<?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[On Humans and Agents]]></title><description><![CDATA[Commenting on AI with insights from history and classical philosophy]]></description><link>https://www.onhumansandagents.com</link><image><url>https://www.onhumansandagents.com/img/substack.png</url><title>On Humans and Agents</title><link>https://www.onhumansandagents.com</link></image><generator>Substack</generator><lastBuildDate>Wed, 18 Nov 2026 11:42:25 GMT</lastBuildDate><atom:link href="https://www.onhumansandagents.com/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[Michael Mangialardi]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[michaelmangialardi@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[michaelmangialardi@substack.com]]></itunes:email><itunes:name><![CDATA[Michael Mangialardi]]></itunes:name></itunes:owner><itunes:author><![CDATA[Michael Mangialardi]]></itunes:author><googleplay:owner><![CDATA[michaelmangialardi@substack.com]]></googleplay:owner><googleplay:email><![CDATA[michaelmangialardi@substack.com]]></googleplay:email><googleplay:author><![CDATA[Michael Mangialardi]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[Delegating the Mechanical: Early Modern Information Management and the Lesson for AI]]></title><description><![CDATA[AI threatens the panache of the scholarly archetype. Recapitulating that archetype my save us from unnecessary mechanization of human activity.]]></description><link>https://www.onhumansandagents.com/p/early-modern-information-management-ai</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/early-modern-information-management-ai</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Tue, 17 Nov 2026 20:36:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!zkqU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!zkqU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!zkqU!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!zkqU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/eb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1796242,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/225387253?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!zkqU!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!zkqU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb8517b7-7ff6-43e7-a6e5-27b362e50cc3_1125x750.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><h2>The Proliferation of Note-Taking Methodologies</h2><p>To enrich my commentary on AI, I&#8217;ve sought to draw from historical parallels.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>Recently, I worked through <em><a href="https://www.amazon.com/Too-Much-Know-Scholarly-Information-ebook/dp/B004G5Z780/ref=sr_1_1?crid=1SLCWSPJCQW24&amp;dib=eyJ2IjoiMSJ9.JtgqATUKAapsZ6bgFAmyhNm0RDURoxJsAVzJxH0EsXbV5qbmyfPsHl84W-5wlma3M_a5vQgsx9yG4B9UP0wYdPu6ZZIYnK1wkzerMUMYwBYi7Osr8ko7xEOFXXCdGC5nPdCh33sg690YH4quum8D7MsRpM8saSTj88EjnLtg4msSJqpH-fdBVYKu_nw-sW0lXQt2v73GVDl1nnylvOr9JjRWRuColeSM5lPV72K0Kas.1Feaa6uZjXKprO8dyGH4uCuI_ILRj8crZOG5PH8pEXs&amp;dib_tag=se&amp;keywords=too+much+to+read&amp;qid=1794939191&amp;sprefix=too+much+to+re%2Caps%2C204&amp;sr=8-1">Too Much to Know: Managing Scholarly Information before the Modern Age</a> </em>by Ann M. Blair (Yale University Press, 2010).</p><p>When the printing press came onto the European scene, scholars could gain an advantage by reading more widely.</p><p>Yet, the proliferation of texts could also cause the scholar to get lost in the sea of information. As <a href="https://catholiclibrary.org/library/view?docId=Medieval-EN/XCT.044.html&amp;chunk.id=00000003">Thomas Aquinas (-1274) encouraged Brother John:</a></p><blockquote><p>Since you have asked me how one should set about to acquire the treasure of knowledge, this is my advice to you concerning it: namely, that you should choose to enter, not straightway into the ocean, but by way of the little streams; for difficult things ought to be reached by way of easy ones.</p></blockquote><p>To navigate the journey to the ocean through little stream, early modern scholars became more disciplined in notetaking.</p><p><a href="https://takenote.library.harvard.edu/sites/g/files/omnuum10456/files/2025-11/placcius_1_0.jpg">Marginal notes</a>, <a href="https://publicdomainreview.org/collection/john-lockes-method-for-common-place-books-1685/">commonplace books</a>, <a href="https://archive.org/details/proverbschieflyt00blaniala">compilations</a>, <a href="https://quod.lib.umich.edu/d/did/">encyclopedias</a>, and other reference books flourished during this period. </p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!ZTqi!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ZTqi!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!ZTqi!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg" width="1456" height="958" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:958,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:2072943,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/225387253?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!ZTqi!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ZTqi!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a238bb3-8e81-4f23-8fcc-9a0ba56418e8_1569x1032.jpeg 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><figcaption class="image-caption">Credit: <a href="https://takenote.library.harvard.edu/73-ultimate-piece-office-furniture">Houghton Library via Harvard University</a></figcaption></figure></div><p>One of my favorites is Leibniz&#8217; note closet. Leibniz took notes on slips of paper, and then arranged them by topic name.</p><p>Initially, this might seem neat but archaic. However, the implementation details around these methodologies is quite analogous with our integration of AI.</p><h2>Identifying Mechanical Work</h2><p>Many early modern scholars merely took notes and leveraged reference books to acquire vast knowledge without losing depth in primary subjects.</p><p>There were some professional readers who did the reading and the compiling of notes for another. (<a href="https://www.onhumansandagents.com/p/oakdale">I&#8217;ve even written a short story that speculates what a factory of professional readers could look like when carried to the extreme</a>.)</p><p>However, this was not common as scholars wanted to preserve reading as the noble, enriching act.</p><p>Some, like Leibniz, nevertheless saw &#8220;mechanical&#8221; steps in the note-taking process that could be delegated.</p><p>Copying marginal notes into slips and ordering the slips in the note closet were the mechanical steps that Leibniz was willing to delegate.</p><p>But the act of reading and categorizing topics required higher-level &#8220;judgment,&#8221; and that was worth preserving.</p><div><hr></div><p>It fascinates me that this recognition of non-mechanical work and the unwillingness to delegate it away was so perfusive amongst the early modern intelligentsia.</p><p>I ask myself whether Leibniz would be amazed by the ability to leverage AI to delegate the non-mechanical&#8212;he could spawn agents to read, share, and organize research notes&#8212;or, whether he would be appalled that intelligence itself is becoming mechnical.</p><p>I&#8217;m inclined to think that there would be more to inform his judgment than research potential alone. The panache of the scholarly experience was no small part of his refusal to delegate non-mechanical work, and the intrusion of AI into that panache would be a primordial alarm. </p><h2>Personal Level: Preserving Non-Mechanical Work</h2><p>It would do us good, both at a personal and organizational level, to consider what constitutes &#8220;mechanical&#8221;, AI-worthy work and what work is non-mechanical and panache-preserving (all changes in custom since Leibniz aside).</p><p>Personally, I love the panache of the scholarly archetype, and so, I&#8217;ve embraced a very minimalistic use of AI:</p><ul><li><p>If it is normative for my professional line of work to leverage AI, and I cannot compete without following suit, then I will use AI.</p><ul><li><p>When I do software engineering work (my main source of income), I utilize AI.</p></li><li><p>When I teach (a part-time income stream), I don&#8217;t use AI in any part of the process. (It is not normative</p></li><li><p>When I work at Starbucks (another part-time income stream), I enjoy an analog day.</p></li><li><p>Since June, I&#8217;ve eliminated AI from my Substack writing process. (I used to use for checking my work, though I regret this.)</p></li><li><p>I don&#8217;t use AI for my graduate work&#8212;not evening for gathering recommended resources on a topic or grammar checking.</p></li></ul></li><li><p>If AI is needed to establish a just relation, then I may use it.</p><ul><li><p>The adoption of AI in HR workflows has resulted in taller list of required and preferred qualifications in job requirements&#8212;and AI systems tightly evaluate your application for these elevated requirement. To level the playing field, I will use AI to ascertain gaps in my experience and find resources for closing those gaps.</p></li><li><p>I suspect a similar phenomenon has emerged with legal documents. I will use AI to help sift through the bureaucratic text in, say, a refinanced car loan proposal.</p></li></ul></li><li><p>Otherwise, no AI. I also have a heavy mix of analog experiences:</p><ul><li><p>I print out most online article I&#8217;d like to read.</p></li><li><p>I tend an organic garden and take care of goats and chickens. </p></li><li><p>I teach at a classical school and work at Starbucks to get plenty of digital-free interactions.</p></li><li><p>I often involve my wife for feedback on my writing and ideas. I have other contacts that may do the same.</p></li><li><p>I read physical books and have a marginal note system. I don&#8217;t view the work of reviewing my notes and organizing them as &#8220;mechanical.&#8221;</p></li><li><p>I pause and think &#8220;is there someone local I could financially support to do this work&#8221; if I do want to delegate; or, I do the work myself.</p></li></ul></li></ul><div><hr></div><p>A healthy first step right now is for people to reflect upon and be transparent about their delegation to AI.</p><p>Early modern approaches to scholarship can help us <em>feel </em>the limits of AI usage to preserve panache, serving as an archetype.</p><p>Thankfully, I was well-read on this archetype before AI&#8212;although I regret to say that my discipline in resisting AI delegation has not always been as strong as it is now.</p><p>But it&#8217;s added a lot of joy to overcome the delegation problem.</p><p>In short, we should be careful about making the joy of human learning mechanical.</p><h2>Organizational Level: Preserving Non-Mechanical Work</h2><p>The same applies at an organizational level in corporate America, even just the practical end of finances is in view.</p><p>AI integration has not been a smooth rollout. Not just at a psychological and ethical level, <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">but with respect to maximal return on investment</a>. </p><p>I suspect that treating all corporate work as mechanical has introduced more problems than it solves in some domains&#8212;beyond just <a href="https://www.onhumansandagents.com/p/ai-governance-and-payoff-without-coordination-bottleneck">the coordination problem</a>.</p><p>For example, HR systems have used AI to optimize their workflows&#8212;but now we have a soulless process that&#8217;s stagnated the job application market.</p><p>Draining the life out of potential workers by <a href="https://www.nytimes.com/interactive/2026/10/28/style/gen-z-job-market.html?unlocked_article_code=1.MVE.wTXR.otcEfJItl_nj&amp;smid=nytcore-ios-share&amp;utm_source=substack&amp;utm_medium=email">presenting the job market as an abyss </a>is hardly good for employers or the economy.</p><p>I&#8217;m sure my readers can multiple examples.</p><h2>The Delicacy of the Matter</h2><p>I feel the tension around AI, and I&#8217;m sure many of you do too.</p><p>I&#8217;m not one to be enthused by utopian visions of universal income. A future without diverse work for the exchange of goods reminds me of the grey town in Lewis&#8217; <em>The Great Divorce</em> rather than anything sanguine.</p><p>I&#8217;m also highly concerned about <a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">worker displacement</a>, though I&#8217;ve charted <a href="https://www.onhumansandagents.com/p/software-engineering-alternative-plans">a path forward</a> despite the challenges.</p><p>At the same time, AI is part of <a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">a global race</a> toward national defense, cybersecurity, and economic advancement. It is here to stay even if it reminds me of Grendel&#8212;and, at times, I&#8217;d be content with a Beowulf to come slay it.</p><p>We may not be able to slay Grendel, but we can keep him at bay.</p><p>We can accept the economic reality that requires AI while also enriching our view of human work&#8212;and preserving panache&#8212;by not viewing humanity as inevitably mechanical.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Software Engineering is (At Best) Plateauing: What Should You Do?]]></title><description><![CDATA[Create an antifragile plan in case software engineering cannot be salvaged]]></description><link>https://www.onhumansandagents.com/p/software-engineering-alternative-plans</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/software-engineering-alternative-plans</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Tue, 17 Nov 2026 03:16:37 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!dQHq!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.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/$s_!dQHq!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dQHq!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 424w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 848w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 1272w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!dQHq!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png" width="500" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:500,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:329230,&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.onhumansandagents.com/i/225274936?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!dQHq!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 424w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 848w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.png 1272w, https://substackcdn.com/image/fetch/$s_!dQHq!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6524e1ec-d992-46b9-92f4-d488c951d09b_500x750.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><h2>Software Engineers Need a Backup Plan</h2><p><a href="https://www.lennysnewsletter.com/p/state-of-the-tech-job-market-late">Software engineering is plateauing</a>.</p><p>The job market is volatile.