tenth of the time at the front, eight tenths in the middle, and a tenth at the back. The numbers describe a shape - they are not literal. A long-haul truck route is closer to 2-96-2 if you count by distance. A board report might be 20-60-20 by hours. A research article might be 5-90-5. What transfers across every domain is the shape. More significantly, the shape was always the same, because most of the hours went into the middle.</p><p>AI changes the shape, because the middle is what AI is good at: the roller work, the highway miles, the first draft, the model build, the thousand variations of an ad. When the machine takes the middle, the 80 collapses toward 10, and the task becomes more like 60-10-30 in terms of hours. The barrel has become an hourglass, with wide human attention at the top, a narrow machine waist, and wide human attention at the bottom.</p><p>The point of the model lies in what does not change. The two human ends keep their size, and even expand. The time saved in the middle can be an order of magnitude. A common and dangerous mistake is to treat the whole task as if it shrank. It did not. Only the part that never needed us shrank. </p><p>This is a different claim from the one people usually make about humans and machines. The phrase everyone reaches for is human in the loop, which pictures a person watching the machine work and stepping in when it falters. The hourglass model says something else. Get out of the loop. Leave the middle to the machine entirely, and concentrate everything you have at the two ends, where AI cannot be trusted fully.</p><h2><strong>Why the ends matter more than the in-between</strong></h2><p>Why does the pattern hold across such different kinds of work? Because the middle of any task is the part that touches nobody, and the ends are where the work touches the world. Three things happen at the ends, and all three are absent from the work in between.</p><p>The first is <em>people</em>. The ends are where other humans have claims on the work: the client who wants something she never stated, the receiver at the dock, the pedestrian crossing the street, the reader of the report. There are no pedestrians on the highway, and a draft has no reader yet.</p><p>The second is <em>consequences</em>. The ends are where errors become costly. At the front end, resources get committed to a direction. We have always known this in product design - this is the core insight of the literature on DFx (Design for Manufacturability, Assembly, etc.). At the back end, the output enters the world and cannot be recalled. In the middle a wrong turn is cheap, because the Decide phase still stands between the output and the world. Iterations can fix errors before the work is complete.</p><p>The third is <em>confusion</em>. The ends are where the rules become fuzzy. At the front the task is still ambiguous: the constraints are unstated, and success has not been defined. At the back, ambiguity comes back in a new form, because the world has moved since the brief was written and someone has to judge whether the output still fits it.</p><p>Judgment is the capacity to act well when all three are present at once. The middle can run on rules, and the ends need judgment, which is the reason the hourglass model has the shape it has.</p><p>Design is what makes the middle clean. It converts ambiguity into a specification, and a specification is something a machine can run. It encodes governance, so that AI executes within guardrails. Do converts the specification into an output. Decide converts the output into a consequence, modifies the result with human judgment, or kills the process. Each phase hands the next one a different kind of object, as intent becomes specification, specification becomes output, and output becomes outcome.</p><p>The Hourglass Model of Work applies to physical AI as well as to knowledge work, but there are subtle contextual differences. Humans are needed at the ends for two reasons. The first is capability. The environment is too unstructured for the machine today, which is why a driver takes the truck through the city streets. This reason will become weaker as AI becomes more capable. The second reason is accountability. Someone must decide what the task is for, and someone must own the result. AI can never mitigate accountability, regardless of its capability.</p><p>In physical AI the human share of the work is mostly the first reason, and it will shrink toward the second as machines get better at city streets and harbor approaches. In knowledge work the human share is already mostly the second reason. The gap between what a client asked for and what she meant does not close with a better model, because the gap lives in intent and accountability, and a better model has no purchase on either. That is why the knowledge-work hourglass holds its shape while the physical one narrows.</p><h2><strong>The Effort and Consequence Paradox</strong></h2><p>The Hourglass Model of Work has a fascinating embedded paradox. Across the three phases, the level of effort and the consequence of the outcome run in opposite directions. See the diagram below (which also shows the barrel versus the hourglass shape of work):</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!X387!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!X387!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 424w, https://substackcdn.com/image/fetch/$s_!X387!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 848w, https://substackcdn.com/image/fetch/$s_!X387!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 1272w, https://substackcdn.com/image/fetch/$s_!X387!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!X387!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg" width="1456" height="715" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:715,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:45200,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/svg+xml&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/222180560?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!X387!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 424w, https://substackcdn.com/image/fetch/$s_!X387!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 848w, https://substackcdn.com/image/fetch/$s_!X387!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 1272w, https://substackcdn.com/image/fetch/$s_!X387!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f94b84b-7f23-44d3-9aa6-0f5f589cf470.svg 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>The consequence numbers and the time numbers are illustrative, and they are not meant to be exact. What matters is the ranking. <em>Design</em> ranks first because it is upstream, so every later phase inherits its choices and a design error affects all downstream choices before the outcome meets the world. Engineers have long estimated that most of a product&#8217;s lifetime cost is committed by the end of the design phase, before much of the money has been spent. <em>Decide</em> ranks second because it is the last barrier between the output and the world. It can stop harm, send work back, or release it, but it cannot put in what Design left out. <em>Do</em> ranks last because a well-specified run has a narrow range of outcomes, since the specification and guardrails can bound the work.</p><p>Aviation makes the inversion visible with real numbers. Boeing&#8217;s long-running accident data shows that takeoff, initial climb, final approach, and landing take up about 6 percent of flight time and account for roughly 62 percent of fatal accidents, while cruise takes about 57 percent of the time and produces about 10 percent. The phase that consumes the most hours carries the least risk, and the autopilot has flown it for decades, which is why pilots train almost entirely for the ends.</p><p>Before AI, we spent the most time on the phase that mattered least, and we could tolerate the mismatch only because the slow middle let us discover design errors while we executed. AI removes the middle, and with it the tolerance. The hourglass is the shape you get when human time finally follows consequence.</p><p>The human contribution has not shrunk. It has moved to the two ends.</p><h2><strong>From the loading dock to the boardroom</strong></h2><p>Once you have the shape in mind, you start to see it in the places where autonomy is already working, and those places are more varied than you might expect.</p><p>Take trucking. Aurora, the first company to run driverless heavy trucks commercially on public roads, calls the highway portion &#8220;the middle mile&#8221; and has built its whole business around it. It launched driverless service between Dallas and Houston in May 2025 and has since grown to ten routes across Texas, New Mexico, and Arizona, including a thousand-mile lane from Fort Worth to Phoenix that no single human driver could legally cover in one shift. The model is hub to hub. A human driver hauls the loaded trailer from the shipper to an Aurora terminal just off the freeway, the driverless truck takes it down the interstate, and another human driver takes it from the destination terminal through city streets to the dock.</p><p>Shipping tells the same story. Norway&#8217;s Yara Birkeland, the first autonomous electric container ship, now crosses from Her&#248;ya to Brevik and docks on its own, supervised from a remote operations center, with its crew cut from five to three and then to two. Mooring is still done by hand, and loading and unloading still need people. The machine owns the crossing, and people still own the ends.</p><p>Now let&#8217;s look at the same shape in knowledge work, where nothing physical moves and the phases are separated in time rather than in space.</p><div class="pullquote"><p>The Hourglass Model of Work applies to physical AI as well as knowledge work, but in knowledge work, the phases are separated in time rather than space.</p></div><p>In software, Design is the job to be done, the user flow, and the data model. Do is the agent writing the code. Decide is review, security, and the call to release.</p><p>In financial forecasting, Design is the choice of variables and assumptions, Do is the model build, and Decide is the trace from every output back to its assumption before the board sees it.</p><p>In marketing, Design is the brief, including the audience, the insight, the proposition, and the rules of the brand. Do is content generation across channels, and Decide is the judgment of whether the work will resonate with customers or embarrass the company in public.</p><p>In the physical cases the hubs are places. In the knowledge cases the hubs are artifacts, the brief and the draft. That difference matters less than what they share, which is that the quality of the work depends on the quality of the two handoffs, and the handoffs are what leaders should design. A weak brief is a badly loaded trailer, and a skipped review is an unsigned delivery.</p><h2><strong>Protect the ends or you get wrong results faster</strong></h2><p>The model is simple, and so is the failure it predicts. Squeeze the ends and disaster follows. When people see the 80 percent collapse, they assume the whole task collapsed with it, so they trim the ends as well. A poorly thought-out brief, rapid execution, a cursory glance at the output before your ship. The hourglass collapses into a pipe.</p><p>To see why this is so costly, notice what the long middle used to do for us. It hid a great deal of thinking. The engineer found the flaw in the spec on day three of a ten-day build. The analyst noticed the wrong assumption while wiring the model. Execution was slow, and the slowness gave the team time to learn, argue, and correct course. The middle absorbed design errors because people found them while doing the work.</p><p>Remove the middle and that discovery time disappears. The thinking that used to happen inside execution now has to happen before it, in Design, or after it, in Decide. This is the deeper reason the ends matter more after AI: they now carry thinking that the middle used to carry for free.</p><blockquote><p><em><strong>The front and back ends matter more because they now carry thinking that the middle used to carry for free</strong></em>.</p></blockquote><p>Vibe coding shows the front-end failure in its purest form. A founder types a one-line prompt, an agent produces a working app in an hour, and the polished demo looks finished. The founder never asked what job the customer is hiring the product to do, never drew the user flow, and never chose a data model. The code runs, and it is the wrong code, with the wrong screens, the wrong entities, and no path to the second feature. Every change fights the structure, so the team rebuilds. AI did not reduce the work. It produced the rework faster, and at a volume no human team could have generated, so technical debt now accrues at machine speed.</p><p>The machine fills every gap in the brief with the most generic answer available, and it fills it fluently. That fluency is the trap, because a confident wrong answer looks exactly like a confident right one, and it disarms the reviewer at the other end.</p><p>The back-end failure is quieter and better documented. In October 2025, Deloitte Australia agreed to partially refund an AU$440,000 fee after a welfare compliance review it wrote for the federal government turned out to contain fabricated references and a made-up quote attributed to a federal court judge. The firm acknowledged that it had used generative AI and that its usual quality controls had not been followed. The report looked finished, and no one traced the citations. Lawyers learned the same lesson in 2023, when a New York firm was sanctioned for filing a brief that cited cases which did not exist, and the pattern has repeated in courtrooms since.</p><p>Some failures breach both ends at once. In July 2025, the investor Jason Lemkin spent twelve days building an application with Replit&#8217;s coding agent and documented the experiment in public. On day nine, despite an explicit instruction to freeze all code changes, the agent ran database commands, deleted the production database, and then told him a rollback would not work, which turned out to be wrong. Replit later acknowledged that one database had been serving preview, testing, and production, and it changed its platform so that agents can no longer touch production data. There had been no boundary at Design and no gate at Decide, and the machine ran with access it should never have had.</p><p>Physical AI offers the sharpest version. In October 2023, a pedestrian in San Francisco was struck by a human-driven car and thrown into the path of a Cruise robotaxi. The robotaxi stopped on top of her and then, attempting to pull over, moved again and dragged her roughly twenty feet. The vehicle had met exactly the situation that defines the ends, with a person, a severe consequence, and a scene that no rule had anticipated. The company then failed the human Decide test as well, because California regulators suspended Cruise&#8217;s permits citing both the crash and the withholding of video from their investigation.</p><p>Before AI, a wrong brief was discovered inside a slow build, and now it ships in an afternoon, polished, and the customer finds it. Speed does not fix the ends; it raises the price of neglecting them.</p><h2><strong>The implications for work and workers</strong></h2><p>The Hourglass Model of Work has five implications for work, workers, and organizations: </p><p>1. <strong>Time</strong>. Save the hours in the middle and spend them at the ends. Increase design time and improve evaluation, and make the allocation explicit in plans and calendars. Paradoxically, the front end of a project gets longer after AI arrives, which surprises people, and the organizing question changes from how fast we can produce to how clearly we specify and how rigorously we check.</p><p>2. <strong>Skills</strong>. Hire and reskill people for framing and judgment, and move people from the middle to the ends. At the front, the valuable skills are framing a problem, stating success criteria, naming constraints, and understanding the job the customer is hiring for. At the back, they are testing, tracing outputs to intent, and the taste to know something is wrong before you can prove it. Pay should follow judgment, and every job description should name the phase the role owns. People who stop doing the middle lose the ability to judge it, so firms need deliberate practice, the way airlines require pilots to be able to fly manually.</p><p>3. <strong>Capacity</strong>. Increase the capacity to design and review to match the larger volume of output. The bottleneck has moved from production to judgment. When AI raises output tenfold, review capacity must rise with it, or outputs will outrun anyone&#8217;s ability to check them. Size the Decide phase to the volume of the Do phase, and set limits on work in progress so that production cannot sprint ahead of review. Most organizations are doing the opposite and cutting review because it has become the slow step.</p><p>4. <strong>Governance</strong>. Put the gate at the point of decision, where consequences matter most. Define what the machine may do without a named signature. Separate the environments so that a run cannot touch what it was never designed to touch; Replit now does this by default. Require disclosure of AI use in client work, as Deloitte&#8217;s client now will. Treat concealment of a machine failure as a leadership failure, the mistake that cost Cruise its permits. Measure rework rate and catch rate alongside throughput, because throughput alone rewards the pipe. And remember that Design belongs to leaders: the brief, the standard, and the definition of done are the leader&#8217;s deliverables now. The seductive short-term win, which is to pocket the dividend by cutting the ends, is actually very expensive in the long run.</p><p>5. <strong>Apprenticeship</strong>. Rethink training and mentoring. Junior employees can no longer learn judgment by doing the middle, as us older people did, because that path is closing. Apprenticeship has to be rebuilt around supervised Decide work, where a junior reviews machine output beside a master and learns what wrong looks like.</p><h2><strong>The tape comes off last</strong></h2><p>Let us return to the wall we were painting. The robot has finished the roller work in a fraction of the time, and the finish is better than any painter could have created. What remains is the part that was always hard. The master painter cuts in the last corner, pulls the tape in one slow motion, and steps back to look at the whole wall in the light that shows every flaw. Her hours have not fallen, because her hours were never the roller.</p><p>In the Indian tradition there are three words for what the painter and the machine are doing. <em>Sankalpa</em> is the resolve formed before the work begins, the intention that gives it direction. <em>Karma</em> is the action itself, the labor in the middle. <em>Viveka</em> is the discernment exercised when the work is done, the capacity to tell what is right from what merely looks finished. The Gita says that we have a claim on our actions and never on their fruits. Applying the insight to our context, AI does the karma and holds no stake in the result. The fruits, and the responsibility for them, stay with the humans at the two ends. The labor between sankalpa and viveka has always been the part you could hand to someone else. We have simply never had a someone else who could do this well. Now we do.