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>And <a href="https://www.nytimes.com/interactive/2026/10/28/style/gen-z-job-market.html?unlocked_article_code=1.MVE.wTXR.otcEfJItl_nj&amp;smid=nytcore-ios-share&amp;utm_source=substack&amp;utm_medium=email">applying for jobs has become an extremely brittle and frustrating process.</a> </p><p>All because of AI integration&#8212;which <a href="https://www.wsj.com/cio-journal/ai-is-testing-the-limits-of-corporate-governance-f69dda7d">isn&#8217;t going anywhere</a>.</p><div><hr></div><p>Don&#8217;t bank on an AI rollback.</p><p>And don&#8217;t let wishful thinking&#8212;which tells you that there&#8217;s still a need for software engineers&#8212;prevent you from taking action.</p><p><a href="https://www.onhumansandagents.com/p/anthropic-ai-expectations-gap">Anthropic thinks that software engineers can be replaced by 2027</a>; the public thinks otherwise.</p><p>There is enough risk that any software engineer should assume Anthropic is right and create a backup plan.</p><div><hr></div><p>I&#8217;ve gone through this process&#8212;and all the grief that comes with it.</p><p>I didn&#8217;t have the best backup plan in place. I got laid off in early 2025 as a shadow AI organization absorbed capital, and I&#8217;ve been trying to land on my feet ever since.</p><p>But I see with eyes wide open now and have solid advice for anyone preparing for the decline of software engineering.</p><h2>Here&#8217;s What I Would Do</h2><h3>#1 - Recognize You&#8217;ll Need Resilience</h3><p>That sounds platitudinous, but it needs to be said.</p><p>As software engineers, we are excellent at problem-solving, maximizing efficiency, and developing an ethos that cuts through bureaucracy.</p><p>And we are told that America is a meritocracy, rewarding workers based on their merits.</p><p>So, software engineers are likely to assume that the market will reward their skills and offer a panoply of opportunities&#8212;even when changing fields.</p><p>That may be true if you can connect directly with business leaders, but the HR machinery will filter you out.</p><p>They aren&#8217;t going to extract transferrable potential in your skills and experience; as of now, they will look for buzzwords and credentials, declining you if you don&#8217;t fit.</p><p>Given that, a software engineer&#8217;s backup plan will quickly run into the following:</p><ul><li><p>Systems that haven&#8217;t adapted to reality of worker displacement driven by AI</p><ul><li><p>HR isn&#8217;t going to see your cross-domain potential and accept your software engineering experience as an alternative qualification</p></li><li><p>There aren&#8217;t &#8220;micro-credentials&#8221; that can qualify you for another domain with little time investment</p></li><li><p>Job listings won&#8217;t go above their estimated salary range because you are used to a higher salary</p></li></ul></li><li><p>People and organizations that won&#8217;t reward you as openly for your problem-solving, or worse, don&#8217;t want you to solve their problems in the first place</p></li><li><p>Lower paying jobs with higher professional standards</p><ul><li><p><em>Example</em>: A job at McDonald&#8217;s will have a dress code a software engineer is not accustomed to</p></li></ul></li></ul><p>In short, you will have higher expectations for what you can do and how you should be compensated than the market will offer you.</p><p>This dissonance is incredibly frustrating when you first encounter it.</p><p>But you need resilience. You need to accept these realities as they are, not as they should be.</p><p>I&#8217;m writing extensively on this topic to promote change. I&#8217;m not complacent. But I could save you a lot of trouble to avoid solving a macro-level problem that is out of your reach.</p><p>While you can&#8217;t solve the dissonance right now, you can create an optimal path within it.</p><h3>#2 - Move to the Relevant Organization</h3><p>If you&#8217;re not in an organization (or team) that is AI-forward, then you have an expiration date.</p><p>Try to make an internal transition to an organization/team in your company that is receiving investment.</p><p>Consider taking a pay cut and switching into a role outside of software engineering if it is AI-proof. It&#8217;s easier to make a career change internally than in the wild.</p><h3>#3 - Save Money</h3><p>Don&#8217;t expect a severance to completely cover you until you land a new job.</p><p>The more margin the better. Some savings is better than none. Don&#8217;t wait, and throw any savings into an online savings account or invest in stocks.</p><p>I&#8217;d aim to save 18 months worth of your current rate of pay. Better too much than too little.</p><h3>#4 - Start a Consulting Business</h3><p>AI is causing trust to decline.</p><p>When anyone can make a resume, for example, your resume carries less weight in an interview process. The ripple effect is wide.</p><p>Given that, consultants will have value even if they do the work that others can now DIY.</p><p>A small business owner is willing to pay for the trust that comes from you (an experienced professional) building the software.</p><p>While I think the risk of worker displacement is serious enough to write this post, I do think the remaining value of trust can be overlooked.</p><p>Some people&#8217;s trust will remain calibrated to pre-AI expertise. Plenty of small business owners will trust an experienced software engineer over an up-and-coming AI firm.</p><p>Take advantage of this window&#8212;the margins are great. Start a consulting business.</p><p>I recommend <em>Book Yourself Solid</em> as a good book for learning how to market yourself.</p><h3>#5 - Consider a Flexible Part-Time Job in Hospitality and Food Service</h3><p>Jobs in Hospitality and Food Service demand the soft skills that many software engineers are weak in.</p><p>The laid back ethos of software engineering is a perk, but you&#8217;ll need to change if you want to transfer your value into a new domain.</p><p>Applying analytical skills into physical work is also very satisfying.</p><p>The bar to entry is low, hours tend to be flexible, and there are some nice benefits.</p><p>Extra income, sharpened soft skills, and free stuff may be worth the 20 extra hours. Plus, you have a fallback that will anchor you amidst the volatility, and you&#8217;ll likely learn more than you expect.</p><h3>#6 - Teach Part-Time</h3><p>Similarly, education has a formality to it that a software engineer would do well to absorb.</p><p>Software engineers will enjoy teaching, and the formality (and dealing with kids) part will build your resilience.  </p><h3>#7 - Imagine a New Lifestyle</h3><p>Trimming a budget means embracing a new lifestyle.</p><p>Seeing this is as an opportunity to master frugality, cultivate soft skills, and develop impenetrable resilience is not a detour for an entrepreneurially minded person.</p><p>Don&#8217;t do something you hate, and aim to do something you love.</p><p>You&#8217;d be surprised at how meaningful work that pays less will create a high quality of life you thought was associated with higher income.</p><p>A full pivot from software engineering into a less volatile and highly enjoyable field is worth the pay cut&#8212;but there is a baseline you&#8217;ll need to hit, and it&#8217;s going to require more work than you may be accustomed to.</p><h3>#8 - Chase One of the Emerging Software Engineering-Adjacent Roles</h3><p>I&#8217;ve intentionally kept this point last since I think many may jump to it first. I still think the previous points are worth considering even if this is the route you go.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!bUZ2!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bUZ2!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 424w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 848w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 1272w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!bUZ2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png" width="1456" height="1592" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/075a565e-4879-48af-866c-73980da381a9_2912x3184.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1592,&quot;width&quot;:1456,&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/$s_!bUZ2!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 424w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 848w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.png 1272w, https://substackcdn.com/image/fetch/$s_!bUZ2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F075a565e-4879-48af-866c-73980da381a9_2912x3184.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><figcaption class="image-caption">Credit: <strong><a href="https://x.com/lennysan/status/2120547244953031043">Lenny Rachitsky</a></strong></figcaption></figure></div><p>AI companies are growing more than the rest of corporate America, but they&#8217;re concentrated in San Fransico.</p><p>Unless you want to move there, then I&#8217;d look into forward deployed engineering, hardware engineer, or security engineering.</p><p>Product management has also been surface in these kinds of conversations, but I&#8217;d imagine it&#8217;s a harder sell.</p><p>In any case, you&#8217;ll need more than &#8220;I can learn this&#8221; or &#8220;here are some toy projects&#8221; to make the transition. And, again, when trust is low legitimate efforts in reading or small projects won&#8217;t carry the same weight as they used to.</p><p>Pursue formal opportunities rather than personal projects, even if it&#8217;s so easy to do personal projects right now with AI.</p><div><hr></div><p>Software engineering is plateauing. It&#8217;s a weird time. But you do have options if you play your cards right.</p><p>Perhaps there&#8217;s a way to salvage software engineering. I&#8217;ll continue to think about that myself, but it&#8217;d be wise to have a backup.</p><p></p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[The Counter Regulation: Europe’s Challenge to the American AI Build-Out]]></title><description><![CDATA[America&#8217;s lack of regulation may, counterintuitively, be an advantage should regulation be desired in the global market]]></description><link>https://www.onhumansandagents.com/p/the-counter-regulation-europes-challenge</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/the-counter-regulation-europes-challenge</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Fri, 13 Nov 2026 15:38:44 GMT</pubDate><enclosure url="https://substack-video.s3.amazonaws.com/video_upload/post/224762484/6f6e83ca-2792-4be0-a44b-c2de86037cdf/transcoded-1794584314.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p>
      <p>
          <a href="https://www.onhumansandagents.com/p/the-counter-regulation-europes-challenge">
              Read more
          </a>
      </p>
   ]]></content:encoded></item><item><title><![CDATA[Europe’s Challenge to the American Plan]]></title><description><![CDATA[America&#8217;s lack of regulation may, counterintuitively, be an advantage should regulation be desired in the global market]]></description><link>https://www.onhumansandagents.com/p/europes-challenge-to-the-american-ai-plan</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/europes-challenge-to-the-american-ai-plan</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Thu, 12 Nov 2026 22:13:37 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!66N4!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!66N4!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!66N4!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!66N4!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!66N4!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!66N4!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!66N4!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1221018,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/224650133?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!66N4!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!66N4!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!66N4!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!66N4!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F45cf0a6d-9f1b-44ce-9f2c-3bca91b00671_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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 <a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">American plan for AI</a> removed the burden of discordant state regulation, placing instead a federal review on the export of AI technologies. </p><p>Without intrusive regulations, and with the capaciousness for training frontier models at an elite scale&#8212;thanks to galvanized US hyperscalers providing direct access to Nvidia GPUs&#8212;America&#8217;s Anthropic and OpenAI are beating out the capabilities of European challengers.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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><h2>European Challengers</h2><p><a href="https://www.anygen.io/showcase/aleph-alpha-europe-answer-to/index.html">Aleph Alpha</a> (via Germany) and <a href="https://mistral.ai/">Minstral</a> (via France) are European answers to America&#8217;s closed frontier models. </p><p>These models are attractive to European companies as they provide off-the-shelf compliance with the <a href="https://artificialintelligenceact.eu/high-level-summary/">EU AI Act</a>. They could also attract global consumers who value  off-the-shelf peace of mind that European regulation intrinsically offers for both security and social costs.</p><p>Aleph Alpha is tailored for highly regulated industries&#8212;such as defense, banking, and government&#8212;as it allows on-premise sovereignty.</p><p>Mistral aims to compete more directly with Anthropic and Open AI as a powerful general-AI model provider, yet compliant with European regulation and more affordable.</p><h2>Reviewing the Global Competition</h2><p><a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">China&#8217;s plan</a> is to provide the cheapest AI models for the global market that could also become the most powerful through recursive improvement enabled by high rates of adoption. </p><p>The American plan has been to remove the burden of regulation, marshall its build-tout to take advantage its distinct access to GPUs, and offer the best capabilities to approved consumers in the global market for a premium price.</p><p>China is leading in embodied AI but lagging behind the capabilities of American AI technologies.</p><p>The UK appears to be immediately focused on integrating American AI technologies across various domains through centralized institutions evenly spread out across the nation&#8212;ironically <a href="https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early">resembling America&#8217;s approach during the British Empire&#8217;s free trade dominance during the Industrial Revolution</a>.</p><p>Europe&#8217;s plan is less monolithic than the competition, and unlikely to offer the most capacious models; yet, Europe may continue to produce models that stay competitive in performance and include regulation for an attractive offering.</p><p>The uncertainty is whether the increased <em>competence</em> of AI technologies always translates to increased opportunities.</p><p>If <a href="https://www.onhumansandagents.com/p/ai-governance-and-payoff-without-coordination-bottleneck">AI technologies require effective </a><em><a href="https://www.onhumansandagents.com/p/ai-governance-and-payoff-without-coordination-bottleneck">coordination</a></em>, then the most performant models may not necessarily provide enough of an edge to justify the premium cost.</p><h2>Europe&#8217;s Challenge</h2><p>Europe&#8217;s opportunity is in its AI technologies with out-of-the-box regulation.</p><p>This could also turn out to be a weakness with respect to the global market&#8212;though overall economic success is not limited to winning the global market for AI models.</p><p>The British Empire dominated free trade during the Industrial Revolution, yet nations like American and France were able to effectively counter through a protectionist approach.</p><p>But AI&#8217;s impact may not be completely analogous with industrialization. The risks seem higher.</p><p>If history repeats itself, then nations will focus on primarily protecting themselves when risks are high.</p><p>Therefore, it becomes apparent that AI build-out requires intrusive regulation to thwart risks to defense, cybersecurity, and public interest, then nations will want to self-regulate rather than getting off-the-shelf regulation.</p><p>Europe&#8217;s pre-regulated technologies may be inconvenient and/or duplicative with regulation at home.</p><h2>America&#8217;s Opportunity</h2><p>Given that, America has an opportunity to provide the global market with the most powerful AI models and a gold standard apparatus for implementing regulation.</p><p>Just as US hyperscalers provided the apparatus for gold standard software implementation, AI companies grown in the US may provide gold standard platforms and tools so other nations can apply custom regulation to their technologies.</p><p>This means that the global market could couple premier US models and gold standard regulation tools. Or, they could pick and choose.</p><p>Now, there&#8217;s nothing preventing from the UK and Europe from providing this self-regulation apparatus.</p><p>However, the UK does not seem too focused on the global market, and Europe is forced to provide narrow regulation in its technologies.</p><p>America&#8217;s lack of regulation may, counterintuitively, be an advantage should regulation be desired in the global market.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[How to Govern AI While Still Capturing the Payoff]]></title><description><![CDATA[Integrate AI to multiply competence without birthing a coordination bottleneck]]></description><link>https://www.onhumansandagents.com/p/ai-governance-and-payoff-without-coordination-bottleneck</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/ai-governance-and-payoff-without-coordination-bottleneck</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Wed, 11 Nov 2026 17:28:36 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!s2Tx!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.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/$s_!s2Tx!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!s2Tx!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 424w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 848w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 1272w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!s2Tx!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png" width="986" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:986,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:260584,&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.onhumansandagents.com/i/224476054?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!s2Tx!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 424w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 848w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 1272w, https://substackcdn.com/image/fetch/$s_!s2Tx!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F745bce59-4542-4f96-a97c-6771fc2afc04_986x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>The Coordination Problem</h2><p>The integration of AI increases <em>competence </em>but can also spike <em>coordination </em>efforts.</p><p>Consider the following examples:</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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><ul><li><p><em>Increased work </em>- AI, for instance, helps job applicants prepare for interviews&#8212;but it also helps recruiters raise the intensity of desired qualifications and evaluations. All things considered, the job application process has created a <a href="https://www.nytimes.com/interactive/2026/10/28/style/gen-z-job-market.html?unlocked_article_code=1.MVE.wTXR.otcEfJItl_nj&amp;smid=nytcore-ios-share&amp;utm_source=substack&amp;utm_medium=email">bottleneck for the job market</a>. The same phenomenon is manifesting with cybersecurity.</p></li><li><p><em>Increased bureaucracy </em>- <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">AI integration thrives when companies consolidate on platform and tooling</a>. Yet, <strong>the uniform integration of AI is most effective when the task is distributed directly to the overseer of an organizational slice</strong>, <a href="https://www.onhumansandagents.com/p/how-to-successfully-integrate-ai">allowing them to restructure their teams and systems around the capabilities of AI</a>&#8212;rather than weaving AI into assumed structures shaped by pre-AI capabilities.</p></li><li><p><em>Increased risk of over-engineering </em>- The <a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">AI race between America and China</a> has a  ripple effect on other nations. <a href="https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early">The UK is taking a protectionist approach</a>. The EU has mobilized to enact an <a href="https://artificialintelligenceact.eu/high-level-summary/">AI Act.</a> Global companies are forming an <a href="https://www.charteroftrust.com/">alliance on cybersecurity</a>. In 2025, American companies rushed to adopt AI early to gain an edge. <a href="https://www.onhumansandagents.com/p/researchers-explorers-and-stitchers">Exploring software engineers</a> have obsessed over harness engineering techniques around <a href="https://www.onhumansandagents.com/p/atomic-ai-engineering-methodology">prompting, skills, agent swarms/organizations</a>, and the like. Yet, <a href="https://www.onhumansandagents.com/p/simple-ai-integration-claude-dreams">frontier models continue to rapidly evolve and absorb the work that we expected from human organizations</a>&#8212;and the latest research only predicts that <a href="https://metr.org/blog/2026-11-05-anthropic-ai-rd-acceleration-assessment">models will accelerate exponentially</a>.</p><ul><li><p><strong>Any attempt to marshall human organizations to optimize around AI capabilities quickly becomes outdated and ineffective as models evolve.</strong></p></li><li><p>The need to move fast also doesn&#8217;t create the conditions for imagining how human organizations may be assuming structures optimized for pre-AI capabilities. </p></li><li><p>Organizations need to swiftly mobilize as AI capabilities evolve, and that requires discerning how to effectively tear down (and not just set up).</p></li></ul></li></ul><p>Given this, AI governance is most effective when lightweight and minimalistic, creating seamless reporting and restructuring mechanisms.</p><p>Otherwise, the potential for <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">payoff will be obscured</a>, and the will to integrate evaporates.</p><h2>The Coordination Solution</h2><p>Here are rules of thumb to maintain lightweight and minimalistic AI governance to thwart increased work, bureaucracy, and risk of over-engineering:</p><ul><li><p><strong>Consolidate on minimalistic platform and tooling</strong></p><ul><li><p><em>Example</em>: Claude Code licenses for all is good strategy even if there are greater opportunities for cost-optimization with the nitty-gritty of harness engineering.</p></li></ul></li><li><p><strong>Pin agents to people rather than creating agent organizations mediated by a central team</strong></p><ul><li><p>It is much easier for accurate reports to be gathered from personal testimony of agent use rather than complex calculations of the costs and value of central agents-as-a-service teams.</p></li></ul></li><li><p><strong>Make leaders of organizational slices directly responsible for reporting AI integration, eliminating unnecessary middlemen</strong></p><ul><li><p>Strategy needs to come from the top, but it doesn&#8217;t require a bloated middle</p></li></ul></li><li><p><strong>Don&#8217;t make every domain utilize AI</strong></p><ul><li><p>HR&#8217;s heavy use of AI has only created stagnation and breeded cynicism, for example</p></li><li><p>Only domains that heavily rely on <em>competence</em> (and not coordination)&#8212;such as software engineers&#8212;should integrate AI</p></li><li><p>Blindly throwing AI at teams is the new <a href="https://web.eecs.umich.edu/~weimerw/2018-481/readings/mythical-man-month.pdf">man-month fallacy</a></p></li></ul></li><li><p><strong>Invest in researchers/strategists around AI integration</strong></p><ul><li><p>Have them mine the latest AI news and research as a full-time activity</p></li><li><p>Give them a direct seat at the executive roundtable</p></li></ul></li><li><p> <strong>Re-deploy resources; Don&#8217;t displace them</strong></p><ul><li><p>Workers won&#8217;t easily adapt if they sense a risk of displacement&#8212;it&#8217;s foolish to try to fool people</p></li><li><p>Coordination needs to set up and tear down swiftly as AI capabilities evolve. </p><ul><li><p>A culture where workers can be re-deployed to new strategies will invigorate your best talent.</p></li><li><p>It also allows companies to coordinate around the current and anticipated states of AI capabilities.</p></li></ul></li><li><p>Pre-AI, companies could afford to blindly layoff without significant pushback.</p></li><li><p>Post-AI, trust is paramount, and social costs are costly. The job application market is also incredibly ineffective in the present, and scaling up cannot be effective with a revolving door.</p></li></ul></li></ul><p>In short, organizations should govern AI because blindly throwing AI won&#8217;t reach a payoff.</p><p>But governance that squares itself with a post-AI reality (instead of assuming pre-AI strategies and structures) can reach the prized payoff.</p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[How to Successfully Integrate AI in Software Organizations ]]></title><description><![CDATA[Strategic organizational structure is paramount to AI integration (and AI transforms the ideal structure)]]></description><link>https://www.onhumansandagents.com/p/how-to-successfully-integrate-ai</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/how-to-successfully-integrate-ai</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Mon, 09 Nov 2026 20:42:10 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!5Lb6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!5Lb6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!5Lb6!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!5Lb6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:875051,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/224184815?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!5Lb6!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!5Lb6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8cc825d-984d-43e7-951c-abc1a60b24bc_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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><h2>Coordination is King</h2><p>Architectural design in software engineering organizations primarily reflects optimizations for <em>coordination</em>.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>A microservices architecture, where business automation is distributed across a panoply of systems, is primarily valuable for its organizational convenience.</p><p>Just as Adam Smith encouraged specialization to scale an economy, microservices allow specialized teams to form around a slice of business automation.</p><p>In theory, this focus will produce better results than if attention is split across multiple domains that compose business automation.</p><p>Software engineering leaders often primarily see the impact on technical <em>competence</em>&#8212;but they are not the ones at the helm of the company.</p><p>Those in charge at the highest level care most about what optimizes for coordination rather than technical competence.</p><p>Microservices allow for distributed teams that can coordinate effectively, increasing output.</p><div><hr></div><p>High-performing technologies companies&#8212;think Stripe, Meta, Google&#8212;become the organizational models for the rest of the field.</p><p>Even if smaller organizations don&#8217;t have the opportunity to scale resources through specialized, distributed teams, they still adopt microservices.</p><div><hr></div><p><strong>With AI, team structures are no longer the paramount factor for coordination.</strong></p><p>Let me explain by applying this insight to software engineering organizations.</p><div><hr></div><h2>Reimagining Software Team Topologies for AI Integration</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!ZyIJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg" width="720" height="405" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:405,&quot;width&quot;:720,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;QRC (Team topologies, 200525) v1.0&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="QRC (Team topologies, 200525) v1.0" title="QRC (Team topologies, 200525) v1.0" srcset="https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ZyIJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dafeeab-ec0a-4edf-a344-e5f221c7037a_720x405.jpeg 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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><figcaption class="image-caption">Credit: <a href="https://hennyportman.wordpress.com/2020/05/25/review-team-topologies/">Henry Portman</a></figcaption></figure></div><h3>Summary of the Team Topologies Schema</h3><p>I found <em><a href="https://teamtopologies.com/">Team Topologies</a></em> (summarized above) to be the gold standard for organization software engineering organizations.</p><p>&#8220;Stream&#8221; teams make contributions to a standalone <em>slice</em>. A slice could be the entire software application when smaller in scope, or a microservice for larger organizations/products.