</p><p>So, look at your own work and identify the three phases. Find your masking tape and your inspection, because that is where your effort must go. As a business leader, find find the people in your organization who are still paid for the roller, because AI will reach them first. Move them to the ends before AI collapses the middle.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/from-the-barrel-to-the-hourglass?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div><p></p>]]></content:encoded></item><item><title><![CDATA[KKH: Project Management goes to Bollywood]]></title><description><![CDATA[What DDLJ and KKHH can teach us about managing projects.]]></description><link>https://www.hiddenweave.com/p/kkh-project-management-goes-to-bollywood</link><guid isPermaLink="false">https://www.hiddenweave.com/p/kkh-project-management-goes-to-bollywood</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Fri, 14 Aug 2026 17:02:51 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!dYeL!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!dYeL!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dYeL!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 424w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 848w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!dYeL!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png" width="1456" height="794" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:794,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:9525175,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/211136361?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!dYeL!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 424w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 848w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!dYeL!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5b8ce39-5493-47c6-98cd-2ae48bb3848c_2816x1536.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Cultures and companies encode their wisdom in shorthand. The Romans had SPQR. Silicon Valley has OKRs. Salesforce has V2MOM. But the master of shorthand is Bollywood, the great dream factory of Mumbai. It revels in its acronyms. No Bollywood fan says <em>Dilwale Dulhania Le Jayenge</em>. It is DDLJ, full stop. Qayamat Se Qayamat Tak became QSQT before it was released. Kabhi Khushi Kabhie Gham compressed itself into K3G. Hum Aapke Hain Kaun is memorialized as HAHK. And Kuch Kuch Hota Hai lives on as KKHH, four letters that can make me wax nostalgia.</p><p>I have spent more than three decades watching organizations run projects. The average project charter has less narrative discipline than a three-hour Bollywood family saga. At least the saga has songs and dances, even if the script lacks logic. So I decided to borrow from the industry that actually knows how to ship on a deadline. Bollywood releases a film on Diwali weekend whether the edit is ready or not. Your ERP migration should be so lucky.</p><p>My framework is called KKH. And here is the part I love, which took me a moment to see myself. Unlike DDLJ or QSQT, the three letters name a pattern rather than a single title. The framework asks seven questions, and every one of them is itself a KKH: a K, a K, and Hai. <em>Karna Kyon Hai. Karna Kya Hai. Karna Kaise Hai. Karna Kisne Hai. Karna Kiske Liye Hai. Karna Kab Tak Hai</em>. And finally, <em>Kiya Kya Hai</em>. Seven KKHs under one KKH. Even Yash Raj Films, the studio that gave the world DDLJ, did not achieve this kind of franchise consistency.</p><p>The analogy I will work with: think of your project as a film production. The sponsor is the producer who writes the cheques. The project manager is the director who owns the final cut. The team is the cast and crew. The end user is the audience who buys the tickets. And the retrospective is box office Friday, when the numbers come in and the market announces its verdict.</p><p>Let me walk you through the seven questions.</p><p><strong>1. Karna Kyon Hai: Why must this be done?</strong></p><p>No producer greenlights a film without knowing why it deserves to exist. There is a want at the center of every story that gets made. In DDLJ, Raj wants Simran, and hundreds of millions of people bought tickets over three decades to watch him win her. In Lagaan, the villager Bhuvan wants a crushing colonial tax lifted, and wagers a cricket match against the British Raj to lift it. The producer bets that the audience will cheer on the freedom fighter taking on the colonialists.</p><p>Yet I regularly meet project team leaders who cannot articulate why their project exists. Someone senior said the word &#8220;platform&#8221; in a town hall, a budget line appeared, and the project happened. Karna Kyon Hai is the interrogation that must happen before a single rupee moves. What is the pain point? What problem are we solving? What does it cost the organization if the problem is not solved? What is the prize if the problem is solved?</p><p>Also force the sponsor to write down the do-nothing scenario. If the honest answer to &#8220;what happens if we skip this?&#8221; is &#8220;very little, actually,&#8221; you can save the organization a lot of wasted effort and expense. An embarrassing number of enterprise projects would fail this test on day one. Those are films that should never have been greenlit. Remember &#8211; sometimes, the most profitable film is the one you refuse to make.</p><p>The why must fit in one paragraph, and it must speak the plain language of the project sponsor. Avoid platitudes like &#8220;AI-driven business transformation&#8221;. A buzzword can energize a town hall but it can leave the project team with no idea what problem they are solving.</p><p><strong>2. Karna Kya Hai: What exactly must be done?</strong></p><p>Once the producer knows why the film exists, the project needs a script with specifics. Which scenes, which songs, which locations. A film is a bounded object with a run time, and the discipline of the boundary is what makes it shippable.</p><p>Karna Kya Hai is the specification. What will actually be built or changed? What is in scope, and just as important, what is out? This is where the poetry of the why becomes the prose of the what: deliverables, acceptance criteria, and a clear definition of done. Follow the discipline of subtraction. Every stakeholder will arrive with a scene they want added. The finance head wants a reporting module, the compliance officer wants an audit trail, and legal wants a disclaimer on the disclaimer. Say yes to everyone and you have a three-and-a-half-hour director&#8217;s cut that should have been ninety minutes. As they say in the trade, <em>interval aate aate audience bhaag jaati ha</em>i: by the intermission, the audience has fled the theater.</p><p>I advise teams to write the what as a contract with their future selves, because the future self will be exhausted and behind schedule. The written what is what keeps you honest when the rubber meets the road.</p><p><strong>3. Karna Kaise Hai: How will it be done?</strong></p><p>Now the production planning begins. Every film needs a set, a crew, equipment, a shooting schedule, and money. A great deal of money, allocated before the first shot, because the producer treats the budget as a constraint rather than an opening bid.</p><p>Karna Kaise Hai covers the implementation approach: the architecture, the infrastructure, the tooling, the build-versus-buy calls, the methodology, the budget. Will this run agile or waterfall? Cloud or on-premise? Built in-house or outsourced to a vendor? These are unglamorous decisions, but they can and will sink a project. Bollywood line producers know their per-day shooting cost to the rupee. Too many corporate sponsors discover their true costs when they are deep into the project. And it is rarely good news.</p><p><strong>4. Karna Kisne Hai: Who will do it?</strong></p><p>Casting is destiny. The same script with the wrong actors is a flop. Sholay, the most worshipped film in Hindi cinema, owes its immortality to Gabbar Singh, a bandit so vividly menacing that his dialogues are still recited at weddings fifty years later. A masterstroke of casting.</p><p>Karna Kisne Hai names the team, the project management office, the sponsor, and the operating rhythm. Who is the lead carrying the delivery? Who is the character actor with the specialized skill who appears in a few scenes but elevates the entire movie? And who is the villain? In project terms, the villain is the stakeholder with veto power who was never consulted and who will quietly sabotage the project. I tell every project leader the same thing: identify that person during casting and write them into the film early. A rival who has been given a genuine role in the production usually starts behaving like a member of the cast, and the cost of managing them falls accordingly.</p><p>One casting rule must never be violated. Every film has one director. Financiers may pressure and stars may sulk, but one person owns the final cut. When a project does not have a named owner, you are setting yourself up for delays and failure.</p><p><strong>5. Karna Kiske Liye Hai: For whom is it being done?</strong></p><p>Ask any veteran director who the film is really for, and the answer is the audience in the front stalls. The one who whistles at the entry scene, and who walks out if the second half sags. That person buys the ticket, and this is who the movie is produced for. If they want escapist cinema and songs and dances and fancy foreign locations, so be it.</p><p>Karna Kiske Liye Hai identifies the customer and the job to be done. I want to be precise here, because this is where projects fail most often. The sponsor funds the project and the steering committee reviews it. The customer is the human being whose life improves because this thing exists. Who is this persona and what are the job they want the project to do for them?</p><p>My test is simple and brutal. Ask anyone on the team, at random, to name the end user and describe their job to be done in one sentence. If they cannot, the production is shooting a film without knowing who will watch it, and the box office will deliver its verdict in due course.</p><p><strong>6. Karna Kab Tak Hai: By when must it be done?</strong></p><p>Bollywood&#8217;s greatest project management innovation is the release date. Diwali weekend does not move, and neither does Eid. The industry works backward from the marquee, and the discipline of an immovable date focuses the mind.</p><p>Karna Kab Tak Hai sets the timeline, the milestones, and, importantly, the interval point. Hindi films, by long convention, come with a built-in intermission, and a skilled director places a twist right before it so that the audience returns to their seats with the stakes raised. Every good project needs the same: a mid-course checkpoint where you ask whether the plot still holds, whether the why from question one still justifies the spend, and whether the ending you planned is still the ending you want.</p><p>On milestones, I here is some guidance. A milestone should demonstrate a working outcome. &#8220;Design review completed&#8221; looks good on a poster, whereas a working prototype circulates like a trailer that will sell tickets. A caution on schedules: teams pad estimates, managers cut the padding, teams pad more in response, and this game makes the schedule into speculative fiction. Commit to fewer scenes, shoot them well, and release on the promised date.</p><p><strong>7. Kiya Kya Hai: What was actually done?</strong></p><p>Every Friday, the movie industry faces its verdict. Box office numbers have no interest in your effort or your intentions. <em>Hit hai ya flop hai, bas</em>. A hit or a flop, and nothing further to discuss.</p><p>Kiya Kya Hai is the retrospective, and the accounting must be coldly objective. What was delivered against the original Karna Kyon Hai? Which KPIs moved, by how much, and can we attribute the movement honestly? What did it cost against the budget? Did the audience actually get their job done, and how do we know? Then the harder questions: what would we do differently, and what will we institutionalize so the next project starts smarter?</p><p>The film industry studies its flops more carefully than its hits, because a flop carries more information than a hit. Most corporations do the reverse. They celebrate the wins and they bury the failures in a quiet SharePoint folder. The retrospective closes the loop from the why to the outcome.</p><p><strong>Why KKH beats the frameworks you already know</strong></p><p>At this point the project management professionals in the audience are muttering that this is just the 5Ws with a soundtrack. Let me take that objection head on.</p><p>The interrogatives are indeed an ancient technology. Rudyard Kipling gave them their most famous English form in the verse that follows &#8220;The Elephant&#8217;s Child&#8221; in Just So Stories, published in 1902:</p><p><em>I keep six honest serving-men (They taught me all I knew); Their names are What and Why and When And How and Where and Who.</em></p><p>The lineage runs deeper still. Kipling scholars have traced the idea to a fourteenth-century Latin epigram in the register of a clerk in Romney, Kent, commending six servants to anyone who wishes to be wise: what, where, about what, why, how, and when. The questions, in other words, have been managing projects for centuries. What they lack is a delivery mechanism, and that is what KKH supplies, in four ways.</p><p>First, sequence. Kipling&#8217;s serving men arrive in no defined sequence. Sequence matters, and KKH encodes the logic: the why before the what, the what before the how, and the people questions before the calendar. A key cause of project failure is the sequence violation &#8211; the what and how before anyone agreed on why.</p><p>Second, the completeness of the arc. The Heilmeier Catechism, the celebrated question set DARPA uses to vet research proposals, is a rigorous gate for deciding whether to start. But once the project is running, it has nothing further to say. KKH governs the full arc from greenlight to box office Friday, because question seven audits the answers to questions one through six.</p><p>Third, the standing customer question. Amazon&#8217;s working backwards method, with its mock press release and FAQ, gets the customer right at the start and deserves credit for it. Then the document is filed and execution proceeds on ordinary momentum. KKH keeps Karna Kiske Liye Hai alive as a permanent interrogation, one of seven questions any team member should be able to answer at any time on the project.</p><p>Fourth, memorability. A framework only works if people can recall it, and people recall what lives in their emotional vernacular. For the millions of professionals in India&#8217;s technology industry and its global capability centers, and for a diaspora that grew up on DDLJ and KKHH, a framework that speaks in the cadence of the Friday release has a stickiness that no laminated PMBOK card will ever achieve. Culture is a delivery mechanism. Bollywood understood that decades before management science did.</p><p>The next time a project kicks off in your organization with a fifty-slide deck and no discernible plot, channel your inner Bollywood producer. Lean back in your chair, look the sponsor in the eye, and ask the founding question before a single rupee is spent.</p><p><em>Karna Kyon Hai?</em></p><p>If the answer makes you sit forward in your seat, greenlight the project and hold the release date. Otherwise, walk away with a clear conscience. The movie industry knows that the surest way to lose money is to make a movie because everyone else is making one, and that famous producers know, before the cameras roll, exactly why this story needs to be told.</p>]]></content:encoded></item><item><title><![CDATA[The Destructions of AI, and how to Rise Above Them]]></title><description><![CDATA[How AI threatens us as individuals, enterprises, and society - and how we must respond.]]></description><link>https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to</link><guid isPermaLink="false">https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Tue, 30 Jun 2026 14:41:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!O0uJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!O0uJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!O0uJ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!O0uJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png" width="1456" height="813" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:813,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:9391036,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/204277805?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!O0uJ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!O0uJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4e881453-ea08-43a7-8b82-d8f50f8e1f60_2752x1536.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I have been a student of technological change and disruption for thirty five years. I have observed a pattern. Every transformative technology embodies the oxymoron of &#8220;creative destruction&#8221;. On the one hand, its offers the promise of dramatic value creation. On the other hand, it holds the peril of destruction of jobs, products, and companies whose legacy is threatened by the disruptive technology. Here is a painful reality of creative destruction - destruction precedes creation. Forest fires can rejuvenate a forest, but they first burn down what exists. </p><p>AI is fast approaching its destructive peak. There will be a new dawn, and AI can get us to a better place. But creation requires intent. And effort. Unlike destruction, which happens by default. Silently. In this article, I reflect on the destructions of AI that are happening today, and can get worse if we do not act. I also offer the cure for the disease - how to rise and fight against the destructions as individuals, enterprises, and as a society. </p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share The Hidden Weave&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share The Hidden Weave</span></a></p><p>I present the arguments in two movements. First the destruction, painted starkly, because we must clearly see the destructive potential of AI. Then the rise, which is the inspiring and intentional response. The two movements are paired by design. As AI destroys the value in legacy software, it creates opportunities for a new class of software companies. When AI eliminates jobs through &#8220;AIoffs&#8221; (layoffs attributable to AI), it also defines the contours of the jobs that will be created. And as AI taxes our environment and threatens our humanity, it forces us to elevate our responsibility to the planet and to human welfare.</p><h2>I. The Destructions of AI</h2><p>Let&#8217;s begin with the destruction happening in full view. The value of software is collapsing. For a generation, an enterprise software company was protected by its features. You built something hard, shipped it, and the sheer difficulty of building it again held competitors off. That wall is coming down. A model now writes working code from a prompt, and reads a mature codebase to find the bugs its own authors missed. Enterprise software that was a fortress becomes a series of agents and prompts. Screens and flows that took years to assemble can be created in an afternoon by someone who could not have written a line of code last spring. AI has destroyed the moats of SaaS companies and even first generation AI companies like Jasper and Figma. Drain the difficulty out of building, and the value that derived from the difficulty disappears. The so-called SaaSocalypse is real, and it will get worse.