</p><p>&#8220;Enabling&#8221; teams are used like calvary, being marshalled to help perform a transformation across stream teams&#8212;such as when performing rebranding across disparate applications/experiences.</p><p>&#8220;Platform&#8221; teams may maintain central systems and/or expose provide accessible tools that speed up the output of stream teams.</p><p>An example of a &#8220;complicated-subsystem&#8221; team would be a central, legacy billing system that processes invoicing requests from a plethora of products.</p><p>If team structures are no longer the paramount factor for coordination, then how would even the purest software engineering organization be reorganized <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">in light of pervading difficulties with AI integration</a>?</p><div><hr></div><h3>Reimagining Platform Teams</h3><p><strong>The dawn of agents removes the need for platform teams</strong>.</p><p>Recall, platform teams centralized best practices for delivery <em>via systems/tools</em>, but those best practices can now be uniformly applied to all stream teams <em>via a markdown file </em>(skill). </p><p>Platform teams, then, are primarily in place to integrate best practices into agents at the stream level.</p><p>For example, a platform team responsible for CI/CD would remove any machinery (central systems and tools) they&#8217;ve introduced to optimize deployments.</p><p>Instead, they act as an enabling team in the short-term to ensure every stream team gets a skill plus a dedicated resource to delivery seamlessly with best practices. Afterward, the team could become a new stream team or mixed into existing ones.</p><p>A supervisor can absorb observing delivery metrics. </p><p>Reporting value is much simplified. Value is derived by comparing a stream team&#8217;s cost (agents + dedicated resource) against output. </p><div><hr></div><h3>Reimagining Complicated-Subsystem Teams</h3><p>Similarly, <strong>there is no reason for complicated-subsystems.</strong></p><p>If the complexity is due to the technical details, deploy enabling teams as needed to transform.</p><p>If the complexity is due to coordination issues with a central system, then a similar path as platform teams may be taken.</p><p>After transformation, the team is a stream team.</p><div><hr></div><h3>Reimagining Stream Teams</h3><p>Stream teams are the most effective because their work is direct and their impact easily observable.</p><p><strong>Each stream should have one agent per team member, and one dedicated resource for delivery.</strong></p><div><hr></div><p>Since an agent can birth its own organization in a sort &#8220;<a href="https://www.onhumansandagents.com/p/ai-colonization-sub-frontiers">sub-frontier</a>&#8221; connected to an &#8220;author&#8221; (team member) to handle complex tasks, the pinning of a single agent to a single engineer is essential to have traceable costs.</p><p>Agent organizations should never &#8220;float&#8221; as an entity outside of a stream team.</p><div><hr></div><p><strong>The software engineering organization should consolidate on an agent and agent harness.</strong></p><p><strong>Moreover, there should be no harness engineering/optimizing in the beginning.</strong></p><div class="callout-block" data-callout="true"><p>AI integration requires traceable costs within a lean structure. Seamless integration is worth any premiums accrued when resisting optimization.</p></div><p>Even after the initial integration, the organization should lean on the side of simplicity as <a href="https://www.onhumansandagents.com/p/simple-ai-integration-claude-dreams">models continue to absorb the capabilities of platforms and tools</a>.</p><div><hr></div><h3>Reimagining Enabling Teams</h3><p><strong>Enabling teams increase in value with AI integration.</strong></p><p>As the AI ecosystem rapidly evolves, there should be plenty of opportunities for enabling teams to perform transformations across the stream teams, especially since platform and complicated-subsystem tasks become one-time transformations&#8212;as outlined above.</p><p>Enabling teams could also be dual-purpose to perform short-term technical and product experiments to iterate further after a first-pass AI integration phase. </p><p>I&#8217;m undecided whether enabling teams should be a deployable pool or divided up and assigned to stream teams. In either case, it makes sense of both stream and enabling teams to have a distinct supervisor as they provide value in different ways.</p><div><hr></div><h3>Further Transformation</h3><p>After performing the transformation suggested above, only stream and enabling teams remain, yet no engineers are lost as they are merely redistributed into stream or enabling teams.</p><p>It does raise the question, however, as to whether both stream and enabling teams can further reduced to just one agent and worker per team (or, at least, a smaller team than pre-AI integration).</p><p>The blunt answer: <em>probably</em>.</p><p>But that&#8217;s an ethical question that the organization (and its transformation) is subservient to.</p><div><hr></div><h3>Oversight</h3><p>There could still be some room for centralization, but I presume it would be a thin pattern of automated feedback and maintenance of skills&#8212;<a href="https://lethain.com/adopting-verification-pattern/">as Will Larson has recently documented.</a></p><p>My assumption is that this does not require a dedicated platform team.</p><div><hr></div><h2>Guiding Principles</h2><ul><li><p>Optimize for efficient delivery and reporting</p></li><li><p>Consolidate on a single platform for agents and agent harnesses</p></li><li><p>Keep the AI architecture minimalist&#8212;avoid initial cost optimizations that obscure reporting</p></li><li><p>Remove centralization to increase direct contributions enabled by micro transformations (stream and enabling teams)</p></li><li><p>With simplified structures, use plain-English reports to curate data instead of any &#8220;machinery&#8221;</p></li></ul><p></p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Organizations Don’t Need to Extend Models (For Models Extend Themselves)]]></title><description><![CDATA[Claude continues to improve, this time through dreaming; Keeping AI integration simple helps make AI integration a reality for organizations]]></description><link>https://www.onhumansandagents.com/p/simple-ai-integration-claude-dreams</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/simple-ai-integration-claude-dreams</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Sat, 07 Nov 2026 12:41:42 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!EDJH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.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/$s_!EDJH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!EDJH!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 424w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 848w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 1272w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!EDJH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png" width="1000" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1000,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1155702,&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.onhumansandagents.com/i/223863197?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!EDJH!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 424w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 848w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 1272w, https://substackcdn.com/image/fetch/$s_!EDJH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F96a3654f-ff83-4f32-9666-65c68accccc6_1000x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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 href="https://www.anthropic.com/research/components-of-agent-memory">Claude dreams</a>, according to Anthropic&#8217;s latest engineering article.</p><p>And this isn&#8217;t surprising when understood analogously.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>LLMs mirror the brain, and the brain stores images as past and anticipates how to respond to future stimuli.</p><p>Drawing inspiration from nature, Claude can log its past actions and infer how to act in the future in light of them; this reflective enrichment works best in between sessions&#8212;or, for humans, during sleep.</p><p>Opus 5 was the first model to close the learning loop&#8212;the &#8220;dreaming&#8221;&#8212;to predict when information in memory will be relevant again.</p><p>It&#8217;s interesting that model improvements have been incremental enough to allow for model consumers to impose their own improvements. Prompt engineering, agent graphs, and context injection are all examples.</p><p>But eventually the models can do those things themselves, and in a more optimal way.</p><p>Failing to remove the duplication when the new model is set can cause worse performance.</p><p>Claude&#8217;s improvements with memory is just another instance of this pattern.</p><p>Anthropic saw agents perform best when they were given minimal instructions around memory optimization.</p><div><hr></div><p>So long as there are sophisticated processes of the brain still to be emulated, I&#8217;d expect models to progressively improve, and the horizon for completing tasks to lengthen.</p><p>Both the trajectory of model architecture and the <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">feedback from AI integration</a> suggest that companies should invest in consolidating on frontier models and harnesses and integrating them across the organization; they should largely resist the temptation to invest in engineering resources to extend models and harnesses&#8212;frontier models and harnesses are extending themselves.</p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Minimal AGI is Here: AI Replaces Skilled Tasks But Need Not Replace Skilled Workers]]></title><description><![CDATA[By avoiding the integration of agent swarms, pinning an agent to each skilled worker, forming coordination teams, and ensuring worker security, America can reach the AI payoff.]]></description><link>https://www.onhumansandagents.com/p/minimal-agi-is-here-ai-replaces-skilled-tasks</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/minimal-agi-is-here-ai-replaces-skilled-tasks</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Fri, 06 Nov 2026 20:20:19 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!RQ8x!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!RQ8x!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!RQ8x!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!RQ8x!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:670340,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/223808093?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!RQ8x!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!RQ8x!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5686e00-5be7-47a3-b34c-b896f0db7d2f_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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><h2>The Bottleneck in AI Integration</h2><p>AI is <em>competent</em> enough to replace many tasks of skilled professionals.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>But it spikes the <em>coordination</em> effort at the initial integration phase.</p><p>Organizational structures in corporate America are notoriously ineffective at coordination relative to lean startups.</p><p>This is why <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">AI integration has been inefficient</a> and also why <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">Apollo is skeptical about AI&#8217;s potential</a>.</p><p>The lack of a payoff in the AI economy is not due to a lack of <em>competence</em>; it is due to a lack of <em>coordination</em>.</p><h2>Anthropic&#8217;s Understanding of the Bottleneck</h2><p><a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">Anthropic believes that the lack of </a><em><a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">coordination</a></em><a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response"> is in no small part due to workers resisting integration</a>.</p><p>(Workers don&#8217;t want to contribute to the driving down of their wages&#8212;at best&#8212;or their displacement&#8212;at worst.)</p><p>Anthropic&#8217;s proposed solution: provide financial incentives for people to integrate and ensure work remains for them.</p><p>But if company shares are that financial incentive, then that&#8217;s hardly attractive at non-frontier companies with stagnate valuation.</p><p>Also two years of mass layoffs has created an atmosphere where the promise of &#8220;you&#8217;ll still have work&#8221; falls on deaf ears.</p><p>Anthropic doesn&#8217;t have a solution to these problems they gesture at.</p><h2>A Macro Solution: Economic Absorption and Redistribution</h2><p>Both OpenAI and Anthropic continue to release domain-specific tools&#8212;<a href="https://support.claude.com/en/articles/18185620-use-personal-finance-in-claude">Anthropic released personal finance capabilities today</a>&#8212;that add more competence for corporate American to integrate, spiking the coordination problem even more.</p><p>This also means that both OpenAI and Anthropic continue to absorb economic opportunities from startups in the frontier AI orbit&#8212;especially when trust is low and AI proliferates many unknown startups. </p><p>This leaves three options:</p><ul><li><p>OpenAI and Anthropic continue to absorb economic opportunities, and they reinvest their gains into corporate America to help solve the integration problem. </p><ul><li><p><a href="https://www.anthropic.com/news/public-benefit-fund">Anthropic&#8217;s creation of a public fund</a> gestures at this potential.</p></li><li><p>The risk is that the reinvestment is left to <a href="https://www.onhumansandagents.com/p/anthropics-global-charity">the whim of the voluntary giving </a>of these companies, causing capital to land far away from local needs.</p></li></ul></li><li><p>OpenAI and Anthropic are prevented from absorbing economic opportunities beyond mere AI models.</p></li><li><p>OpenAI and Anthropic distribute their economic opportunities evenly across America&#8217;s regions, effectively extracting t<a href="https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early">he UK model</a> into America&#8217;s private sector.</p></li></ul><h2>A Micro Solution: Accelerate Human Coordinators</h2><p>Macro solutions are complicated and politically-entangled, but there is a clear solution within the immediate control of corporate America: accelerate human coordinators.</p><p>While AI can replace tasks of skilled professionals, they are not yet effective at solving the coordination problem at scale, <a href="https://institute.deepmind.com/essays/the-economics-of-intelligence/">according to a recent paper from Google DeepMind</a>.