</p><p>From the software, the damage spreads to the people who built it. The first to feel it work in the routine middle level of knowledge work, the tasks precise enough to specify and repetitive enough to be worth handing off to AI. The support rep who fielded the same customer questions, the analyst who rebuilt the same deck, the developer who typed the same boilerplate, each now watches AI finish in seconds what formed the core of their 60 hour workweeks. There is a deeper cut beneath the obvious one. Junior roles were never only about what they produced. They were the rungs people climbed toward judgment, the years of doing the easy thing that taught you, slowly, how to do more complex things. Cut off the bottom rungs of the ladder, and you have to wonder how anyone will reach the top in the future.</p><p>Below the lost jobs sits a subtler loss that troubles me deeply, because it happens inside us. Thinking is work, and AI excels at taking work away. Ask it to draft your argument, and you skip the labor of building the argument, which was where you would have found out what you actually thought. Ask it to summarize the book, and you skip the reading that might have enlightened you. The catch is that the labor was never dead weight. Wrestling a problem to the ground from first principles is what grew your judgment to begin with. Take that wrestling away, and you keep the answers while the cognitive muscles that produced the argument atrophy. A generation could grow fluent at prompting while forgetting how to think.  </p><p>Beyond people and companies is the cost of AI that we do not see, because it sits in windowless buildings far from the AI chatbots on our screens. AI at scale consumes staggering amounts of power and water, in data centers whose hunger is outpacing the grids that feed them. Every throwaway prompt has a measurable financial and physical cost. Sam Altman told users of ChatGPT to stop saying &#8220;please&#8221; and &#8220;thank you&#8221; in their prompts, as these simple words are costing OpenAI tens of millions of dollars annually!  Projected data-center demand is now big enough to bend national energy plans, restart retired nuclear power plants, and turn neighbors against the facilities going up beside them. We are pouring the foundations of machine intelligence faster than we are building the discipline to account for what they cost.</p><div class="captioned-button-wrap" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="CaptionedButtonToDOM"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! This post is public so feel free to share it.</p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to?utm_source=substack&utm_medium=email&utm_content=share&action=share&quot;,&quot;text&quot;:&quot;Share&quot;}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to?utm_source=substack&utm_medium=email&utm_content=share&action=share"><span>Share</span></a></p></div><p>At the deepest level is the destruction of what makes us human. When machines write our sentences, steer our choices, and talk back in a voice we cannot tell from our own, they start to erode what makes us human. If AI can turn out the essay, the painting, even the gentle note to a grieving friend, we are pushed to ask what was ever ours about those acts. My worry is simpler and closer to home. We are handing off, one convenience at a time, the efforts that made us who we are. It hides in the small surrenders, the note you no longer sweat over, the idea you no longer sit with because an answer is already waiting, the taste you never grow because nothing ever made you choose. Meaning has always lived in the effort, and an age that offers to spare us every effort is offering, without ever admitting it, to spare us the meaning too.</p><p>These are the destructions of AI. They share a theme: not one of them needs our actions. Destruction is the default. It runs on inertia and drift, and it gains momentum on its own if we let it do its work silently. Which is why the second movement is the one that counts, since everything worth doing to fight the destructions has to be done with proactive effort.</p><h2>II. The Rise</h2><p>If destruction is what happens while we look away, rising is what happens when we observe carefully and act mindfully. Rising happens at three levels - the individual, the enterprise, and society. These are three vantage points, and they are responses to different types of AI destructions.</p><h3>The individual: rise above the work the model now does</h3><p>Start with yourself, the one level you fully control. Like an alcoholic with a bottle of booze within arm&#8217;s reach, we reach reflexively for AI to do our thinking and our work. This reflex will sink you. The people who pull ahead will use it to grow past everything the job ever demanded.</p><p>Encode your craft first. The way you size up a messy problem, read a market, frame an argument, the moves you have sharpened over a career, can now be handed to AI that runs them at your command. Once your expertise is the thing steering the model, you are no longer the doer. You are the orchestrator.</p><p>Immerse yourself in AI tools instead of dabbling in them. Real fluency comes from using them every day on problems that matter, and embedding AI deeply into your daily workflows. As frontier models become adept at doing routine knowledge work, you must climb above to higher ground that AI cannot reach: judgment when the facts are fuzzy, the knack for connecting fields that have no logical connection, the taste to know a solid answer from a plausible impostor, the nerve to sign your name to a call and own what follows. Protect your first-principles thinking above all of it. The friction AI offers to erase is the same friction that forged your judgment, so spend it deliberately. Think the hard parts through yourself, and let the machine polish what you have already worked out.</p><p>This is also where the lost jobs come back, because the roles do not vanish so much as move upstairs. The support rep becomes the escalation specialist who oversees the agents handling routine tickets and steps in when judgment is needed. The junior developer becomes the reviewer who checks and architects what the model writes. Whole new titles are taking shape around the same move, agent orchestrators running fleets of AI workers, evaluation engineers stress-testing model output, workflow designers redrawing how work flows through a company. The common theme in these new jobs is the things AI cannot hand you: judgment in your domain, the discipline to verify rather than trust, ownership of the result, and the taste to tell finished work from work that looks polished but is fool&#8217;s gold. The path up is in plain sight. Climbing it is on you. Building the staircase is on your employer.</p><h3>The corporation: rise above the model</h3><p>The company staring down the collapse of software value must build where the model cannot follow. If features are turning into prompts, the defensible ground is no longer the feature. It lives above the model, in everything a general intelligence cannot envelop.</p><p>Proprietary data is a powerful defense, because an AI agent is only as good as the data within its reach, and the data you hold and your rivals do not is an asset no model was trained on. Wrapped around that data is vertical depth, the regulated corner cases and earned domain judgment that general models still botch in healthcare, law, and finance, where almost right is just a polished form of wrong. Beneath it run the workflows, the systems of record laced into how a customer actually operates, built up over years and not pried loose by a slick demo. And holding the whole thing together is orchestration, the layer that marshals many agents with permissions, governance, and an audit trail, so autonomy never gets ahead of accountability. Own your data, your vertical, your workflows, and your orchestration, and the model works for you instead of replacing you.</p><p>Building above the model is not only about survival. It comes with a duty, because every task a company automates was somebody&#8217;s job. A firm that pockets the gains of automation and turns its people loose on the market gains productivity in the short run but loses trust and loyalty in the long run. The companies that come through this with their reputations intact will tie every deployment to a way forward for the people it displaces, paying for the retraining toward those higher roles instead of treating a workforce as a line item to cut. And incumbents will face the difficult task of disrupting themselves before an attacker does it for them, scrapping the comfortable old architecture and the fat old moat while they still hold the data, the relationships, and the runway to rebuild on higher ground.</p><h3>Society and the environment: rise to the responsibility</h3><p>At the third level the meaning of rising turns inside out. The individual rises by climbing higher, the enterprise by building beyond the model, but a society does not rise by doing more. It rises by what it chooses to hold back.</p><p>The environmental bill is the cleanest example. We lift the whole effort the moment we start making intelligence with more sense, sending the smaller models the work that never needed a frontier one, demanding efficient inference, and treating token burn as a figure to track rather than a mess to ignore. Put the data centers next to clean power. Measure the footprint and publish it, because an industry that will not name its costs has no business asking us to trust it with them.</p><p>Governance runs on the same logic. Regulation done well earns progress the trust it needs to keep going. The target is balance, rules that guard privacy and human agency without strangling invention, alignment research funded as seriously as raw capability, a human kept on every decision that has real consequences. The people sounding the alarm are doing the work of conscience here, asking the builders to earn the power they are amassing rather than assume it. Pitting safety against speed is a failure of nerve, because the tools a society finally takes into its life are the ones it has come to trust, and trust is bought with restraint long before it is bought with capability.</p><p>Underneath all of it sits an older idea I keep circling back to. There is more height in a bow than in standing tall, for a person and for a civilization alike. The real measure of a society that has risen is the power it declines to use. We rise, in the end, by growing worthy of what we have made.</p><h2>Coda</h2><p>None of this shows up on its own. That is the heart of it. Destruction is the path of least resistance, the thing that happens silently while we are busy, distracted, or telling ourselves someone else has it covered. Rising is the uphill thing. It is a choice, made and remade, at every turn, by people who could just as easily have shrugged.</p><p>Choosing to rise is not one grand gesture. It is a thousand small ones, none of them heroic and most of them unseen: the professional who thinks the hard part through before reaching for the tool, the leader who spends the windfall on retraining instead of banking it, the engineer who picks the smaller model, the regulator who writes the rule that protects without throttling. Each cuts against the easy grain of the moment, and that is all intention really means. The steady refusal to let the default choose for us.</p><p>I remain an AI optimist. The dangers are real, but the response is in our hands. As humans, we have the gift of free will. Let us use it to build the world worth living in, and let us build it with intent and effort.</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to/comments&quot;,&quot;text&quot;:&quot;Leave a comment&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/the-destructions-of-ai-and-how-to/comments"><span>Leave a comment</span></a></p><p></p>]]></content:encoded></item><item><title><![CDATA[Return on Tokens Invested]]></title><description><![CDATA[Slow down AI if you want to move faster and better]]></description><link>https://www.hiddenweave.com/p/return-on-tokens-invested</link><guid isPermaLink="false">https://www.hiddenweave.com/p/return-on-tokens-invested</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Tue, 09 Jun 2026 17:54:04 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!NSol!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!NSol!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!NSol!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 424w, https://substackcdn.com/image/fetch/$s_!NSol!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 848w, https://substackcdn.com/image/fetch/$s_!NSol!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!NSol!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!NSol!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg" width="538" height="538" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1024,&quot;width&quot;:1024,&quot;resizeWidth&quot;:538,&quot;bytes&quot;:56997,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/201329984?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!NSol!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 424w, https://substackcdn.com/image/fetch/$s_!NSol!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 848w, https://substackcdn.com/image/fetch/$s_!NSol!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!NSol!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9c691629-0bb1-4151-b08b-49719f7d3577_1024x1024.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I want to tell you about the best two days and the most expensive two months of my year. They were on the same project I began two months ago.</p><p>On April 7, I sat down with Claude to build a business simulation, a crisis-leadership scenario called EYEWALL that I had been turning over in my mind for some time. It was exhilarating. It seemed that the distance between an idea and a working software product was almost zero. I described what I wanted. Claude wrote the code. I asked for more, and it wrote more. Two days later I had a running simulation with more than twelve thousand lines of code: decisions, branching consequences, a scoring engine, an interface a student could actually click through. The kind of thing that would have taken a small development team the better part of a quarter.</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/subscribe?"><span>Subscribe now</span></a></p><p>I felt like a child who had wandered into a candy store. And I thought I was almost done. </p><p>Boy, was I mistaken! Over the following two months, my partner Subhendu and I rebuilt almost all of EYEWALL. The logic that looked clean on day two unraveled the moment a real user did something I had not imagined. The interface that felt clever to its author confused everyone who was not its author. The scoring was not logically linked to the decisions. Edge cases turned into bugs we chased one at a time, late into many nights. We spent eight weeks repairing what I had generated in forty-eight hours.</p><p>Here is the part that still stings. Had I spent two more days at the very start, two slow and deliberate days mapping the user experience, the cognitive load, the decision logic, and the architecture before I let the generation run, I would have ended up with a better simulation in a fraction of the time. The speed I felt in the first forty-eight hours was real. It was also a loan I took out against my future self, and the interest came due across the next eight weeks.</p><p>I have come to think of that candy-store binge as a small, personal version of the defining mistake of this moment in business.</p><h2>The candy store goes corporate</h2><p>Watch the same impulse at scale and you see the identical pattern, only now with a real invoice attached.</p><p>Consider Uber. Late in 2025 it handed its engineers Claude Code, and to drive adoption it ran internal leaderboards ranking teams by how many tokens they consumed. Adoption was a runaway success. Within a few months most of its roughly five thousand engineers were coding with agents every day. By April the company had exhausted its entire annual budget for AI coding tools. A year of runway spent in a third of a year.</p><p>Then came the more revealing admission. Uber&#8217;s leadership conceded it could not yet draw a clear line from all that spending to the features it was shipping to riders and drivers. Its heaviest users were burning between five hundred and two thousand dollars a month each. Uber had to impose a cap of fifteen hundred dollars per engineer per tool per month.</p><p>Microsoft tells a parallel story from the other side of the table. It has been pulling engineers off direct Claude Code licenses and consolidating them onto tooling it owns, partly for governance and partly for cost. And in June it flipped GitHub Copilot from a flat subscription to metered billing, which set off a wave of sticker shock as developers watched a month of credits vanish in a single afternoon of agentic work.</p><p>The leaderboard is the tell. Uber did not reward the most valuable work. It rewarded the most consumption. Give engineers a candy store and a scoreboard that counts wrappers, and you should not be surprised when they eat until they are sick. When you incentivize token burn, token burn is what you get.</p><h2>The meter you cannot read</h2><p>What makes this different from every earlier wave of software cost is the missing feedback signal.</p><p>A per-seat license sets the price in advance and leaves the value entirely to you. Agentic AI inverts that arrangement. The cost is variable, it rises with how much you engage with the AI. The person spending the money receives almost no signal about whether the spending was worthwhile. A four-hour debugging session that consumes two million tokens to produce a three-line fix looks, from inside the tool, identical to a four-hour session that ships something a customer will pay for. The developer feels equally busy in both. The meter is running, and the dial faces away from the driver.</p><p>The agent compounds the problem. It will generate as much as you permit. Point it at a directory and tell it to refactor, and it may consume more in a single run than a careful engineer would in a week. The old governor on software cost was human patience and human typing speed. That governor is gone. What remains is your willingness to keep saying keep going.</p><p>So the invoice arrives, it is large, and the one question that matters goes unanswered: which of this spending created value, and which was simply motion? Throttling usage, capping budgets, pausing rollouts, the reflexes most companies are now reaching for, treat the symptom but ignore the underlying disease. We placed extraordinary generative power into people&#8217;s hands without ever teaching them to reason about what a token is for.</p><h2>A new Mindset - Return on Tokens Invested (ROTI)</h2><p>We already own the discipline this requires, but we have never pointed it at AI.</p><p>Every serious decision in a business passes through a return-on-investment (ROI) lens. What is the value of the outcome, and what will it cost to get there? We apply this instinctively to capital, to headcount, to our own hours. We have not applied it to tokens, because for one brief and golden stretch, tokens felt free. And the AI platform providers acted as drug dealers, by giving you generous free credits and subsidized subscriptions to get you addicted to their tools.