</p><p>As they note, we are in a &#8220;minimal AGI&#8221; phase, even if <a href="https://www.onhumansandagents.com/p/anthropic-ai-expectations-gap">American workers are lagging to take notice of this</a>.</p><p>In this phase, workers that can accelerate AI integration through effective coordination and trust-building will be increasingly valuable.</p><p>Software engineers, mathematicians, and other workers must at risk to be displaced would be very effective in these roles. It&#8217;d be both just and effective to grant a fast-track into these roles.</p><p>These coordination roles could also be the place to move other displaced workers.</p><div><hr></div><p>The paper also suggests that displacement may not be as severe as we&#8217;d predict. When we distinguish minimal AGI from further stages of AGI, consolidating on technology, pinning a worker to an agent, and moving workloads from the worker to an agent&#8212;while staying blind to the temptation of integrating and coordinating agent swarms&#8212;can maximally preserve existing structures. <a href="https://www.onhumansandagents.com/p/ai-colonization-sub-frontiers">It can also make organizational activity, and ultimately cost and value, much more easy to track</a>&#8212;a paramount problem according to the latest <a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">Wharton AI Adoption report</a>.</p><div><hr></div><p>These challenges are the challenges in implementing this micro-solution:</p><ul><li><p>It requires shifting AI integration single C-level individuals to a &#8220;calvary&#8221; of coordinators, and bureaucratic habits might stubbornly resist.</p></li><li><p> It requires fluidity between roles that may clash into HR-maintained gates.</p></li><li><p>Most importantly, it could drive down wages for non-coordinating workers:</p><ul><li><p>The Google DeepMind paper writes, &#8220;workers whose skills are matched by AI will see their wages decline toward the cost of compute, while specialists and human coordinators can demand even higher wages than before.&#8221;</p></li><li><p>In this event, the problem has only been softened from fears of complete displacement to reduced wages.</p></li><li><p>However, if corporate America has already committed to AI adoption, then perhaps the coordination efforts will save enough cost to offset the temptation to reduce wages.</p></li></ul></li></ul><h2>Between a Macro and Micro Solution</h2><p>At some point, America cannot have its cake and eat it to.</p><p>We cannot point fingers to critique macro solutions yet not do our apart at the micro level.</p><p>But mico solutions would no doubt benefit from the macro solutions.</p><p>No matter which layer hits a breakthrough first, the value of human workers needs to be paramount to build out for the common good.</p><p><strong>The goal is clear: integrate</strong><em><strong> </strong></em><strong>AI to replace</strong> <strong>skilled tasks while maintaining skilled workers.</strong></p><p>The old adage is still true for a new problem: <em>If there is a will, there is a way.</em></p><p></p><p></p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[AI and the Cold War: AI’s Mark on America’s Journey in Free Trade]]></title><description><![CDATA[The AI build-out mirrors both the Industrial Revolution and the Cold War. But history need not repeat itself; technological breakthroughs change thing]]></description><link>https://www.onhumansandagents.com/p/ai-and-the-cold-war</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/ai-and-the-cold-war</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Thu, 05 Nov 2026 20:02:20 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Srq3!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.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/$s_!Srq3!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Srq3!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 424w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 848w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 1272w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Srq3!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png" width="1456" height="1092" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1092,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:11178537,&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.onhumansandagents.com/i/223665944?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Srq3!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 424w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 848w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 1272w, https://substackcdn.com/image/fetch/$s_!Srq3!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb5e8e105-3f21-475e-b014-fc9f52948ea1_3248x2436.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>Nations protect themselves.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>Gunpowder changed for good how nations protect themselves.</p><p>Ancient weapons were relatively cheap and easy to produce.</p><p>After firearms, as Adam Smith observed, national defense an expensive, perennial line item.<a class="footnote-anchor" data-component-name="FootnoteAnchorToDOM" id="footnote-anchor-1" href="#footnote-1" target="_self">1</a></p><p>When nations feel threatened, they raise their walls. As walls rise, international trade falls.</p><p>The Industrialization Revolution blossomed the imagination for a peaceful world united by free trade; but it also provided a militaristic edge to the winner, and someone must win unless all nations agree to complete equality (and follow through by will or by force).</p><p>Threatened by the emergent winner, nations, like the United States, raised their walls and scaled industrialization at home to thwart the British Empire. Other European nations followed suit.</p><p>The world wars unleashed the stockpiles of firearms, causing unprecedented havoc.</p><p>From the ashes, peace emerged in Europe and a post-war consensus helped restrain bellicose behavior, and free trade helped cement this consensus.</p><p>But the consensus wasn&#8217;t global. The Cold War between democracy and communism loomed over the world.</p><p>Trade was freer, but limited as discernment and defenses were in high demand.</p><p>The end of the Cold War saw a resurgence of the post-war hope for wider peace and freer trade.</p><div><hr></div><p>AI is the new gunpowder, and it is the new costly line item to maintain nation defense.</p><p>Nations feel threatened during this build-out, and walls are being raised higher.</p><p>It also promises a revitalization of trade for those aligned in a neo-Cold War.</p><p><a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">America and China are racing to recapitulate the dominance of the British Empire</a> during the Industrial Revolution. They will attempt to lower walls to exchange their technologies. Surely, some will accept the trade-off.</p><p>The <a href="https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early">UK is staying out of the global race</a>, building defenses, and integrating American AI technology until it scales up its own (as a counter example).</p><p>Both approaches may work, and neither formally causes peace (or war).</p><div><hr></div><p>Industrialization came with promises of global peace and utopian dreams, and it was not so. Yet the world did also improve in some respects.</p><p>AI came with promises of global peace and utopian dreams; this will not be so.</p><p>The extent of free trade will depend on the global theater.</p><p>World peace is never the fruit of technology <em>per se</em>. (Nor is war.)</p><p>But technology may change some things for good, though the extraction of its fruits requires harnessing.</p><p><a href="https://lithub.com/artificial-intelligence-is-a-horror-story/">AI may not become a horror story</a>, and AI may be less than <a href="https://darioamodei.com/essay/machines-of-loving-grace">machines of loving grace</a>.</p><p>Then again, gunpowder, and now AI, changes things.</p><div class="footnote" data-component-name="FootnoteToDOM"><a id="footnote-1" href="#footnote-anchor-1" class="footnote-number" contenteditable="false" target="_self">1</a><div class="footnote-content"><p>Smith, <em><a href="https://www.gutenberg.org/files/3300/3300-h/3300-h.htm">Wealth of Nations</a></em>, Bk. 5, Ch. 1, Pt. 1.</p></div></div>]]></content:encoded></item><item><title><![CDATA[AI Empire vs. Protectionism: Comparing the American and British Plan for AI]]></title><description><![CDATA[Listen now (20 mins) | America is primarily focused on presenting AI technologies to a global market; Britain is primarily focused on integrating foreign AI technology evenly across its regions]]></description><link>https://www.onhumansandagents.com/p/ai-empire-vs-protectionism-comparing</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/ai-empire-vs-protectionism-comparing</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Wed, 04 Nov 2026 19:14:46 GMT</pubDate><enclosure url="https://api.substack.com/feed/podcast/223514953/3277b2043456e1556296b4126f218f13.mp3" length="0" type="audio/mpeg"/><content:encoded><![CDATA[<p>I am publishing this episode for free to give my audience a taste of what&#8217;s to come. Please consider subscribing to support my work and get access to future podcast episodes.</p>]]></content:encoded></item><item><title><![CDATA[On Sub-Frontiers: Transformational Integration of AI Through Colonization]]></title><description><![CDATA[Agent sprawl must be harnessed by pinning agents to people and forming the organizational structure around that model]]></description><link>https://www.onhumansandagents.com/p/ai-colonization-sub-frontiers</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/ai-colonization-sub-frontiers</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Wed, 04 Nov 2026 17:45:30 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!mp_F!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!mp_F!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!mp_F!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 424w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 848w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 1272w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!mp_F!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png" width="1260" height="709" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:709,&quot;width&quot;:1260,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1588760,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/223415032?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!mp_F!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 424w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 848w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 1272w, https://substackcdn.com/image/fetch/$s_!mp_F!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7ad5844d-f1f2-4970-9620-6ee8a15591f1_1260x709.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>We&#8217;ve come to call emerging AI models &#8220;frontier AI.&#8221;</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>Yet, the execution of these models within a particular context is like a frontier of its own.</p><p>In his essay &#8220;<a href="https://coolcalvary.com/wp-content/uploads/2018/10/on-fairy-stories1.pdf">On Fairy Stories</a>,&#8221; JRR Tolkien described fantasy stories as &#8220;sub-creations,&#8221; stories that participate in reality&#8212;even when they operate within a distinct, fantastical world.</p><p>I&#8217;d like to call the inner-workings of a model in a particular context a &#8220;<strong>sub-frontier</strong>,&#8221; a world marked by the activity of frontier agents participating in a goal mediated by a human author.</p><p>You can think of an agent harness, an AI-enabled API layer, etc. as the &#8220;sub-frontier.&#8221;</p><div><hr></div><p>Like sub-creations, we interact with sub-frontiers through words and rely on our imagination to apprehend the activity and how it relates to the immediate chapter/task, as well as the end story/goal.</p><p>Movies help animate sub-creations, making them tangible to us.</p><p>Graphs help animate sub-frontiers, making them more tangible and predictable.</p><p>Yet, an avid reader knows that movies cannot do justice to one&#8217;s imagination.</p><p>Likewise, the best way to understand a sub-frontier is to pin an individual to it who can use words to describe what they imagine is happening.</p><p><strong>The best way for an organization to imagine the costs and impacts of AI integration&#8212;<a href="https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026">a common problem</a>&#8212;is to surface the summaries of individuals who imaginatively author the sub-frontiers.</strong></p><div><hr></div><p>This makes the process incredibly simple: empower the people closest to a sub-frontier to act decisively and trust their summaries as <em>the </em>metrics that inform reporting by higher-ups.</p><p>The pre-AI machinery of structured reporting mechanisms lose value in a post-AI landscape.</p><div><hr></div><p>Concretely, this looks like having a single individual responsible for every agent, and giving every agent a clear goal.</p><p>Each slice of an organization would then have a team of people, if needing more than one individual, that corresponds to a team of agents (and their sub-frontiers).</p><p>This collection of sub-frontiers (i.e., team of people/agents) forms a <strong>colony</strong> representing the activity of an organizational slice.</p><p>The costs and value of a sub-frontier map to a person&#8217;s account.</p><p>The costs and value of the colony map to the collective accounts of the team members.</p><p>The costs and value of the company map to the costs and values of the colonies.</p><div><hr></div><p>The anti-pattern should also be clear:</p><ul><li><p>Agents cannot be pinned to multiple people.</p></li><li><p>Colonies cannot be pinned to multiple teams.</p></li><li><p>Costs and value are not determined through reporting machinery, but rather through the plain summaries of people responsible for an agent.</p></li><li><p>There cannot be multiple agent technologies/platforms at play.</p></li><li><p>Software systems should not be centralized if they create a bottleneck&#8212;this should be re-evaluated according to the new organizational structure.</p></li></ul><div><hr></div><p>The goal in this transformation is reached if organizations can grasp costs and value, with a direct path to who is responsible for what.</p><p>If the restructuring doesn&#8217;t attain that goal, then <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">AI cannot reach its payoff</a>. </p><p>This far in, it is worth a shot.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Ordering Loves: The Tension Behind Anthropic’s Positioning to Redistribute Capital]]></title><description><![CDATA[Anthropic is positioned well to redistribute capital. Will their global mission conflict with their local consumers?]]></description><link>https://www.onhumansandagents.com/p/anthropics-global-charity</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/anthropics-global-charity</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Tue, 03 Nov 2026 19:01:12 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!8vjj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.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/$s_!8vjj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!8vjj!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!8vjj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:780016,&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.onhumansandagents.com/i/223335301?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!8vjj!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!8vjj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd006953e-aee4-46b5-b55e-4c07317e7f04_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>Committed to the Common Good</h2><p>Anthropic has invested $200 million into <a href="https://www.anthropic.com/news/economic-futures-research-fund-agenda">proactive research on frontier AI&#8217;s impact on the economy</a>.</p><p>Their co-founders have each pledged to dedicate 80% of their personal equity in Anthropic to philanthropic causes.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>Most recently, Anthropic has committed a portion of its shares to the <a href="https://www.anthropic.com/news/public-benefit-fund">Anthropic Public Benefit Fund</a>: 5% of Anthropic&#8217;s shares have been committed, with an additional 2% added at the board&#8217;s discretion for every 50% increase in the company&#8217;s value above a set baseline.</p><p>In their <a href="https://www.sec.gov/Archives/edgar/data/2130357/000162828026067168/anthropicpbc-sx1.htm?utm_source=substack&amp;utm_medium=email#ica694c859b8e458dbf6a73f56dde9d42_2965">IPO filing</a>, a summary deck mentions that they also have active collaborations with the WHO&#8217;s Regional Office for Africa, Gates Foundation, International Rescue Committee.</p><p>They&#8217;ve made clear to both investors and the public that they are committed to the public good.</p><h2>Defining to the Common Good</h2><p>The filing lists four ways in which Anthropic can improve the quality of human life:</p><ul><li><p>Accelerate progress in biology and medicine</p></li><li><p>Stimulate economic growth</p></li><li><p>Raise the quality of government services and build technological state capacity</p></li><li><p>Diffuse the benefits of AI <em>throughout the world</em></p></li></ul><p>If done indirectly through the exporting of Claude to the global market paired with direct contributions into the American economy, this could align with the <a href="https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory">American plan</a>.</p><p>The Industrial Revolution may be tellin, as <a href="https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early">the British Empire&#8217;s free trade approach</a> resonated with the aspirations of ideologues to transform the world.</p><p>If done directly through a distribution of resources and capital, the global mission could conflict with the emerging need to catalyze AI integration through financial incentives&#8212;<a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">as Anthropic identified</a>.</p><p>Anthropic has positioned itself to proactively lead through the ripple effects of an AI economy, making material contributions. </p><p>That&#8217;s half the battle.</p><p>The other half will be how they choose to order their loves as they redistribute their capital.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Solving Ineffective AI Integration: Insights from Wharton’s 2026 AI Adoption Report]]></title><description><![CDATA[Companies need clear guidelines on supervising cost and security, standardized tooling and frameworks, and empowered agents and workers (and less reorganization)]]></description><link>https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/solving-ineffective-ai-integration-wharton-2026</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Tue, 03 Nov 2026 14:42:15 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!IXhM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.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/$s_!IXhM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!IXhM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!IXhM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:852279,&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.onhumansandagents.com/i/223267237?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!IXhM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!IXhM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F44cb9692-8c0f-4721-9801-e1f34790ad7f_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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 href="https://knowledge.wharton.upenn.edu/special-report/2026-ai-adoption-report/">Knowledge at Wharton | 2026 AI Adoption Report</a></p><h2>Key Takeaways</h2><ul><li><p>AI integration continues to plague organizations.</p></li><li><p>Despite a high degree of internal training, organizations have not made clear:</p><ul><li><p>The role of AI and how to measure its success</p></li><li><p>How to empower agents and their supervisors to act decisively, eliminating  second-guessing around costs, security, and approved platform &amp; tools</p></li></ul></li><li><p>Reorganization seems constant, potentially adding to <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">subtle resistance</a></p></li></ul><h2>Action Items</h2><ul><li><p>Companies must empower by:</p><ul><li><p>Standardizing platforms &amp; tooling&#8212;while options proliferate, frontier models are powerful when driven by confident and strategic workers</p></li><li><p>Establishing reusable frameworks for evaluating and governing costs</p></li><li><p>Providing clear guidelines and examples that remove doubts about AI autonomy, ownership &amp; accountability, and data security&#8212;as well as the most efficient means to adjudicate any lingering doubts</p></li></ul></li><li><p>AI works best bottom-up when complemented with centralized patterns&#8212;resist reorganizing since centralized patterns moved the needle the most and reorganizations are costly</p></li></ul><h2>Incomplete Integration</h2><p>&#8220;<strong>In order to derive maximum value from AI, companies must use it to its full potential.</strong> This means transforming processes at a fundamental level.&#8221;</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>&#8220;Adoption is no longer the question. Employees largely report that AI is embedded in core operations and some workflows have been redesigned around it. <strong>While positive returns are near-universal, though, substantial ones are not.</strong>&#8221;</p><p>&#8220;Most departments have changed how they work, but <strong>few have finished fully implementing</strong>.&#8221;</p><h2>Unclear Vision</h2><p>&#8220;Asked what AI's primary job is over the next 12 to 24 months, 52% of managers+ name improving productivity or reducing costs. <strong>Only 22% look to AI to grow the business</strong>, whether through new revenue, new products, or better customer experience.&#8221;</p><h2>Bottlenecks</h2><h3>Weak Ownership and Empowerment</h3><p>&#8220;Accountability for agent errors is unsettled and oversight is not the gap. <strong>No single owner reaches 30% accountability</strong>: a governance or risk team leads (28%), the deploying employee&#8217;s manager (26%), the employee (25%), the vendor (9%).&#8221;</p><p>&#8220;Most departments let AI do the work, but <strong>few let it act autonomously</strong>.&#8221;</p><p>&#8220;<strong>The barriers [to granting AI autonomy] managers+ cite</strong> are trust and risk (<strong>data security 43%</strong>, regulation 24%, distrust of the AI&#8217;s reasoning 23%); a shortage of skills ranks last at 8%.&#8221;</p><h3>Illusive Costs</h3><p>&#8220;<strong>The most-cited barrier to stronger returns is the cost</strong> of tools, compute, and licenses.&#8221;</p><p>&#8220;40% of managers+ are very confident their organization can track total AI spend once costs bundled into other software and cloud services are counted. The other <strong>60% are attempting to prove a return against a number they are less sure of.</strong>&#8221;</p><h3>Disruption Competing With Standardization</h3><p>&#8220;<strong>52% of managers+ say their department has changed its management structure</strong> because of AI, 37% report staff reductions, and 33% report slowed or frozen hiring.&#8221;</p><p>&#8220;50% of organizations that <strong>standardize on a single approved platform</strong> report strongly positive ROI.&#8221;</p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[The UK Adopts a UK-First Plan for AI]]></title><description><![CDATA[The UK is emphasizing localism via centralized institutions distributed across the nation in its AI build-out]]></description><link>https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Tue, 03 Nov 2026 02:05:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!sKMj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!sKMj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!sKMj!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!sKMj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1327091,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/223246486?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!sKMj!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!sKMj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>Summarized this article in conjunction with others in this podcast episode: </p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;98ca65e6-efb2-4412-9634-6f0a309a4b27&quot;,&quot;caption&quot;:&quot;I am publishing this episode for free to give my audience a taste of what&#8217;s to come. Please consider subscribing to support my work and get access to future podcast episodes.&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;lg&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;AI Empire vs. Protectionism: Comparing the American and British Plan for AI&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:222541085,&quot;name&quot;:&quot;Michael Mangialardi&quot;,&quot;bio&quot;:&quot;Covering the borderland between agents and humans with insights from classical philosophy.&quot;,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8ed7f0da-08b7-4b73-af37-632fe0470cbb_806x806.jpeg&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2026-11-04T19:14:46.867Z&quot;,&quot;cover_image&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/737201d7-79b4-4c32-8d7c-9ee0217b7848_6016x3649.jpeg&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://www.onhumansandagents.com/p/ai-empire-vs-protectionism-comparing&quot;,&quot;section_name&quot;:null,&quot;video_upload_id&quot;:null,&quot;id&quot;:223514953,&quot;type&quot;:&quot;podcast&quot;,&quot;reaction_count&quot;:0,&quot;comment_count&quot;:0,&quot;publication_id&quot;:7414701,&quot;publication_name&quot;:&quot;On Humans and Agents&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!HhGi!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8e19b227-db96-4f97-998c-7aa10610cdad_751x751.png&quot;,&quot;belowTheFold&quot;:false,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><h2>Comparing the AI Plans of America and the UK</h2><p><a href="https://www.gov.uk/government/news/chancellor-unveils-britains-ambitious-plan-for-ai?utm_source=substack&amp;utm_medium=email">Britain unveiled an ambitious plan for AI</a> that emphasizes:</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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><ul><li><p>Distributing the opportunities of AI across local economies</p></li><li><p>Investing in UK AI companies, including physical and industrial AI</p></li><li><p>Establishing research institutions to deliver innovative and safe adoption of AI according to British values</p></li><li><p>Scaling up AI hardware capabilities to drive reindustralization</p></li><li><p>Building resilience against cyber, biological, and other emerging AI-enabled threats</p></li></ul><p>The United Kingdom is not aiming to win the global AI market; instead, Britain is focusing on ubiquitous integration to create independence from foreign actors.</p><p><a href="https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf">America&#8217;s AI Action Plan</a> focuses on:</p><ul><li><p>Securing the largest AI ecosystem to set global standards and consumption, analogous to the space race</p></li><li><p>Spurring innovation by ensuring states remove discordant, unnecessary regulation in accordance with the <a href="https://www.whitehouse.gov/presidential-actions/2025/12/eliminating-state-law-obstruction-of-national-artificial-intelligence-policy/">national policy framework</a> </p></li><li><p>Building vast AI infrastructure and the energy to supply it</p></li><li><p>Introducing high-paying jobs, resisting worker replacement, and evaluation AI&#8217;s impact on the labor market</p></li><li><p>Exempting open models from additional regulations</p></li><li><p>Nudging AI adoption by supporting a &#8220;try-first&#8221; culture </p></li><li><p>Prioritizing investment in AI-enabled manufacturing</p></li><li><p>Bolstering cybersecurity and national defense</p></li><li><p>Advancing the science of AI</p></li></ul><p>While sharing talking points, the key differences in the American plan are:</p><ul><li><p>The primary focus on winning in the global market</p></li><li><p>The reliance on market forces rather than centralized institutions</p></li><li><p>The inability to selectively distribute the AI economy across regions</p></li></ul><p>Comparing the two, the paramount difficulty of the American plan is navigating the cross-pressures between frontier AI and AI integration.</p><h2>Cross-Pressures: Frontier AI and AI Integration</h2><p>Given the primacy of winning the global market, the removal of state regulation has accelerated frontier AI, especially Anthropic and OpenAI.</p><p>These companies have produced AI technologies leading at a global scale.</p><p>However, the models from these outlets have become so powerful that they&#8217;ve proven a key assumption in the American plan to be false: that the tools would merely augment, and not replace, workers.</p><p>Either workers are being replaced, or there is <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">subtle resistance to integrate AI</a>.</p><p>Moreover, since AI models require a concentration of resources, these companies have consolidated power.</p><p>Seeing growing valuation, these companies have been able to provide financial incentives to accelerate AI integration within their companies. <a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">Other companies lack this means to overcoming resistance to AI integration.