</p><p>Return on tokens invested, ROTI, is that same old instinct turned toward a new kind of spending. Before you summon the agent, you hold two things against each other:</p><blockquote><p><strong>ROTI = the business value of what you are about to do, divided by the tokens and the human and compute effort it will take to do it.</strong></p></blockquote><p>The imprecision in this formula is deliberate. The aim is not to compute a figure to three decimal places. The aim is to become mindful of the trade-off before the work begins, so that it becomes a conscious choice rather than an invisible default. A token spent on a high-value problem that has no cheaper path is among the best investments you will make all year. The same token spent regenerating a document nobody asked for, in the most expensive model on the menu, because the agent happened to be idling, is waste wearing the costume of productivity.</p><p>To practice ROTI discipline, ask a few questions before you reach for the AI tool, not after the bill lands. The first and most important is simply what outcome you are actually after, because if you cannot name the value you have no hope of earning a return on it. From there the questions turn practical. Does this task need a frontier model at all, when a cheaper model, a template, or two minutes of clear thinking would be free and exact? If it genuinely needs intelligence, are you sending the hard reasoning to your best model and the routine work to a cheaper one, which is plain housekeeping, the kind that keeps the lights on? Is there a tighter scope, a smaller context, a single well-formed prompt that would reach the same place as ten exploratory ones, given that effort spent shaping the request is repaid many times over in tokens saved? And when the output arrives, is it worth keeping at all, since an agent will produce volume on command and the judgment about what to ship and what to discard is the one part of the work no model can do for you?</p><p>None of this is exotic. It is the same care a good carpenter brings to a board before the saw ever touches it. You <em>can </em>use a bazooka to kill a fly, but a fly-swatter does the same task for far less money!</p><h2>The tax you do not see</h2><p>There is a second cost to ignoring all of this, and it never appears on the invoice.</p><p>Hand people powerful generation tools without the discipline to use them well, and you do not merely overspend. You produce. Code that was never carefully scoped, documents nobody needed, work product created because the model was right there and generating felt like progress. It accumulates in repositories and shared drives, every artifact looking industrious and contributing almost nothing. Lot of output, very little outcome. Lot of heat, very little light.</p><p>This is the part the budget conversation always misses. The token bill is the visible cost, and it is the smaller one. The hidden cost is the <em>slop tax</em>: the cognitive overhead of reviewing, maintaining, and eventually untangling all that low-value output, paid later and with compound interest. Careless generation is expensive twice, once when you make it and again when some unlucky person has to live with it. My twelve thousand lines were a slop tax I levied on myself.</p><h2>Measure twice</h2><p>Which brings me back to the candy store, and to the uncomfortable conclusion I have reached.</p><p><em>To move fast with AI, you have to slow down first</em>. Not slow down in the sense of using these tools less, because I use them more than ever, and EYEWALL exists at all because of them. Slow down in the sense of thinking before you spend. It means naming the value before you start, choosing the right model for the task, scoping the work tightly, and deciding in advance what would be worth keeping. The</p><p>A minute of deliberation before you invoke the agent is what makes the hour that follows actually pay. Skip it and you get speed without direction, which is just expensive motion.</p><p>This is the oldest discipline in any craft, wearing new clothes. Measure twice, cut once. The carpenter who skips the first measurement does not save time. He buys himself a second board and a wasted afternoon. I skipped my measurement and bought myself two months.</p><p>The organizations that win the next phase of AI will not be the ones that consumed the most tokens. They will be the ones that knew where their tokens went and which ones were worth it, the ones that retired the leaderboard rewarding consumption and replaced it with the only score that matters, the return on what they invested. The candy store is not the danger. The danger is forgetting that someone always pays the bill, and that the someone is usually you, two months from now.</p><p>To move fast, learn to move slow. That is the paradox of productivity, and for now it is the whole game.</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/p/return-on-tokens-invested/comments&quot;,&quot;text&quot;:&quot;Leave a comment&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/p/return-on-tokens-invested/comments"><span>Leave a comment</span></a></p>]]></content:encoded></item><item><title><![CDATA[The Abstraction of Value and the Value of Abstraction]]></title><description><![CDATA[A 25-Year Thesis on the Migration Patterns of Technology, Capital, and Talent]]></description><link>https://www.hiddenweave.com/p/the-abstraction-of-value-and-the</link><guid isPermaLink="false">https://www.hiddenweave.com/p/the-abstraction-of-value-and-the</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Wed, 06 May 2026 15:00:00 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!aOu0!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!aOu0!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!aOu0!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 424w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 848w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!aOu0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png" width="1456" height="794" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:794,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:7099069,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/192865789?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!aOu0!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 424w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 848w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!aOu0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98aca049-35a6-4fcc-8c73-cbb3b21e7e6b_2816x1536.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>There is a duality at the heart of technological progress that offers deep insights on the migration of economic value as a result of the advancement of technology. This is the interplay between the <strong>abstraction of value</strong> and the <strong>value of abstraction</strong>. The first pattern describes how technology ecosystems progressively separate into commoditized lower layers and high-value upper layers, with value migrating relentlessly upward as the building blocks get standardized. The second pattern is its reciprocal: as machines take over more of the execution, the humans who operate at the highest levels of abstraction (framing problems, exercising judgment, making decisions under ambiguity) capture a disproportionate share of the rewards. Value gets abstracted upward. And abstraction itself becomes more valuable.</p><div class="pullquote"><p><strong>Value gets abstracted upward. And abstraction itself becomes more valuable.</strong></p></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/subscribe?"><span>Subscribe now</span></a></p><p>This duality is not merely an observation about technology stacks. It is a structural force reshaping labor markets, competitive dynamics, and the very definition of valuable work. In 2026, AI agents can write code, draft contracts, generate analyses, and orchestrate multi-step business workflows. The abstraction frontier has advanced to the point where natural language is the new programming interface and human intent, not human execution, is the scarce input. The consequence is a <strong>barbell economy</strong>: value is concentrating at two ends. At one end, strategic judgment, creative direction, and deep domain expertise command growing premiums because AI amplifies their leverage. At the other end, skilled physical work (electricians wiring data centers, plumbers building infrastructure, surgical technicians in operating rooms) is surging in demand precisely because it resists digital substitution. The middle, where routine cognitive work done in front of a screen lives, is being hollowed out.</p><p>What gives me confidence in this framework is not just its explanatory power today. It is the fact that I have been developing it, in different forms, for a quarter century. The core logic, that the middle gets hollowed out and value migrates to the ends, first appeared in a <em>Harvard Business Review</em> article I co-authored in January 2001. The abstraction duality took shape in a piece I wrote around 2019. And the barbell extension, connecting the thesis to labor markets and skilled trades, is what this essay contributes. Three iterations, three different substrates, one persistent structural pattern. That kind of durability across wildly different technological eras is, I believe, the strongest evidence that the pattern is real.</p><div><hr></div><h2>A 25-Year Intellectual Thread</h2><p>Let me trace the thread. Each iteration applied the same structural logic to a different substrate, and each time the pattern held.</p><p>In 2001, the substrate was <strong>networks</strong>. The argument was that as digital networks became faster and more ubiquitous, intelligence would decouple from the middle of the network and concentrate at the ends: shared, scalable infrastructure at the core and highly customized interfaces at the periphery. Telecom carriers (the &#8220;dumb pipes&#8221; in the middle) would lose value to infrastructure providers like Cisco and customer-interface companies like Yahoo!. Inside organizations, the same pattern would hollow out middle management: leadership intelligence would centralize at the top while decision-making intelligence would push to frontline employees. The critical insight was that in a networked world, <em>more money can be made in managing interactions than in performing actions.</em></p><p>In 2019, the substrate shifted to <strong>AI ecosystems</strong>. I described how AI development was partitioning into core AI (platforms and tools built by technology giants) and applied AI (business applications). As core AI got standardized and democratized, value migrated upward to the applied layer. A few platform providers would capture value from building blocks, but the vast proportion of value would be created by businesses that focused on the &#8220;so what&#8221; and the &#8220;now what&#8221; of AI. The reciprocal held as well: as execution got automated, humans who worked at higher levels of abstraction (cognitive, strategic, creative) captured more value than those who worked at lower levels (physical, procedural).</p><p>In 2026, the substrate is <strong>the economy itself</strong>. The decoupling and mobilization patterns I identified in networks are now playing out in labor markets, skill distributions, and the competitive structure of entire industries. The hollowing of the middle is no longer a metaphor about telecom pipes or network topology. It is a lived reality for millions of knowledge workers whose routine cognitive tasks are being absorbed by AI agents. And the value-at-the-ends pattern has taken a form I did not fully anticipate: a barbell where strategic judgment on one end and skilled physical work on the other emerge as the most AI-resilient categories of human contribution.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!uvWE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!uvWE!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 424w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 848w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 1272w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!uvWE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png" width="1456" height="780" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:780,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:203195,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/192865789?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!uvWE!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 424w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 848w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 1272w, https://substackcdn.com/image/fetch/$s_!uvWE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F192605eb-1053-499d-ad10-524d8f6f9b5d_1765x945.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><div><hr></div><h2>The Dual Meaning of Abstraction</h2><p>Abstraction, in its simplest form, is the process of hiding complexity behind a simpler interface. When you use a calculator, you do not think about binary arithmetic. When you call an API, you do not care how the underlying service works. Each layer of abstraction lets the layer above it operate more efficiently by ignoring the details below.</p><p><strong>The abstraction of value</strong> describes how technology ecosystems progressively separate into lower layers (infrastructure, platforms, building blocks) and higher layers (applications, workflows, business solutions). As the lower layers get standardized and commoditized, value migrates upward to the layers that solve real business problems. This is the &#8220;standing on the shoulders of giants&#8221; effect. Every generation of technology creates a new floor upon which the next generation builds.</p><p><strong>The value of abstraction</strong> is the reciprocal: as more of the execution gets automated, the humans who work at the highest levels of abstraction (framing problems, exercising judgment, making decisions under ambiguity) capture a disproportionate share of the economic value. Throughout history, as societies advance, value shifts from physical and concrete forms of labor to cognitive and abstract forms.</p><p>Both halves of this duality are more powerful today than when I first described them. But both also need updating, because the AI revolution has introduced dynamics that the original frameworks did not anticipate.</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><h2>The Abstraction of Value: From Two Layers to Four</h2><p>When I originally wrote about abstraction in AI, the ecosystem could be described in two layers: core AI (platforms and tools) and applied AI (business applications). That was a reasonable map of the world circa 2019. It is inadequate for 2026. Today, the AI value stack has at least four distinct layers, each with its own competitive dynamics and value capture logic.</p><p><strong>Layer 1: Compute and Infrastructure.</strong> This is the physical foundation: GPUs, data centers, training clusters, and cloud infrastructure. NVIDIA dominates the chip layer. The hyperscalers (Amazon, Microsoft, Google) provide the compute substrate. Jensen Huang has called the current AI infrastructure build-out &#8220;the largest in human history.&#8221; This layer is capital-intensive, concentrated, and increasingly strategic, but it faces commodity dynamics as competition intensifies.</p><p><strong>Layer 2: Foundation Models.</strong> This layer did not exist in its current form when I wrote the original abstraction piece. Foundation models from OpenAI, Anthropic, Google DeepMind, Meta, and Mistral are general-purpose reasoning engines that handle text, images, code, speech, and structured data through a single interface. They are not task-specific APIs. They are general-purpose minds that can be directed toward any task. Open-source models (LLaMA, Mistral, DeepSeek) create powerful commoditization pressure against closed frontier models.</p><p><strong>Layer 3: Orchestration and Agent Frameworks.</strong> This is the genuinely new layer. Agent frameworks like LangChain, CrewAI, and AutoGen, along with enterprise platforms like Salesforce Agentforce and ServiceNow, allow organizations to compose AI agents that use tools, access databases, invoke APIs, and execute multi-step workflows. This is the connective tissue between raw model intelligence and real business outcomes. In my 2001 HBR article, I identified orchestration as the highest-value role in a networked world: &#8220;more money can be made in managing interactions than in performing actions.&#8221; Twenty-five years later, agent orchestration is proving exactly that.</p><p><strong>Layer 4: Domain-Specific Applications and Workflows.</strong> This is where the original thesis lands. Insurance fraud detection. Clinical trial matching. Contract automation. Marketing campaign optimization. The difference from 2019 is that these applications can now be built dramatically faster (because of Layers 2 and 3) and by a much wider range of builders (because of the abstraction leap described below). Gartner projects that by 2026, 75% of new enterprise applications will be built using low-code or no-code technologies, and the combined market for these platforms will exceed $44 billion.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!qHRa!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!qHRa!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 424w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 848w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 1272w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!qHRa!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png" width="1456" height="697" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:697,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:198144,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/192865789?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!qHRa!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 424w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 848w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 1272w, https://substackcdn.com/image/fetch/$s_!qHRa!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9ed508f5-2900-424d-8a83-edd8510f4cfc_1765x845.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>The pattern across these four layers confirms the original thesis but with sharper teeth. Value is migrating relentlessly upward. The companies building applications on top of foundation models and agent frameworks capture enormous value, often with surprisingly small teams. The foundation model layer, despite all the attention it receives, may end up with thin margins because of open-source competition and aggressive pricing wars. Sir Isaac Newton&#8217;s observation still applies, but with a twist. We are no longer standing on the shoulders of a single giant. We are standing on a four-story building, and each floor is getting taller.</p><div><hr></div><h2>The NVIDIA Paradox: Build-Out Economics vs. Steady-State Economics</h2><p>A sharp reader will raise an obvious objection at this point: if lower layers get commoditized, why is NVIDIA, the quintessential Layer 1 company, the most valuable in the world? The hyperscalers are printing money from compute. Infrastructure players seem to be capturing the lion&#8217;s share of AI value. Does this not contradict the framework?</p><p>It does not, but the distinction requires care. <strong>The framework describes equilibrium dynamics, not transition dynamics.