</a></p><p>As a result, the frontier AI companies are moving fast ahead of the rest of corporate America, absorbing research and development. The consolidation problem deepens.</p><p>Besides investments and talent, these companies have also absorbed research and development.</p><p>Concentrated in the west coast, especially Silicon Valley, AI opportunities are not evenly distributed across the country.</p><p>These pressures have stifled the expectations of reaching a shared payoff in the AI economy, eroding the will to build data centers.</p><h2>The UK Advantage</h2><p>By contrast, the UK is centralizing the AI build-out, giving them the opportunity to selectively distribute.</p><p>Consolidation can only occur in AI segments designed upfront while research can remain competitive in the private sector.</p><p>This resolves the cross-pressures America is facing, but it requires lowering the bar. </p><p>Homegrown AI technologies merely have to be &#8220;good enough&#8221; to establish independence; they&#8217;ve made peace with not winning the global market.</p><h2>Roles Reversed: The Inverse of the Industrial Revolution</h2><p>In this way, they resemble the dominant protectionist policy of the young American in the 19th century.</p><p>At that time, the British Empire sought to win the global market through low tariffs and the marshalling of industrialization.</p><p>If history repeats itself, the protectionist policy will become a strong counterforce that many prominent nations will emulate.</p><p>America may have success in the global market as the British Empire did. However, consumers of American technology may defect to protectionism should America become lacking, sluggish, insecure, or too expensive.</p><div><hr></div><p>Now, it is entirely possible that America can win the global market and succeed in creating diverse opportunities and deep integration at home.</p><p>But that requires the growing protectionist spirit in Europe to not defect to Chinese models.</p><p>And it will require a stronger force than the mere removal of state regulation.</p><p>America may not be positioned to introduce centralized institutions to provide this force. </p><p>But it can find ways to redistribute the power consolidated into frontier AI and imbue the national plan with a spirit of localism.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Cheaper Models May Not Win the Global AI Race]]></title><description><![CDATA[America&#8217;s higher price tag in the global market doesn&#8217;t necessarily give China the edge]]></description><link>https://www.onhumansandagents.com/p/cheaper-models-may-not-win-the-global</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/cheaper-models-may-not-win-the-global</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Fri, 30 Oct 2026 05:23:31 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!ENCM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!ENCM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ENCM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!ENCM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1332967,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/222738769?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!ENCM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!ENCM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5528f0a6-2a21-4e83-88ad-f18e723ac8eb_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>America is offering closed and expensive AI; China is offering open and inexpensive AI.</p><p>This is good for China if the global marker is content with &#8220;good enough&#8221; capabilities if it means a lower price tag.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>This is good for America if the premium is worth the additional cost in the eyes of the global market.</p><p>AI is different than manufactured goods; the technology can progressively improve itself as more people use it.</p><p>China&#8217;s approach may accumulate enough consumers to supercharge it&#8212;at least according to &#8220;<a href="https://commanding-heights.com/#part-iii">The Commanding Heights</a>,&#8221; a spicy new take on the trajectory of the global AI economy.</p><p>But as we&#8217;ve seen with Anthropic&#8217;s race against OpenAI, c<a href="https://www.theinformation.com/newsletters/applied-ai/cheaper-models-win-openai-enterprise">heaper models won&#8217;t necessarily win the AI race</a>. America may benefit from the accumulation effect first.</p><p>OpenAI has sought to win the market by this same strategy of &#8220;cheaper yet comparable.&#8221; Yet, the market has favored Anthropic because it&#8217;s superior capabilities.</p><p>Anthropic has also slashed its costs through engineering breakthroughs, offering a more capable <em>and </em>cheaper option.</p><p>Again, AI is different than manufactured goods; there may be more incentive to opt for better quality despite the premium charge when the stakes are massive.</p><p></p><p></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[How the American Empire Plans to Win the AI Race (And What Hurdle They Must Overcome)]]></title><description><![CDATA[America plans to offer expensive-yet-powerful, closed AI technology to a global market by marshalling the private sector. They must solve the bottleneck of worker displacement.]]></description><link>https://www.onhumansandagents.com/p/how-the-american-empire-plans-to</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/how-the-american-empire-plans-to</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Thu, 29 Oct 2026 04:41:15 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/406da069-3068-4099-b642-8ed7021d79a9_3000x3000.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Thanks for checking out this episode, and I&#8217;d value your feedback.</p>
      <p>
          <a href="https://www.onhumansandagents.com/p/how-the-american-empire-plans-to">
              Read more
          </a>
      </p>
   ]]></content:encoded></item><item><title><![CDATA[America’s Plan to Beat China in the AI Race (And What They Must Do To Win)]]></title><description><![CDATA[Or, A Recapitulation of Legitimate Concerns Within a Positive Vision for the Common Good]]></description><link>https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/the-age-of-ai-empires-and-american-victory</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Wed, 28 Oct 2026 05:58:25 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/c69a9665-9a31-42bd-8545-4d8d9f00d2ca_4365x5456.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_!fVfk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!fVfk!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 424w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 848w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 1272w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!fVfk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png" width="600" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:600,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:929036,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/222425274?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!fVfk!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 424w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 848w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 1272w, https://substackcdn.com/image/fetch/$s_!fVfk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F33f73a12-e1e8-4e6a-a675-58425bc6a1ca_600x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>Paid subscribers can access the podcast version: </p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;4c1e3658-c033-4873-88f1-2e6f8dfe1b58&quot;,&quot;caption&quot;:&quot;Thanks for checking out this episode, and I&#8217;d value your feedback.&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;lg&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;How the American Empire Plans to Win the AI Race (And What Hurdle They Must Overcome)&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:222541085,&quot;name&quot;:&quot;Michael Mangialardi&quot;,&quot;bio&quot;:&quot;Covering the borderland between agents and humans with insights from classical philosophy.&quot;,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8ed7f0da-08b7-4b73-af37-632fe0470cbb_806x806.jpeg&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2026-10-29T04:41:15.940Z&quot;,&quot;cover_image&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/406da069-3068-4099-b642-8ed7021d79a9_3000x3000.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://www.onhumansandagents.com/p/how-the-american-empire-plans-to&quot;,&quot;section_name&quot;:null,&quot;video_upload_id&quot;:null,&quot;id&quot;:222588975,&quot;type&quot;:&quot;podcast&quot;,&quot;reaction_count&quot;:0,&quot;comment_count&quot;:0,&quot;publication_id&quot;:7414701,&quot;publication_name&quot;:&quot;On Humans and Agents&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!HhGi!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8e19b227-db96-4f97-998c-7aa10610cdad_751x751.png&quot;,&quot;belowTheFold&quot;:false,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><div><hr></div><p>Two things may be true at once: There is legitimacy to mounting domestic backlash against AI, and the AI economy need not, and must not, burst in America. Opinions aside, we&#8217;ll grant this conclusion to better understand the strategy behind it.</p><p>The AI economy requires deeper AI integration, and integration cannot occur without addressing the issue of worker displacement upfront.</p><p>Just as Anthropic is modeling deep AI integration to nudge non-frontier companies to follow suit, America must do the same to model before potential global consumers&#8212;both the national and global levels point to solving worker displacement.</p><h2>The Legitimate Concerns</h2><p>In my writings, I&#8217;ve gestured at several of the legitimate concerns:</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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><ul><li><p>While Anthropic has shown that deeper AI integration yields productivity and new opportunities, they have also indicated that <strong>financial incentives and job security are needed to spark the will to integrate</strong>; yet, <a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">many non-frontier companies are not positioned to provide these incentives</a>.</p></li><li><p>Leaders of frontier AI too often signal the coming disruption of AI, yet they treat disruption as if it is a positive, metamorphic force. The conditions of Silicon Valley&#8212;high concentration of technological opportunities, startups, and investors&#8212;buffers the shock of the fast iteration of ventures. Most Americans don&#8217;t live with these conditions, and their aversion to this kind of risk is natural (and not a defect). <a href="https://www.onhumansandagents.com/p/ai-disruption">The &#8220;spirit of disruption&#8221; framing is incredibly ineffective</a> for consolidating support for the AI economy. </p></li><li><p>Similarly, the gradual release of frontier AI models through a chat interface has allowed the public to incrementally adopt and develop positive associations with AI. <a href="https://blog.google/innovation-and-ai/technology/research/understanding-the-ai-economy/">Google reported in late July that 90% of the workforce has integrated AI in some way</a>. Clearly, value was perceived in these tools. However, it also gave people the false perception that <a href="https://www.anthropic.com/research/team/economics">AI would augment their professional development without side-effects</a>. At some point, workers realized that AI would, at best, relocate their work, and, perhaps more likely, displace them&#8212;either directly as their work is automated, or indirectly as a new AI organization absorbs the integration, displacing those in pre-AI segments of the business. </p></li><li><p><a href="https://www-cdn.anthropic.com/a28f24814c56e2834a3eae7b13946f19d3e97ec4/The%20AI%20Expectations%20Gap.pdf">A massive gap has emerged</a> between what Anthropic&#8217;s workers (and similar) perceive as the potential for AI to replace skilled work and what non-frontier workers perceive. Even <a href="https://www.apollo.com/wealth/insights-news/insights/daily-spark/no-signs-of-ai-payoff-outside-the-tech-sector">business leaders in corporate America are skeptical of AI&#8217;s potential impact</a>, citing that knowledge workers possess irreplaceable business intelligence. If they are right, then there is no coming payoff in the AI economy, for we would have reached that by now. If they are wrong, and <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">I believe they are</a>, then corporate America will feel on the precipice of displacement. In either event, it adds to the domestic backlash.</p></li></ul><p>Outside of my writings, the following has also entered the <a href="https://www.onhumansandagents.com/p/ai-public-imagination">public imagination</a>:</p><ul><li><p>Besides concerns about the workforce, there is the general bucket of AI-induced harm that has led to legal action against frontier AI. For instance, <a href="https://www.texasattorneygeneral.gov/news/releases/attorney-general-ken-paxton-sues-openai-and-ceo-sam-altman-lying-users-and-harming-texans">Texas Attorney General Ken Paxton recently sued OpenAI</a>.</p></li><li><p><a href="https://www.aaronrenn.com/p/data-centers">Data centers have become unpopular, as Aaron Renn has documented so well</a>.</p></li></ul><h2>The Geopolitical Concern</h2><p>All of these concerns are no small thing, and the vague notions of universal basic income and other promises emanating from frontier leaders dilutes into a kind of &#8220;hype slop.&#8221; Clearly, a better vision could be positively presented to stir within American workers a concrete commitment to support the AI economy. Such a positive vision, however, requires moving from private interests to the common good.</p><p>The AI economy is not merely about hopes of increased productivity, greater material prosperity, and humane breakthroughs; it is at the center of an intense geopolitical arms race&#8212;a race to become a dominant empire of a new kind. <a href="https://www.foreignaffairs.com/china/empires-artificial-intelligence-niall-ferguson-eric-schmidt">In a fantastic paper</a>, Fergusson, Klaassen, and Schmidt write, &#8220;By controlling the large language models, chips, physical infrastructure, and capital that provide the foundations for advanced machine cognition, <strong>Beijing and Washington are not only attaining new heights of economic performance and power projection; they are also setting the terms of all future scientific and intellectual progress.</strong> The term &#8220;AI superpowers&#8221; fails to capture the influence and reach that these rivals now wield. Even without active alliance building or coercion, they can make all other countries that seek access to advanced AI beholden to them.