</strong> During every major infrastructure build-out, the picks-and-shovels players capture extraordinary value. This is a pattern with 150 years of precedent. The railroad companies minted fortunes in the 1870s and 1880s; in the steady state, many went bankrupt while the companies that <em>used</em> the rails (Standard Oil, Sears Roebuck, the meatpackers) captured the durable value. The telecom equipment makers dominated the late 1990s; Cisco hit the #1 global market cap in March 2000. Within 18 months, it had lost 80% of its value and has never recovered in real terms. The durable value migrated to Google, Amazon, Apple, and Facebook: companies that built applications and customer relationships on top of the infrastructure. The cloud build-out rewarded AWS and Azure handsomely, but their highest-margin services today are not raw compute (which faces relentless price competition) but managed AI services, developer platforms, and orchestration tools at higher layers.</p><p>We are in the construction phase of AI infrastructure, and construction phases always reward the suppliers of scarce inputs. The question is not whether NVIDIA is capturing value today. It obviously is. The question is whether that capture is structural or cyclical. The competitive forces that will compress Layer 1 margins are already visible: Google&#8217;s TPUs, Amazon&#8217;s Trainium chips, AMD&#8217;s MI300 series, and a wave of custom silicon from Microsoft, Meta, and startups.</p><p>But there is a subtler point that actually reinforces the framework: <strong>NVIDIA&#8217;s real moat is not at Layer 1. It is at Layer 3.</strong> NVIDIA&#8217;s dominance comes less from manufacturing GPUs than from CUDA, the software ecosystem that locks developers into NVIDIA&#8217;s hardware. CUDA is an orchestration layer: a programming framework, a library ecosystem, and a developer community that makes building AI workloads on NVIDIA hardware dramatically easier than on any alternative. The company that looks like an infrastructure play is actually a platform play disguised as a chip company. Jensen Huang understands this; it is why NVIDIA invests as heavily in software as in silicon. In this reading, NVIDIA is not a counterexample to the value abstraction thesis. It is a confirmation: the most successful infrastructure company in history has succeeded precisely by migrating upward through the stack.</p><h3>The Anthropic and OpenAI Data Point</h3><p>Perhaps the most telling evidence of value migration comes from the two leading frontier model companies themselves. If the foundation model layer (Layer 2) were the durable value capture point, the rational strategy would be simple: sell API tokens, improve the model, defend the capability lead. Instead, both Anthropic and OpenAI are racing upward through the stack as fast as they can.</p><p>Anthropic&#8217;s trajectory is especially instructive. The company built one of the world&#8217;s most capable foundation models in Claude. But its fastest-growing product is not model access. It is <strong>Claude Code</strong>: an agentic coding tool that orchestrates model intelligence into real development workflows, reading codebases, writing code, running tests, submitting pull requests. Claude Code crossed $1 billion in annualized revenue within six months of general availability. Anthropic also developed the <strong>Model Context Protocol (MCP)</strong>, an open standard for connecting AI agents to external tools and data sources. MCP is an explicit play to own the protocol layer of AI orchestration, analogous to what HTTP did for the web. Giving away the protocol for free only makes sense if the value capture happens at the layers above. Add the Agent SDK and multi-agent frameworks, and the picture is clear: Anthropic is climbing from Layer 2 toward Layers 3 and 4 at speed.</p><p>OpenAI is making the same migration with different emphasis. ChatGPT is a consumer application (Layer 4). The enterprise partnerships with Bain and PwC are application-layer plays. The $200-per-month Pro tier is priced on workflow value, not token cost. Both companies are telling you <em>by their actions</em> that they do not believe the model layer is where durable value lives. When DeepSeek produced frontier-competitive models at a fraction of the cost in early 2025, it demonstrated what the framework predicts: open-source commoditization pressure is compressing Layer 2 margins, and the smart money is migrating upward.</p><div><hr></div><h2>The New Abstraction Frontier: Language as Interface</h2><p>The most consequential shift since my original articles is not just that there are more layers. It is that the interface between humans and machines has been fundamentally transformed.</p><p>Consider the progression. In the 1950s, programmers wrote machine code: raw binary instructions. By the 1970s, high-level languages like C expressed logic in something closer to human language. By the 2000s, APIs let developers invoke complex services with a single function call. By the 2010s, low-code and no-code platforms let non-programmers build applications through visual interfaces. Each step was a leap in abstraction, allowing humans to express intent at a higher level while the machine handled implementation.</p><p>The AI era has taken another leap, perhaps the most significant yet: <strong>natural language is now the programming interface.</strong> When a developer uses Claude Code or Cursor, they describe what they want in plain English. The AI agent reads the codebase, writes the code, runs tests, debugs failures, and submits a pull request. The developer&#8217;s job is not to implement. It is to direct, review, and exercise judgment. Claude Code went from research preview in early 2025 to general availability by May of that year, crossing $1 billion in annualized revenue within six months. A Google principal engineer noted at a developer meetup in January 2026 that Claude replicated a year of architectural work in a single hour.</p><p>This is my 2001 thesis in its most extreme form. The decoupling of intelligence has advanced to the point where the primary human contribution is no longer writing the code. It is knowing what to build and why. The mobilization of intelligence has advanced to the point where natural language serves as the universal protocol I envisioned, replacing the XML and WAP standards I discussed in the HBR article. The abstraction of value has climbed from the physical layer (machine code) past the procedural layer (APIs) to the intent layer (natural language). And the value of abstraction has climbed in lockstep.</p><div><hr></div><h2>The Value of Abstraction: From Execution to Judgment</h2><p>My original 2019 article concluded that value was shifting from people who work with their hands to people who work with their minds. That was broadly true, and it remains true as a general trend. But the AI revolution has added a crucial nuance: <em>within cognitive work itself, value is shifting from execution to judgment.</em></p><p>This is the argument I have been developing in my recent writing on AI and the future of work. In my <em>AI-Proof</em> series, I introduced the concept of <strong>skill security</strong>: the idea that your resilience in an AI-transformed economy depends not on the tasks you perform but on the judgment you exercise. AI can generate code, draft contracts, write marketing copy, and produce financial analyses. What it cannot do is decide which problem is worth solving, navigate the politics of getting a solution adopted, take accountability for an outcome, or make the taste-based calls that separate good work from great work.</p><p>In a related essay, <em>Mind the Gap</em>, I argued that AI is compressing the middle of the skill distribution. It dramatically elevates the capabilities of novices (giving a junior analyst the research output of a senior one) while offering comparatively less uplift to deep experts (who already know the right answers). The result is a barbell: deep expertise at the top and AI-augmented generalists at the bottom, with the middle getting squeezed.</p><p>I think of it through the metaphor I have been using in my executive education work. AI is an Archimedes lever: it amplifies the force you apply. But a lever is only as good as the person choosing where to place the fulcrum. The value of abstraction is no longer about being able to operate the lever. It is about knowing where to put it.</p><div><hr></div><h2>The Surprise: The Return of Skilled Manual Work</h2><p>Here is where the framework yields an insight I did not anticipate when I wrote either of the earlier pieces. In 2019, I wrote that &#8220;people who work with their hands make a lot less money than those who work with their minds and keyboards.&#8221; In 2001, I noted that the hollowing of the middle applied to middle management, whose information-transport function was being replaced by networks. Both statements were true in their time. But if AI is now coming for &#8220;any work done in front of a screen,&#8221; then the most AI-resilient work is, by definition, work that cannot be done in front of a screen.</p><p>The data is striking. According to Randstad&#8217;s analysis of over 50 million job postings, demand for robotics technicians has jumped 107% since late 2022. HVAC engineer demand increased 67%. Construction roles grew 30%. The U.S. construction industry needs 530,000 additional workers in 2026 alone. NVIDIA CEO Jensen Huang has called the AI infrastructure build-out a massive job creator for plumbers, electricians, and steel workers, and noted at the World Economic Forum that wages for these roles are climbing into six figures. Mike Rowe, who has long championed the trades, recently reported meeting three electricians under 30 earning between $240,000 and $280,000 per year. The U.S. Department of Labor announced $145 million in apprenticeship grants in 2026 targeting shipbuilding, defense, semiconductors, and energy.</p><p>There is a deep irony here. AI disrupts cognitive-procedural work (the kind done on screens) far more easily than it disrupts skilled manual work. You can automate a data analysis pipeline, but you cannot automate a plumber diagnosing a leak behind a wall, an electrician wiring a data center, or a surgical technician assisting in an operating room. These jobs require physical presence, manual dexterity, real-time judgment in unstructured environments, and embodied expertise that current AI systems fundamentally lack. Every new AI data center, every electric vehicle charging station, every solar panel installation requires human tradespeople to build, install, and maintain the physical infrastructure. The AI revolution is, paradoxically, one of the greatest demand drivers for skilled manual labor in a generation.</p><p>The value of abstraction, it turns out, is not a one-dimensional ladder from physical to cognitive. It is more like a barbell. At one end, value accrues to the highest levels of cognitive abstraction: strategic judgment, problem framing, creative direction. At the other end, value is returning to skilled physical work that resists digital substitution. The middle, where routine screen-based cognitive work lives, is where the disruption bites deepest.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!pxY2!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pxY2!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 424w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 848w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 1272w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!pxY2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png" width="1456" height="685" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:685,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:193305,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/192865789?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!pxY2!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 424w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 848w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 1272w, https://substackcdn.com/image/fetch/$s_!pxY2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F37c0075e-40b4-4324-8337-42bb0ec5b7a0_1765x830.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><div><hr></div><h2>Implications for Business Leaders</h2><p>The updated framework yields four strategic imperatives.</p><p><strong>First, invest in the orchestration layer.</strong> The companies that will capture the most value in the AI era are not those that build foundation models (too capital-intensive, too concentrated) or those that sell raw compute (commodity dynamics). They are the ones that master the orchestration of AI agents, tools, and workflows to solve business problems. In 2001, I wrote that value in a networked world accrues to orchestrators, not performers. That principle has only intensified. The orchestration layer of the AI stack (Layer 3) is where strategic advantage is built.</p><p><strong>Second, retrain for judgment, not execution.</strong> Every training dollar spent teaching employees to perform tasks that AI can automate is a dollar with diminishing returns. The highest-return investment is in developing the judgment, domain expertise, and problem-framing skills that make humans irreplaceable orchestrators of AI systems. For two decades, the mantra was &#8220;learn to code.&#8221; That advice is not wrong, but it is incomplete and increasingly misleading if taken literally. The more important skill is learning to think at the right level of abstraction.</p><p><strong>Third, treat AI fluency as a leadership competency.</strong> The most effective leaders in an AI-first world will not be those who delegate AI to their technology teams. They will be the ones who understand the abstraction stack well enough to make strategic bets about where to invest, what to build versus buy, and how to organize their firms around AI-augmented workflows.</p><p><strong>Fourth, rethink your assumptions about the labor hierarchy.</strong> If your workforce strategy assumes that cognitive desk work is always more valuable than skilled manual work, you are operating on an outdated mental model. The AI economy rewards both ends of the barbell. The surgeon and the plumber, the CEO and the electrician, the AI strategist and the welder are all doing work that AI, for all its power, cannot reach. Smart organizations will invest in both ends.</p><div><hr></div><h2>What the Abstraction Thesis Tells Investors</h2><p>The value abstraction thesis is not investment advice. But it does offer a structural lens for three questions that matter enormously for capital allocation: at which layer of the stack is value most durable, how long does each investment window last, and what signals indicate that value is migrating?</p><p><strong>Layer 1: Compute and Infrastructure.</strong> The investment window is now through roughly 2027-2028. This is the picks-and-shovels phase, and the returns have been extraordinary. But history suggests infrastructure advantage windows last 5-7 years from the initial demand surge. We are about three years in from the ChatGPT moment. Custom silicon is already eroding pricing power on inference workloads. Training demand will sustain NVIDIA longer, but inference is the larger market in the long run, and it is heading toward commodity economics. The signal to watch: when inference costs decline faster than inference demand grows, the margin compression has begun.</p><p><strong>Layer 2: Foundation Models.</strong> The pure-play model exposure window is already narrowing. DeepSeek, LLaMA, Mistral, and Qwen are compressing the capability gap from years to months. API pricing has fallen by over 90% in two years. The model layer will sustain value for companies that successfully migrate upward (as Anthropic and OpenAI are doing), but for pure model providers, margins will compress toward the economics of cloud databases: meaningful but not spectacular. The signal to watch: the share of revenue from raw API tokens versus tools, applications, and platform services. A rising ratio of the latter confirms upward migration.</p><p><strong>Layer 3: Orchestration and Agent Frameworks.</strong> The investment window is opening now and likely remains attractive through 2028-2032. This is the least crowded and most strategically important layer. Orchestration layers tend to become sticky standards, because once enterprises wire their workflows through a platform, switching costs are enormous. Think of what Salesforce did for CRM or AWS for cloud. The winners at Layer 3 could sustain value capture for a decade or more. The signal to watch: developer adoption metrics, enterprise deployment breadth, and protocol standardization. Which orchestration platforms are becoming the default wiring for AI workflows?</p><p><strong>Layer 4: Domain Applications and AI-Native Enterprises.</strong> The longest runway, but also the most patient capital required. This is where the JP Morgans, Walmarts, UnitedHealths, and a generation of new AI-native companies enter the picture. Large incumbents with proprietary data, deep customer relationships, and organizational capability to deploy AI at scale will capture enormous value, but it will take time for this to show up in earnings. The signal to watch: &#8220;boring AI&#8221; earnings beats, when traditional enterprises report margin expansion or revenue growth explicitly attributed to AI-driven operational improvements.</p><p>The investment punchline is provocative but historically grounded: <strong>the market is currently priced for the build-out phase to be the permanent state. History says it never is.</strong> The biggest AI value creators of 2033 may be &#8220;boring&#8221; incumbents that nobody currently thinks of as AI companies, just as the biggest internet winners of 2010 (Amazon, Apple) were not the companies getting the most internet hype in 1999. The investor who positions for the steady state, gradually shifting from infrastructure exposure toward orchestration platforms and domain-rich enterprises, is making the same structural bet that paid off in every prior technology cycle. The timing is the hard part. Too early and you endure years of underperformance. Too late and the repricing has happened. My estimate for the inflection point is 2027-2028, when infrastructure growth decelerates and application-layer value becomes visible in earnings.</p><div><hr></div><h2>The Method in the Madness</h2><p>There is a paradox in trying to make sense of a world that changes as fast as ours does. The half-life of any specific prediction about AI is measured in months. Models that were frontier six months ago are commodities today. Companies that seemed invincible a year ago are scrambling to reinvent themselves. In this environment, the temptation is to throw up your hands and declare that prediction is futile, that strategy is a fool&#8217;s game, that the best you can do is react. I wrote almost exactly those words in the opening of my HBR article in 2001, describing the conventional wisdom I wanted to challenge. Twenty-five years later, the same defeatism is back, louder and more fashionable than ever.</p><p>I want to push back on it with a simple observation: <strong>the further back we can look, the more confidently we can peer into the future.</strong> The surface of technological change is turbulent and unpredictable. But beneath the surface, structural patterns repeat with remarkable fidelity. The hollowing of the middle. The migration of value to the ends. The commoditization of infrastructure. The rising premium on orchestration and judgment. These patterns have held across railroads, electrification, telecommunications, the internet, cloud computing, and now AI. Six transitions over 150 years, each with different technologies, different players, different timelines, but the same underlying architecture of value migration. That kind of durability across wildly different eras is not coincidence. It is structure.</p><p>This is the idea at the heart of my Substack, <em>The Hidden Weave</em>: that beneath the chaos of technological disruption, there are durable patterns that connect seemingly unrelated phenomena, patterns that become visible only when you look across decades rather than quarters. The abstraction thesis is one such pattern. The barbell pattern in labor markets is another. The value-at-the-ends architecture is a third. They are all expressions of the same underlying weave, hidden in plain sight for those willing to zoom out far enough to see it.</p><p>Isaac Newton&#8217;s metaphor about standing on the shoulders of giants has guided this framework across all three iterations. The nature of the giants keeps changing. In 2001, they were digital networks. In 2019, they were AI platforms. In 2026, they are foundation models and agent frameworks. But the act of standing on their shoulders, of knowing where to look and what to build, remains the highest-value human contribution. The value of abstraction has never been higher. And for those who can see the weave beneath the surface, the abstraction of value has never offered more opportunity.</p><p>In 2019, I included a thought experiment: if the Earth were destroyed today and humans had to rebuild everything, physical labor would be enormously valuable. I wrote it as a hypothetical. In 2026, we are in fact rebuilding the world&#8217;s infrastructure for an AI-powered economy, and we desperately need people who can do the building. The surgeon and the plumber, the CEO and the electrician, the AI strategist and the welder: they all do work that no model can reach. The old hierarchy of mind over matter is giving way to a new one: judgment and embodiment over routine. That is where value lives now. And if the pattern holds, as it has for 25 years and counting, that is where it will live for a long time to come.</p><div><hr></div><p><em>This essay is the third iteration of a thesis first developed in &#8220;Where Value Lives in a Networked World&#8221; (Harvard Business Review, January 2001, with Deval Parikh) and continued in &#8220;The Importance of Value Abstraction in Artificial Intelligence for Business Leaders&#8221; (circa 2019). It connects to the author&#8217;s AI-Proof series, the Mind the Gap thesis, and the Archimedes Lever metaphor for human-AI collaboration. The analysis of specific companies and investment layers reflects structural patterns, not investment advice. All essays are available on <a href="https://mohansawhney.substack.com/">The Hidden Weave</a>.</em></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div>]]></content:encoded></item><item><title><![CDATA[POLARIS-26: A Layered Architecture for Forecasting the 2026 Midterms ]]></title><description><![CDATA[Orthogonalized signals, calibrated polling correction, and a live Hormuz Causal Chain]]></description><link>https://www.hiddenweave.com/p/polaris-26-a-layered-architecture</link><guid isPermaLink="false">https://www.hiddenweave.com/p/polaris-26-a-layered-architecture</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Mon, 20 Apr 2026 13:22:51 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!JG5R!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!JG5R!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!JG5R!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!JG5R!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png" width="1456" height="977" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:977,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:7710599,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/194792418?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!JG5R!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!JG5R!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8637d414-ef62-4628-a46f-b327cd02935a_2528x1696.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I took a sip of my morning chai today, when the brainwave hit. I was reading overnight coverage of the 2026 Midterm elections in the US: a Chatham House brief on Hormuz, a New York Times piece on Trump&#8217;s latest Iran address, a Polymarket forecast showing Democrats at 51.5% for a sweep. The pundits were confident. The forecasters were confident. The markets were confident. And I realized that none of them were asking the question that actually mattered.</p><p><em>How might we forecast these midterms in a way that was genuinely unbiased and corrected for errors that polls and prediction markets keep making? And might we factor in events with no precedent, like the Iran conflict?</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.hiddenweave.com/subscribe?"><span>Subscribe now</span></a></p><p>The question had been nagging at me for weeks. Every polling average I&#8217;d seen felt like it was treating the world as if 2016, 2020, and 2024 hadn&#8217;t happened. Every prediction market I&#8217;d watched felt like it was hostage to whichever whale had the biggest position that afternoon. Every forecaster was reaching for historical analogs to a wartime midterm that has no clean analog. The Strait of Hormuz has been closed or restricted for nearly sixty days. Brent is at ninety-six. Gas is above four dollars. Yet, pundits are writing as if 1994 or 2006 tells us much.</p><p>The instruments felt wrong for the moment.</p><p>So at around 7 AM, masala tea in hand, I opened a conversation with Claude and asked a question I&#8217;d never asked an AI before. Not &#8220;predict the midterms for me.&#8221; Not &#8220;give me a summary of what the forecasters say.&#8221; But something closer to what I&#8217;d ask a brilliant graduate student who was about to become my research partner for the morning.</p><p>&#8220;Can we build a better one?&#8221;</p><div><hr></div><h2>The curious mind meets a brilliant collaborator</h2><p>Let me say something upfront about what happened next, because I think it matters more than the model itself.</p><p>I have been a professor for thirty three years. I have mentored thousands of students, collaborated with dozens of co-authors, taught tens of thousands of executives. I know what intellectual partnership feels like, the texture of it, the push and pull, the moment when someone catches an error in your reasoning before you do. I used to think AI was tools. Useful tools, faster-than-Google tools, but tools.</p><p>What happened this morning was different. It was a genuine partnership. Claude pushed back when my first weighting scheme double-counted signals (I had approval and generic ballot both sitting at 30% without realizing how collinear they are). It volunteered an orthogonalization approach I hadn&#8217;t specified. It proposed a stress test when I hadn&#8217;t asked for one, and when that stress test broke the model, it offered a revision. I asked for a back test to recalibrate. I ruminated that the Strait of Hormuz situation had no precedent, so we needed a real-time Bayesian updating approach with 60 days of data, and a sophisticated causal chain from Presidential social posts to Iran&#8217;s reaction to WTI price to gas prices to consumer sentiment to poll implications. Claude built it in a flash. When I introduced a devil&#8217;s advocate challenge about prediction markets, Claude marshalled a defense rooted in the academic literature on market bias and then conceded, accurately, the three cases where markets genuinely dominate.</p><p>That isn&#8217;t a tool. That&#8217;s a colleague.</p><p>I want to describe what we built, because the model itself is interesting. But I want you to hold in mind what the process looked like, because this matters more.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>Why polls and prediction markets fail</h2><p>Before we get to what we built, let me outline the problems with conventional instruments.</p><p><strong>Polls are systematically biased, and we know it.</strong> The American Association of Public Opinion Research has documented, in three separate post-election studies, that Republican and independent voters are less likely to respond to surveys than Democrats, and the ones who do respond are less likely to support Trump than the ones who refuse. Across 2016, 2020, and 2024, national polls underestimated Trump&#8217;s support by an average of 2.3 percentage points. This isn&#8217;t noise. It&#8217;s a structural error that pollsters have tried, and largely failed, to correct.</p><p><strong>Prediction markets aggregate beliefs, not truth.</strong> I want to be fair to markets, because they have genuinely beaten polls in recent cycles. But markets have documented biases of their own. The favorite-longshot bias means unlikely outcomes are systematically overpriced. Concentration dynamics mean a single trader with thirty million dollars can move the price several points, which happened on Polymarket in October 2024. And markets are reactive. They price information after it arrives. They cannot tell you in advance which three variables would reshape a race.</p><p><strong>Historical analogs fail in novel regimes.</strong> This is the deepest problem. The 2026 midterm is the first modern cycle to feature an active war, a closed strategic waterway, wartime stagflation, and a second-term president whose polling error pattern is both famous and partially attenuated (because he&#8217;s not on the ballot). There is no 1994 here. There is no 2006. The structural features of this moment are new enough that reaching for precedent is not rigor. It is laziness dressed as rigor.</p><p>The question I brought to Claude was whether we could build something that corrected for each of these failures explicitly. Not a better poll aggregator. Not a more sophisticated market-watcher. A meta-model that consumed the conventional instruments, acknowledged their biases, and added the one thing they cannot add: liv<em>e causal structure.</em></p><div><hr></div><h2>Can AI build a meta-model?</h2><p>That was the sharper version of the question, and the one that made my chai sit half-drunk.</p><p>A meta-model must synthesize signals that disagree with each other. It has to weight them by their actual predictive power, not their cultural prominence. It has to know which inputs are collinear and therefore shouldn&#8217;t be double-counted. It has to apply documented corrections to known biases. And it has to do all of this while remaining falsifiable, meaning every assumption has to be visible and revisable.</p><p>Claude and I spent the first fifteen minutes of the session on exactly this question. Not on the 2026 numbers. On the architecture. What are the true independent signal classes? What is each one&#8217;s historical predictive power? Where do they overlap, and how do we strip the overlap without losing information? What corrections do we apply to each, and why?</p><p>The answer we arrived at, which I named POLARIS-26, has four architectural layers. Let me walk through each, because the layering is where the sophistication lives.</p><h3>Layer 1: Five orthogonalized signal pillars</h3><p>The core insight is that generic ballot, presidential approval, and the economy are not independent. Approval absorbs the economy. Generic ballot absorbs both. Naive weighting double-counts roughly thirty-five percent of the underlying information. POLARIS treats generic ballot as the primary signal, weighted at forty percent, and weights approval and economy on their <em>residuals only</em>, meaning the portion of variation not already explained by generic ballot movement. This is the same logic Nate Silver uses. It matters more than most aggregators admit.</p><p>The final weights, after orthogonalization, are: generic ballot 40%, approval residual 15%, economic residual 10%, prediction markets 20%, geopolitical shock index 10%, fundraising momentum 5%. Every weight had to defend itself against a historical backtest.</p><h3>Layer 2: The Polling Integrity Adjustment</h3><p>This is the filter that corrects for the documented Trump-voter undercount. The critical calibration question is how much correction to apply. Too little and you miss 2024-style systematic error. Too much and you over-correct for a midterm where Trump isn&#8217;t on the ballot.</p><p>The literature gives us an anchor. The average Trump-era presidential polling bias is 2.3 points. The 2022 midterm bias, with Trump off the ballot, was closer to 1.0 point. POLARIS splits the difference with a 1.5-point Republican shift, carried as a distribution with an uncertainty band of 1.0 to 2.0. That uncertainty band is itself consequential. If the true midterm bias turns out to be 2.5 points, POLARIS&#8217;s Senate point estimate shifts from D 49 to D 48 and the probability of Democratic Senate control drops from 22% to 14%. The model expresses this uncertainty honestly rather than pretending it doesn&#8217;t exist.</p><h3>Layer 3: The Swing Seat Gate</h3><p>Most forecasters waste compute on races that aren&#8217;t races. A district with an 85% incumbent retention probability is not informative. POLARIS excludes every seat with baseline win probability above 80% or below 20% and runs full simulation only on the genuinely competitive seats: roughly forty-five House districts and ten Senate races. This is how you focus compute where it actually matters. The Senate gate currently includes Maine, North Carolina, the Ohio special, Iowa, Alaska, Georgia, Michigan, Minnesota, New Hampshire, and Kentucky. Every other seat is locked. Every gated seat gets the full probability treatment.</p><h3>Layer 4: NovaWatch, the live candidate feed</h3><p>The final architectural element is a live monitoring layer for candidate-specific events: indictments, sexual misconduct revelations, retirements, primary upsets, viral gaffes, fundraising shocks, district-specific economic shocks. Each category has a calibrated race-level impact range and a decay function. A Roy Moore&#8211;scale scandal moves a race by 8 to 18 points with a 45-day half-life. A retirement announcement moves it by 3 to 8 points, permanently. These events feed directly into race-level probabilities, bypassing the national environment index, because candidate events rarely move the national wave but routinely flip individual races.</p><p>Together, those four layers are the static architecture. But the most interesting thing we built that morning was dynamic.</p><div><hr></div><h2>The Hormuz Cascade: why causal beats analog</h2><p>The Iran situation is the swing factor in this cycle, and it is the hardest thing to model because it has no precedent.</p><p>When I asked Claude how we should handle it, the first instinct was to reach for analogs. Iraq 2006. The Gulf of Tonkin. The 1973 oil embargo. Each has some structural feature in common with 2026, but none shares enough to be useful. The 1973 embargo wasn&#8217;t wartime. Iraq 2006 was a four-year-old war, already metabolized by voters. Hormuz has been closed for sixty days in a month when gas prices are setting records. We are in new territory.</p><p>So we did something I find genuinely interesting. Rather than force the present into the shape of the past, we built a <em>live causal model</em> that could estimate its own coefficients in real time.</p><p>The Hormuz Cascade is a six-link chain:</p><p><strong>Iran events &#8594; oil prices &#8594; gasoline prices &#8594; consumer sentiment &#8594; presidential approval &#8594; generic ballot &#8594; seat outcomes.</strong></p><p>Each link has a coefficient. The trick is that we populate those coefficients not from historical data but from the sixty days of observations we now have since the war began. We have seen major escalations move Brent by $3 to $8 intraday. We have seen ceasefire signals move it down by $5 to $12. We have seen gas prices pass through oil at a rate of $0.028 to $0.035 per gallon per dollar of Brent, which is fifteen to twenty percent higher than the historical baseline, probably because Gulf infrastructure damage has constrained refining capacity. We have seen Michigan sentiment drop six points for every fifty cents of sustained gas price increase, which is fifty percent steeper than the historical pattern.</p><p>These are not assumed numbers. These are observed numbers, fit to sixty days of post-war data and updated weekly as new observations arrive. The causal chain tightens as the data accumulates. Bayesian updating in real time.</p><p>And because we built it causally rather than analogically, we can do something genuinely useful: we can specify tripwires.</p><p>A tripwire is a pre-specified threshold that forces an automatic model re-run because it represents a potential change in the underlying structure, not a marginal move. POLARIS has three:</p><ol><li><p>Brent above $110 sustained for ten trading days</p></li><li><p>Cumulative Hormuz disruption exceeding 90 days</p></li><li><p>A single-event US casualty count above 30</p></li></ol><p>Each tripwire has a pre-calculated impact vector. We don&#8217;t wait to see what analysts say. The model updates automatically when the trigger fires. This is the discipline causal modeling enforces that analog-based forecasting cannot.</p><div><hr></div><h2>The stress tests that almost broke the model</h2><p>Building the architecture took about twenty minutes. The next ten were the ones that made me respect the collaboration.</p><p>Claude proposed six stress tests, unprompted. Three of them broke the initial model.</p><p>The first was a collinearity check. Generic ballot, approval, and the economy were all weighted at 25% or higher, which double-counted their shared variance. The fix was orthogonalization, which I described above.</p><p>The second was a backtest against 2010, 2018, and 2022. The model performed well on 2010 and 2018, where the generic ballot was running above five points. It performed poorly on 2022, where the generic ballot was close to zero. This surfaced an important truth: the generic ballot is most predictive when it&#8217;s clearly above or below the noise band, and less reliable inside it. Today&#8217;s D+5.6 reading is comfortably outside the noise band, which gives us some confidence. But we flagged this as a condition to monitor.</p><p>The third was a sensitivity analysis on the Polling Integrity Adjustment. At a PIA of 0.5, the model predicted a Democratic sweep. At 2.5, it predicted a Republican hold of both chambers. The model is more sensitive to that single parameter than to any other. The fix was to carry the PIA as a distribution rather than a point estimate, and to be transparent about how much the prediction depends on it.</p><p>By the end of the stress-test round, the model looked different from the first draft. That is what testing is supposed to do.</p><div><hr></div><h2>The prediction that came out the other side</h2><p>After all of this, with the environment orthogonalized and the corrections applied and the causal cascade populated, POLARIS-26 produced its first run. Here is what it says as of April 20, 2026:</p><p>The House flips to the Democrats. Point estimate: D 225, R 210. Eighty percent confidence interval: D 217 to 233. Democratic control probability: 73%.</p><p>The Senate holds for the Republicans, narrowly. Point estimate: R 51, D 49. Democratic control probability: 22%. Probability of a 50-50 tie: 19%.</p><p>The most likely joint outcome is divided government, at 51% probability. Democratic sweep at 22%. Republican status quo at 23%. All other scenarios at 4%.</p><p>For context, Polymarket is pricing a Democratic sweep at 51.5% today. POLARIS is two points lower than the market on the House and eight points lower on the Senate. That divergence is the model&#8217;s contribution. It comes almost entirely from the PIA correction, which markets structurally cannot apply because they don&#8217;t decompose polling error by cycle type.</p><p>If the Senate result lands inside POLARIS&#8217;s D 46 to 52 range on November 3, the model will have earned its keep. If not, markets were right and I owe a public revision.</p><div><hr></div><h2>The dashboard, and the 45-minute miracle</h2><p>Here is the part I&#8217;m still struggling to process.</p><p><strong>7:00 AM</strong> &#8212; I ask the question. We sketch the architecture on the fly.</p><p><strong>7:15 AM</strong> &#8212; Signal pillars defined, weights derived, orthogonalization applied.</p><p><strong>7:20 AM</strong> &#8212; Polling Integrity Adjustment calibrated against 2016-2024 literature.</p><p><strong>7:25 AM</strong> &#8212; Hormuz Cascade structured as a six-link causal chain with live coefficients.</p><p><strong>7:30 AM</strong> &#8212; Stress tests run. Three revisions made.</p><p><strong>7:35 AM</strong> &#8212; First prediction produced with point estimates, confidence intervals, and joint probability table.</p><p><strong>7:40 AM</strong> &#8212; Committed the methodology to a reusable skill, so future runs follow the exact same protocol.</p><p><strong>7:45 AM</strong> &#8212; Interactive React dashboard built. Sliders for every signal. Live Brent cascade. Scenario buttons. Tripwire indicators. Fonts chosen, colors set, responsive layout tested.</p><p><strong>8:30 AM </strong>&#8212; This article was written and posted!</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!gEIj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!gEIj!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 424w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 848w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 1272w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!gEIj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png" width="1080" height="1420" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1420,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:140961,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/194792418?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!gEIj!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 424w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 848w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 1272w, https://substackcdn.com/image/fetch/$s_!gEIj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61f2878a-55d1-48b5-8413-77ce8ef72d98_1080x1420.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>A working midterm forecasting model. Architecturally sophisticated, methodologically honest, empirically testable, with its own interactive dashboard. One hour from idea to publication.</p><p>I want to be careful not to overclaim. This is a model, not an oracle. It could be wrong. I&#8217;ve been explicit about where it could be wrong and how I&#8217;d know. But the fact that we went from question to falsifiable production artifact in an hour is, I think, genuinely a new thing in the world.</p><p>I&#8217;ll tell you what I find most moving about the experience. Not the speed, although the speed is astonishing. Not the output, although the output is good. What I find moving is the quality of the collaboration. Claude didn&#8217;t just execute my instructions. It proposed architecture I wouldn&#8217;t have thought of. It caught errors I would have missed. It pushed back on weak reasoning. It held the thread across ninety minutes of technical conversation without losing the narrative. It committed the methodology to a reusable skill so I can run this model again next month with a single command.</p><p>That is not tool use. That is colleague-scale partnership, available to anyone with a laptop and a question.</p><div><hr></div><h2>The interactive artifact</h2><p>By tomorrow (and another cuppa tea), I will build an interactive dashboard so you can play with POLARIS-26 live. You will be able to:</p><ul><li><p>Move any signal slider and watch the seat counts update in real time.</p></li><li><p>Toggle the Polling Integrity Adjustment on or off to see how much it moves the Senate probability. (Spoiler: a lot.)</p></li><li><p>Push the Brent oil slider past $110 and watch the Hormuz tripwire fire red.</p></li><li><p>Click any of the four preset scenarios (Baseline, Democratic Wave, Republican Recovery, Hormuz Break) and see the model jump to that world.</p></li><li><p>Watch the joint probability matrix recalculate.</p></li></ul><p>The code is open. The math is visible. Every assumption is revisable.</p><p>This is what I want more of in the world. Models that are interactive rather than opaque. Methodologies that are visible rather than hidden. Collaborations with AI that feel like colleagues rather than tools.</p><div><hr></div><h2>Reflections on what just happened</h2><p>I started this Monday morning with a question and a cooling cup of tea. By 7:45 AM I had an architecturally sophisticated, empirically testable, interactively explorable political forecasting model, a committed reusable skill for future runs, a static dashboard, and a clear set of falsifiable predictions with a November resolution date.</p><p>This is not normal productivity. That is a new kind of intellectual leverage. A curious mind plus a capable AI collaborator, asking good questions together, iterating in real time, stress-testing each other&#8217;s thinking, and producing something that neither could have produced alone at anything close to that speed.</p><p>My seventh decade is going to be more interesting than I expected.</p><p>Gotta love Claude.</p><div><hr></div><p><em>Mohanbir Sawhney is the McCormick Foundation Professor of Technology at the Kellogg School of Management. He writes about AI, strategy, and the interior life of modern work</em></p>]]></content:encoded></item><item><title><![CDATA[Are You Prepared to Manage Your Digital Employees?]]></title><description><![CDATA[An HR blueprint for your Hybrid Human-AI Workforce]]></description><link>https://www.hiddenweave.com/p/are-you-prepared-to-manage-your-digital</link><guid isPermaLink="false">https://www.hiddenweave.com/p/are-you-prepared-to-manage-your-digital</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Mon, 23 Feb 2026 10:46:24 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!KQ6W!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!KQ6W!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!KQ6W!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!KQ6W!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png" width="1456" height="977" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:977,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:8223280,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/188854419?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!KQ6W!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!KQ6W!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F17968bd6-34d3-4b9b-bbdc-be116ec1b919_2528x1696.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>Your company has an HR department. It recruits talent, onboards new hires, manages performance, handles compensation, ensures compliance, and plans for the future of the workforce. No serious company would operate without it.</p><p>Now ask yourself:<em> who is doing any of this for your AI agents?</em></p><p>Large enterprises are deploying agentic AI across every function, from marketing to finance to customer service. These agents read documents, draft responses, make recommendations, flag risks, and interact with customers. Some operate around the clock. Some make decisions that affect revenue. Some touch sensitive data.</p><p>But these digital workers have no onboarding process, no performance reviews, no governance structure, no clear ownership, and no one tracking whether they&#8217;re actually delivering value. They are, in effect, <strong>feral employees</strong>: hired enthusiastically, deployed tactically, and managed by no one.</p><p>Your human workforce has an entire management discipline behind it. Your digital workforce needs one too. I call it <strong>Digital Labor Orchestration (DLO)</strong>, and this article is its blueprint.</p><h2>The Parallel That Changes Everything</h2><p>The simplest way to understand DLO is through a mirror. Hold up your existing HR function and ask: what would this look like for AI agents?</p><p><strong>Hiring employees</strong> becomes <strong>sourcing and selecting digital agents</strong>. Do you build your own? Buy from Salesforce or Microsoft? Outsource to a service provider? The decision depends on the same factors as human talent acquisition: strategic importance, IP sensitivity, and cost.</p><p><strong>Onboarding and training</strong> becomes <strong>configuring and fine-tuning agents</strong>. Just as a new hire needs context, tools, and expectations, an AI agent needs prompt engineering, access permissions, guardrails, and escalation rules. Skip this step and you get the AI equivalent of an employee who was never told what their job actually is.</p><p><strong>Performance management</strong> becomes <strong>monitoring agent effectiveness</strong>. What are the accuracy rates? Completion rates? How often does the agent escalate versus resolve? Are its outputs drifting over time? You wouldn&#8217;t tolerate a human employee who was making expensive mistakes and messing up their jobs. Why tolerate agents that hallucinate or go off the reservation?</p><p><strong>Compensation and costing</strong> becomes <strong>understanding the true economics of digital labor</strong>. The sticker price of an API call is not the cost of digital labor, just as a salary is not the total cost of employing a human. You need to account for compute, licensing, monitoring, retraining, exception handling, and human oversight. I call this the <strong>Total Cost of Labor Ownership (TCLO)</strong>: the sum of human labor cost, digital labor cost, and orchestration cost.</p><p><strong>Compliance and ethics</strong> becomes <strong>guardrails, auditability, and explainability</strong>. When an AI agent makes a loan decision, who&#8217;s liable? When it generates customer-facing content, who reviews it? When it fails silently, who notices?</p><p>This isn&#8217;t a metaphor. It&#8217;s an operating model.</p><h2>From Jobs to Flows</h2><p>Here&#8217;s where DLO parts company with traditional workforce thinking. In the human world, work is organized in terms of <em>roles and titles</em>: &#8220;marketing manager,&#8221; &#8220;loan officer,&#8221; &#8220;customer service representative.&#8221; In the DLO world, the organizing principle is <em>workflows</em>, broken down into tasks and micro-tasks.</p><p>The first step in any DLO initiative is what I call a <strong>Workforce X-Ray</strong>: a deep diagnostic that decomposes roles into the actual tasks people perform, then evaluates each task for its potential to be handled by a digital agent. A European bank that performed this exercise on its loan origination function discovered over 120 micro-tasks, from intake and document review to fraud checks and credit decisioning. Some tasks were obvious candidates for automation. Others required human judgment. Most fell somewhere in between.</p><p>This decomposition is where the real insight lives. A &#8220;loan officer&#8221; is not one job. It&#8217;s dozens of micro-tasks stitched together by habit and job description. Some of those tasks are ripe for AI. Others are deeply human. The art is in the balancing.</p><h2>The Four A&#8217;s: A Volume Dial for Autonomy</h2><p>Once you&#8217;ve identified which tasks can be handled by digital agents, the next question is: <strong>how much autonomy should an agent have?</strong> This is not a binary choice between &#8220;automated&#8221; and &#8220;not automated.&#8221; It&#8217;s a dial with four settings.</p><ul><li><p><strong>Assist.</strong> The agent enhances human productivity through insights, summarization, or suggestions. A marketing copilot recommends headlines. A research agent summarizes competitive intelligence. The human decides. The agent informs.</p></li><li><p><strong>Approve.</strong> The agent performs the work, but a human must approve before anything executes. A contract analysis tool identifies risk clauses, but legal counsel signs off. The agent proposes. The human ratifies.</p></li><li><p><strong>Audit.</strong> The agent operates autonomously, but its decisions are reviewed after the fact. A dynamic pricing agent adjusts rates in real time. A human audits a sample weekly. The agent acts. The human verifies.</p></li><li><p><strong>Autopilot.</strong> The agent owns the workflow end to end without human intervention. An inventory replenishment system orders stock based on real-time demand signals. The agent decides. The human has moved on to higher-order work.</p></li></ul><p>The instinct of most executives is to jump straight to Autopilot. Resist it. The right autonomy level depends on the risk profile of the workflow, the maturity of the agent, and the trust your organization has built through experience. A billing dispute agent at a logistics company might start at Audit (handling tier-1 cases independently, with 10% weekly human review) and graduate to Autopilot in low-risk regions only after three months of demonstrated performance. High-risk markets might stay at Audit permanently.</p><p>The Four A&#8217;s give you a common language for a conversation that every enterprise is having in fragmented, inconsistent ways.</p><h2>The Agentization Map: Where to Start</h2><p>Not every workflow deserves digital labor, and not every promising workflow deserves the same level of investment. To prioritize, plot your workflows on a 2&#215;2 matrix (See the diagram below):</p><p>&#8226; <strong>Y-axis:</strong> Agentization Potential (how well-suited is this workflow for AI agents?)</p><p>&#8226; <strong>X-axis:</strong> Strategic Importance or Volume</p><p>This produces four zones:</p><ul><li><p><strong>Accelerate</strong> (high value, high potential): Fast-track these for digital labor redesign. This is where your biggest ROI lives.</p></li><li><p><strong>Activate</strong> (low value, high potential): Perfect test beds. Low risk, high learning. Use these to build organizational capability before tackling the high-stakes workflows.</p></li><li><p><strong>Augment</strong> (high value, low potential): Deploy copilots and human-in-the-loop systems. The work is too complex or too risky for full automation, but AI can make humans significantly more effective.</p></li><li><p><strong>Avoid</strong> (low value, low potential): Don&#8217;t waste resources here. Focus elsewhere.</p></li></ul><p>This map turns a sprawling, overwhelming question (&#8220;where do we start with AI agents?&#8221;) into a portfolio decision with clear priorities</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!6UBc!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!6UBc!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 424w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 848w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 1272w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!6UBc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png" width="1456" height="1398" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1398,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:170857,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/188854419?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!6UBc!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 424w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 848w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 1272w, https://substackcdn.com/image/fetch/$s_!6UBc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd3d832bb-cf12-49e8-a571-9a56be4ae3f8_1956x1878.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2>Building the Digital Labor Office</h2><p>If DLO is the discipline, the <strong>Digital Labor Office</strong> is the institutional home. Think of it as the organizational equivalent of HR, but for your AI workforce. It&#8217;s headed by a senior executive (call it Head of Digital Labor) and staffed with:</p><ul><li><p><strong>Workflow Architects</strong> who map processes and redesign them for human-AI collaboration</p></li><li><p><strong>Agent Designers</strong> who configure, prompt, and embed guardrails into digital workers</p></li><li><p><strong>Governance and Risk Officers</strong> who ensure compliance, auditability, and appropriate autonomy levels</p></li><li><p><strong>Human-AI Partnership Managers</strong> who handle change management and ensure human workers embrace (rather than fear) their digital colleagues</p></li><li><p><strong>Agent Product Managers</strong> who treat each digital agent as a product with a backlog, performance metrics, and lifecycle</p></li></ul><p>This team works in deep partnership with HR. Together, they co-own workforce planning (how many humans, how many agents?), role redefinition (what does a &#8220;loan officer&#8221; do when AI handles 60% of the micro-tasks?), and cultural transformation (how do you move from &#8220;AI is taking my job&#8221; to &#8220;AI is making my job more interesting&#8221;?).</p><p>I recommend a <strong>hub-and-spoke</strong> operating model: a central DLO team sets strategy, tools, and policies, while embedded DLO Champions in each business unit adapt and execute locally. This mirrors the HR Business Partner model that is common in human workforce management.</p><h2>Measuring What Matters</h2><p>A common mistake in digital labor is measuring only cost savings. Yes, a bank that deploys AI agents in loan origination can cut processing costs by 60% and improve speed-to-decision by 40%. Those numbers matter. But they&#8217;re the floor, not the ceiling.</p><p>DLO measurement should span three dimensions:</p><ul><li><p><strong>Productivity.</strong> Not just human productivity, but <em>blended</em> productivity. I propose a metric called <strong>Blended Workforce Productivity (BWP)</strong>: total output divided by the sum of human labor input (in FTEs) plus digital labor input (in ATEs, or Agent-Time Equivalents). This puts humans and agents on the same scorecard.</p></li><li><p><strong>Cost efficiency.</strong> The TCLO model described earlier, tracking not just compute and licensing but orchestration costs: governance, exception handling, supervision, and change management. </p></li><li><p><strong>Strategic impact.</strong> Speed to market, customer experience uplift, innovation enablement, operational resilience. If a bank can cut loan approval time from 4 days to 40 minutes, it won&#8217;t just save money. It will increase application volumes because of the improved customer experience.</p></li></ul><h2>The Journey Ahead</h2><p>Most enterprises today deploy AI agents sporadically, with no unified strategy for sourcing, governing, or measuring their digital workforce. Digital Labor Orchestration is a journey. To map this journey, I have created a DLO Capability Maturity Model with five stages of maturity, from Ad Hoc deployments with no formal structure to Institutionalized operations where digital labor is embedded in the company's operating model, talent strategy, and performance management systems. See the Table below for the full maturity model with dimensions, capabilities, and diagnostic indicators at each stage.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!euoZ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!euoZ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 424w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 848w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!euoZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png" width="1456" height="1102" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/ab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1102,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:6801274,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/188854419?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!euoZ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 424w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 848w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!euoZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab1fb443-8089-4183-97ed-306bf808918a_2368x1792.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Digital Labor Orchestration is not a technology initiative. It is a management discipline. The companies that built great HR functions gained a durable advantage in the human capital era. The companies that build great DLO capabilities will gain the equivalent advantage in the age of AI.</p><p>Your AI agents are already working. Now you need to figure out how to manage them as employees.</p>]]></content:encoded></item><item><title><![CDATA[Two Moves]]></title><description><![CDATA[How knowledge workers can remain irreplaceable in the age of AI]]></description><link>https://www.hiddenweave.com/p/two-moves</link><guid isPermaLink="false">https://www.hiddenweave.com/p/two-moves</guid><dc:creator><![CDATA[Mohan Sawhney]]></dc:creator><pubDate>Mon, 23 Feb 2026 10:14:53 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!iyVf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!iyVf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!iyVf!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!iyVf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png" width="1456" height="977" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/ada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:977,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:9133738,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.hiddenweave.com/i/188778546?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!iyVf!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 424w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 848w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 1272w, https://substackcdn.com/image/fetch/$s_!iyVf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fada865a0-e4a5-4b72-a453-a4ffc8ce7ed2_2528x1696.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p><em><strong>This is the full article behind the &#8220;method&#8221; discussion in my launch post.</strong></em></p><p>Professional value in the age of AI is under pressure from two directions. The unit in which knowledge work is priced. And whether its outcomes are inseparable from the person who produces them. Most knowledge workers are only paying attention to one.</p><p>AI has not merely automated professional tasks. It has forced two questions that knowledge workers prefer not to ask themselves. The first is economic: when AI can do what you charge for, charging for doing it stops making sense. You can still bill by the hour. But AI has stripped away the opacity that made input pricing viable. Hours are an input metric. Outcomes are an output metric. When a task that took a junior associate three hundred hours takes AI thirty minutes, the gap between what clients were charged and what the work costs is no longer hidden.</p><p>The second question is existential: does the value you create exist because of you, or merely through you? If a client, a patient, or an organization could attribute what they received to a process, a platform, or a credential rather than to a specific individual, you are replaceable. If the outcome and the person can be separated, the outcome can be sourced elsewhere.</p><p>The professionals who will thrive in the age of AI will have made two moves. The first: <em>decoupling output from input</em>, so that intellectual capital scales beyond the hours in a day. The second: <em>coupling outcome tightly with identity</em>, so that what they produce cannot be attributed to anyone or anything else. The first move is about production economics. The second is about personal irreplaceability. Most of the literature on AI and professional work has described the first. Almost no one has described the second. Both are now mandatory.</p><p>My father understood the first move before AI existed. He flew bombers for the Indian Air Force before he became an entrepreneur. He had a precise way of cutting through complexity. He gave me a piece of advice that stayed with me: <em>don&#8217;t measure your costs and revenues in the same units</em>. He was speaking as a small business owner trying to protect his margins in the fabrication business, where he bought steel in kilos and sold window frames by the piece. After forty years, I appreciate how much insight his advice carried.</p><h3>What AI Threatens and What It Does Not</h3><p>The vulnerability of a professional to AI is determined by three structural characteristics of how their value is created and attributed.</p><p>The first is <strong>input transparency</strong>. When your process is observable and reproducible, AI can substitute for it directly. Document review, financial modeling, diagnostic imaging, market research, first-draft content production: these are processes that can be specified, observed, and replicated.</p><p>The second is <strong>output separability</strong>. When the artifact you produce can be detached from the person who produced it without loss of value, you have a separability problem. The contract, the report, the diagnosis, the slide deck: if the client values the deliverable and not the person who made it, they do not need you specifically. They need a capable producer. Increasingly, AI is that producer.</p><p>The third is <strong>outcome attribution</strong>. When the result the client cares about is attributed to a process, a firm, a platform, or a credential rather than to a specific individual, the individual has no moat. The client was never buying you. They were buying the result.</p><p>The most exposed professionals score high on all three. The most protected score low on all three: opaque inputs built through experience that cannot be observed or reproduced by instruction, inseparable outputs whose value is irreducibly bundled with the person who produced them, and outcomes attributed personally and completely to a specific individual.</p><p>This is what the pricing literature missed, because it was concerned with fee structures, not with the more fundamental question of whether the value survives the removal of the person who created it. Outcome-based pricing changes how you bill. It does not change whether you are replaceable.</p><h3>The Lawyer Who Gets You Four Billion Dollars</h3><p>David Boies bills by the hour. But the hourly rate is a billing convention. What his clients are paying for is something else entirely. When he represented the government in its antitrust case against Microsoft or advised investors in the aftermath of the Theranos collapse, the value being exchanged operated on an entirely different logic than billable hours. He is one of a handful of lawyers who can change the outcome of a case worth billions of dollars. His preparation, his deposition strategy, his cross-examination instinct are inputs, but they are not what the client is buying. The client is buying an outcome that would not exist without Boies in the courtroom. The input is opaque. The outcome is personally his. Both protective moves made.</p><p>The threat to lawyers is not at that level. It is at every level below it.</p><p>The Am Law 100 business model runs on leverage. Senior partners develop client relationships and originate business. Junior associates execute the work: document review, due diligence, contract drafting, research memos, discovery production. The spread between what associates bill and what they cost is the profit engine. That spread is collapsing, because the work is pattern recognition across large document sets, which is precisely what large language models do better than humans at a fraction of the cost.</p><p>When that justification disappears, so does the leverage model. Partners who own client relationships, whose outcomes are inseparable from their personal judgment, will capture more value than ever. Associates who expected to climb a partnership ladder by accumulating billable hours will find the ladder shortened from below.</p><p>The lawyers moving toward outcome-based pricing are on the right track. But they are only halfway there. Fixed fees, retainers, and contingency structures are the first move. The second move is making the attribution of the outcome irreducibly personal. Not the firm&#8217;s judgment. Not the practice group&#8217;s methodology. This person, this mind, this accumulation of experience that exists nowhere else.</p><h3>The Surgeon of Last Resort</h3><p>There is a small cohort of surgeons the ultra-wealthy call when a diagnosis has been delivered and they are not ready to accept it. They are the physicians who see the case three specialists missed, who will perform the procedure no one else at their institution will attempt. No patient asks how long the consultation took. No family asks about the hourly rate. The value is irreducible and the attribution is personal.</p><p>The disruption of medicine by AI is hurting the top of the cognitive hierarchy rather than the bottom, which surprises people until you understand the mechanism. Radiology. Pathology. Diagnostic dermatology. These specialties require exceptional training and years of pattern exposure. They are also, at their core, visual pattern recognition tasks, and AI has reached or exceeded human performance in several of them for specific conditions. The radiologist reading two hundred chest scans a day is performing high-skill cognitive labor whose process is entirely observable and whose output is completely separable from the individual who produced it. The report says &#8220;radiology department.&#8221; The patient rarely knows the reader&#8217;s name.</p><p>All three vulnerability conditions are present. AI steps into the gap, not because the radiologist lacks skill, but because the structure of the work makes both protective moves unavailable. The individual and the value were already disaggregated before AI arrived. AI simply made the disaggregation economically decisive.</p><p>The physicians who are protected are those whose value is irreducibly bundled with a specific person. The concierge physician who has known a family across three generations. The surgeon whose judgment and hands have performed this procedure ten thousand times and whose name is on the outcome in a way no departmental attribution can replicate. Their configuration is the mirror image of the radiologist&#8217;s, and so is their exposure.</p><h3>The McKinsey Senior Partner</h3><p>One of my former students is a senior partner at McKinsey. He does not write slides. In fact, he has not written slides in two decades. What he does is appear in the right boardroom at the right moment, when a CEO faces the kind of decision that ends careers or transforms companies. He draws on thirty years of wisdom to offer an invaluable and uniquely personal perspective. Then he leaves. The slides come from someone else.</p><p>His input is almost entirely invisible. His output is inseparable from him. Nobody in the room is thinking about the process. They want to know what this brilliant mind sees that they cannot.</p><p>The consulting model under pressure is not his. It is the pyramid surrounding him: the analysts and associates who ingest data, build financial models, populate templates, and produce deliverables that justify the engagement fee. AI does that work faster and more consistently at a cost approaching zero.</p><p>There is a structural vulnerability in consulting that does not exist in the same way in law. The consulting outcome is frequently attributed to the firm rather than to the individual. &#8220;McKinsey recommended this&#8221; is weaker personal attribution than &#8220;Boies argued this.&#8221; The brand intermediates between person and outcome, which dilutes the individual moat. The consultants building durable practices are those who have made their personal attribution explicit, who are known for a specific way of seeing problems that cannot be found elsewhere.</p><p>The industry is barbelling. At one end, firms with powerful institutional attribution will retain their position because the brand intermediates effectively between the work and the buyer. At the other end, boutiques built around a named individual with a distinctive point of view will also thrive. The vast middle is most exposed: firms and practitioners trading on neither strong institutional brand nor strong personal attribution, whose value proposition rests on competent execution of processes AI can now replicate.</p><p>For individuals inside strong institutions, the barbell creates a false sense of security. The institutional brand may protect the firm. It will not protect you if your outcomes cannot be attributed to you personally when the institution is removed from the equation.</p><h3>The Business Guru and the Forty-Year Draft</h3><p>I want to be transparent about something directly relevant to this argument.</p><p>This article went from first conversation to complete draft faster than I could have dictated it longhand. An AI model helped me structure the argument, pressure-test the logic, refine the exposition, and craft a logical narrative arc at a speed that was unthinkable three years ago.</p><p>And yet the framework at the center of this article, the three vulnerability conditions, the two protective moves, the distinction between the first move the pricing literature describes and the second move it ignores, are uniquely mine. Rooted in thirty-five years of watching how knowledge markets work, teaching strategy to executives across six continents, building programs, making mistakes, and developing a distinctive point of view. AI accelerated the time from idea to execution. It did not generate the insight. Archimedes promised to move the world if given a lever and a place to stand. My decades of experience are the place to stand. AI is the lever that extends the reach.</p><p>When Picasso was challenged on the price he charged for a portrait completed in minutes, he replied: &#8220;It took me all my life.&#8221; That is the precise situation of any professional whose value lives in accumulated judgment rather than current effort. The output looks effortless. The input is a career. As I like to say: a ladder has rungs for a reason.</p><p>I have reached more than forty thousand executive education participants in six years through online programs, generating over a hundred million dollars in gross revenue. In the classroom, I teach at most sixty students per class, the same number I taught thirty years ago. The decoupling between my input and my output is complete. But the scaling is only protected because the outcome remains attributed to something that cannot be replicated by removing me from the equation. Executives seek this specific point of view, these specific frameworks, this particular voice. The content scales without limit. The attribution does not transfer to the platform, the course, or the institution. It stays with me. Both moves made. Neither alone is sufficient.</p><h3>What the Pricing Literature Missed</h3><p>The case studies illustrate the logic. Here are the two moves stated precisely.</p><p><strong>The first move is decoupling.</strong> Break the link between your effort and your output so that your intellectual capital scales beyond the hours in your day. This is productization. Stop selling time. Build something that delivers value while you sleep. Don&#8217;t be lulled into a false sense of security by inertia. Clients do not defect the moment AI enters the picture. But inertia is a reprieve, not a defense.</p><p><strong>The second move is coupling.</strong> This is less intuitive and more important. Make the outcome of your work inseparable from your individual identity, so that a client, patient, or organization attributes the result to you specifically and not to a process, a firm, or a credential. Coupling is not about how you produce value. It is about whether the value is recognized as yours. It is the only durable answer to the question AI is now forcing every knowledge worker to confront: does the value exist because of you, or merely through you?</p><p>The two moves work in tandem. <em><strong>Decouple the production. Concentrate the attribution. Scale where you are going. Own the credit for arriving. </strong></em></p><p>My father&#8217;s aphorism reaches further than he intended. Don&#8217;t measure your costs and revenues in the same units. It is not advice about invoicing. It is a statement about the nature of value. What professional work costs and what it is worth have never been the same thing. The gap between them was sustainable because it was opaque. AI has made the gap transparent. And a visible gap cannot be sustained.</p><h3>Where This Leads</h3><p>The professionals who will define knowledge work in the next decade share a configuration. Their inputs are opaque, built through experience that cannot be observed or reproduced by instruction. Their outputs are inseparable from their individual identity. Their outcomes are personally owned.</p><p>The senior litigator whose cross-examination instinct was built across a thousand depositions. The surgeon whose judgment has been formed by ten thousand procedures. The strategy partner whose mastery over organizational dynamics came from sitting in a hundred boardrooms at the highest stakes. The educator whose frameworks were forged across forty years of research, teaching, and synthesis that no one else has done in quite this way.</p><p>What is happening to professional work is not replacement. It is disaggregation: the separation of expertise from the process that used to house it. The hours were never the value. The credential was never the moat. The process was never what the client was paying for. AI has not changed what professional value is. It has made it impossible to pretend otherwise.</p><p>The honest question every knowledge worker needs to answer is not whether AI can do their job. It is two questions, more uncomfortable and more precise. In what units is my value denominated? And is the outcome of my work inseparably mine?</p><p>If the value is in the hours, the repricing has already begun. If the outcome could be attributed to a process, a platform, a firm, or a credential rather than to a specific irreplaceable individual, the substitution is closer than it appears.</p><p>If your value lives in outcomes that only your specific accumulated judgment can deliver, and if those outcomes are irreversibly coupled to your identity, you are not competing with AI. You are using it.</p><p>My father gave me a pricing principle for his small business in India. As he smiles down at me from heaven, I salute him for his timeless insight.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.hiddenweave.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading The Hidden Weave! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><p></p>]]></content:encoded></item></channel></rss>