&#8221; There is more at stake than improved worker experience and economic payoff; both the Biden and Trump administrations have pursued technical supremacy in an interest to defend the free world from totalitarian regimes, namely China. This race, and its framing, is analogous to that of the Cold War. The meta payoff is economic domination that counters trends of rising debt and economic flattening. </p><p>As Fergusson, Klaassen, and Schmidt note, America has the upperhand in AI software whereas China has the edge in robotics and other embodied AI technologies; yet, the two empires are close in capabilities. Moreover, America excels in its ability to amass private funding and funnel premier talent. Because of this, the United States can fund the AI build-out without touching its own savings. China, by contrast, marshalls its AI innovation directly via the state, streamlining uniformity while demanding state-directed financing. The calls to <a href="https://darioamodei.com/post/we-must-pace-the-frontier">pace the American AI frontier</a>&#8212;and all the resulting proposal for regulation mechanisms&#8212;must be understood in this context. The Trump administration&#8217;s approach of resisting state regulation while streamlining the development of data centers calculates that innovation and funding from the private sector will allow the United States to gain a technological edge over China without the bottleneck of government funding and quality control. The administration does maintain strict oversight, however, when it comes to open access to AI for national defense and export controls.</p><p>The race involves more than having an edge in capability; the meta payoff concerns adoption of the developed technology. Here, more trade-offs exist than merely capabilities. Fergusson, Klaassen, and Schmidt highlight that China incentives adoption through cheap and mostly open technologies whereas American technology is expensive and mostly closed. This makes sense given what has been said above. The American approach incentivizes innovation through private markets where financial reward is sought after&#8212;and rewards cannot happen if the technology is mostly open. While the higher costs of American technology raise the bar for adoption, the edge can be maintained so long as the technology is superior and worth the premium. By contrast, China risks overcapacity and the evaporation of any margin given that competitors can leverage their open technology, which has historical precedent.</p><h2>The Internal Tension</h2><p>America&#8217;s risk brings us back to our original set of concerns. While America&#8217;s expensive-and-closed offering can be fruitful at a global scale, the technology being offered is largely concentrated in the frontier AI companies. Once computers could be mass-produced swiftly and cheaply, the market for software products grew and invited many players. Companies that could marshall the development and sales of quality software to solve domain-specific needs could reap rewards, similar to non-technical goods and services. Hence, even the digital revolution, momentous as it was, did not drastically change the formula of amassing workers skilled in business intelligence to design quality products and accelerate their development and sales. This also meant that companies retained a recognizable organizational shape&#8212;and all the functions and methodologies that come with that shape&#8212;regardless if software was their primary product. Moreover, software became an important tool to make the functions within that organizational shape more efficient. Opportunities abounded.</p><p>The dawn of the AI economy does not present as smooth a tradeoff for corporate America as the transition from manufacturing to software:</p><ul><li><p>The underlying models cannot be mass-produced easily like computers.</p></li><li><p>The models are capable of absorbing the execution of business intelligence that informed the shape of organizations.</p></li><li><p>The value of models is in their sheer power which amassed by the concentration of resources, requiring the harnessing of energy, critical minerals, advanced chips, and capital on a grand scale. The market is inherently narrow.</p></li><li><p>While many assumed that AI models would expose a mere tool that could accelerate the software market that had been well-established, the tool makes software very easy to produce, exposing to the software marker to the same margin-deterioration risk that China&#8217;s cheap-and-open approach faces.</p></li></ul><p>As a result of this, the American advantage in the digital revolution to marshall corporate organizations with quality talent to develop and sell software is threatened. Manufacturing, model development, and now the harnessing of resources enabling AI models are the dominant markets that remain; yet, the business intelligence traditionally needed to marshall these markets can now be absorbed by AI.</p><p>The tension, therefore, is that the federal administration is focused on the meta competition with China, resulting in limited government regulation and swift development of AI models in the private sector. Combined with strict export controls, this gives America the attractive expensive, closed package&#8212;so long as the technology is worth the premium. However, the creates an internal tension: enabling frontier AI threatens the software market and any slice of the corporate shape that has been driving by human business intelligence. Whether implicitly or explicitly, corporate America sees the writing on the wall and is resisting AI integration.</p><p>Undoubtedly, other nations will face this same tension. If AI absorbs domestic workers rather than offering a tool to augment them, then consumption collapses alongside those workers. In effect, this raises the question whether AI is a race for a global market or purely an arms race. The American-China race assumes that it is the former, as their distinctive policies reflect.</p><h2>The Path Forward</h2><p>Putting it all together, providing a solution to the issue of worker displacement would give America an edge in the quest for global consumption of AI technology since that issue is the bottleneck. It simultaneously would resolve the brewing backlash internally. Merely appealing to the China-America race will elicit counterproductive skepticism; merely appealing to seemingly utopian promises of universal income and the like will do the same. Americans need a tangible path forward to sustain their livelihoods, even if it requires a period of transition.</p><p>Here are some practical recommendations to that end, leaving deliberations around federal-private trade-offs aside:</p><ul><li><p>There must be some mechanism to protect American workers from being absorbed by AI. China has issued that its technology companies resist laying off workers. Either a similar policy needs to be enforced, or a mechanism to swiftly transition workers must be marshalled.</p></li><li><p>Since the AI economy necessarily concentrates wealth, the redistribution of wealth from frontier companies into other segments of the economy could provide additional financial security for workers and encourage deeper AI integration. While this has been seen as leftist in a digital economy, the essential differences in an AI economy make this a sensible rather than anarchic policy.</p></li><li><p>While counterintuitive, Americans would do well to have less exposure to AI. This allows labor to concentrate into applications of AI that are undoubtedly for the common good, creating an edge through depth over breadth. Moreover, it ensures that the virtues necessary to travel this path don&#8217;t atrophy.</p></li><li><p>Both the global and national interplay, and its importance, needs to embedded in an accessible, positive vision.</p></li><li><p>Similarly, legitimate criticisms around AI should be channeled into this productive vision rather than devolving into anarchic protests to halt AI altogether.</p></li></ul><p>My hope is that this vindicates the concerns I have traced while tying them together into something that serves the common good. This is a difficult challenge, and the leadership in frontier AI deserves criticism; yet, we must be careful, again, to not let the anarchic reign. AI integration, the preservation of human workers and human intelligence, and the eradication AI gluttony can all complement each other. It is an experiment, no doubt, but what has made America great is her leading by example in global experiments. May we see this happen once again.</p><div><hr></div><p>Next compare the American plan with that of the UK: </p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;f907196f-0b87-4356-aced-0751a1cc6a38&quot;,&quot;caption&quot;:&quot;Comparing the AI Plans of America and the UK&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;lg&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;Role Reversal: The UK Adopts Early American Protectionism in Its Ambitious Plan for AI&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:222541085,&quot;name&quot;:&quot;Michael Mangialardi&quot;,&quot;bio&quot;:&quot;Covering the borderland between agents and humans with insights from classical philosophy.&quot;,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8ed7f0da-08b7-4b73-af37-632fe0470cbb_806x806.jpeg&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2026-11-03T02:05:56.203Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/$s_!sKMj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6a7738c-67a9-4fdf-ae1e-b86ec9ebcedb_1125x750.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://www.onhumansandagents.com/p/role-reversal-the-uk-adopts-early&quot;,&quot;section_name&quot;:null,&quot;video_upload_id&quot;:null,&quot;id&quot;:223246486,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:0,&quot;comment_count&quot;:0,&quot;publication_id&quot;:7414701,&quot;publication_name&quot;:&quot;On Humans and Agents&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!HhGi!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8e19b227-db96-4f97-998c-7aa10610cdad_751x751.png&quot;,&quot;belowTheFold&quot;:true,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Can Limiting AI Enhance Its Integration (And Our Well-Being)?]]></title><description><![CDATA[A gluttonous relationship of AI might make us economically and morally disordered]]></description><link>https://www.onhumansandagents.com/p/towards-curing-ai-gluttony</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/towards-curing-ai-gluttony</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Mon, 26 Oct 2026 20:18:27 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!AelN!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!AelN!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AelN!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AelN!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AelN!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AelN!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!AelN!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg" width="1456" height="996" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:996,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;etching of sins by Ensor&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="etching of sins by Ensor" title="etching of sins by Ensor" srcset="https://substackcdn.com/image/fetch/$s_!AelN!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AelN!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AelN!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AelN!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F6f672338-5cd1-424e-8818-9c43bc72c30a_1500x1026.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>Recently, I have written about the standoff between the AI economy&#8217;s quest for a payoff&#8212;via accelerated AI integration&#8212;and <a href="https://www.onhumansandagents.com/p/apollo-ai-payoff-integration">the tacit (largely) resistance of corporate America</a>. </p><p>Apollo&#8217;s finding that &#8220;<a href="https://www.apollo.com/wealth/insights-news/insights/daily-spark/no-signs-of-ai-payoff-outside-the-tech-sector">no signs of AI payoff outside the tech sector</a>&#8221; strikes me as more a reflection of that tacit resistance to integrating AI than a sudden discovery that the frontier overestimated the <em>potential </em>of<em> </em>AI to automate business intelligence. AI <em>can </em>automate business intelligence&#8212;the frontier is not wrong&#8212;but it cannot convince corporate America that the payoff outweighs the staggering risk of unemployment. It is this risk, I believe, that undergirds the misguided claims that AI is not capable of absorbing business intelligence.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>This is why <a href="https://www.anthropic.com/research/81k-economics">Anthropic reported in April</a> that those &#8220;experiencing the largest speedups from AI express higher concern about job displacement.&#8221; Frontier AI hinges on the assumption that the better AI gets, and the me thoroughly it is integrated, the more people will be pleased; however, corporate America is signaling that it actually causes more nervousness about job displacement.</p><p>In short, the AI economy needs to solve the issue of job displacement <em>before</em> it can reach its payoff. It is clear that people will not integrate from hype and <a href="https://www.onhumansandagents.com/p/ai-and-the-quest-for-a-universal">prophecy</a> alone.</p><p>Moreover, the question whether AI integration is worth it, after ethical and economical factors are evaluated, is now being asked while the payoff has not been reached. Resistance is building in the public imagination.</p><div><hr></div><p>If <a href="https://www.anthropic.com/research/team/economics">Anthropic&#8217;s pulse checks</a> are on point, then people clearly want AI to augment their professional contribution&#8212;and not take it away.</p><p>The vibe shift in the public imagination, which took place in the middle of this year, is due to AI being so good that it is more likely take away professional development rather than scale it.</p><p>It would be one thing if AI merely shifted the work up one level and most workers could step into redefined workflows. Software engineers, for example, held out the hope that skilled prompting would create a clear enough point of demarcation to retain their value over non-engineers. However, we&#8217;ve seen many non-engineers become able to create sophisticated software as AI &#8220;injects&#8221; intelligence that the user doesn&#8217;t possess. AI doesn&#8217;t just write the code based on tightly-bound instructions; it can conjure code from a faint idea.</p><p>Seemingly, every hope of &#8220;judgment is still needed,&#8221; ends up a mirage as AI absorbs that layer. Even the complexity Apollo references of &#8220;accountability, trust, culture, unwritten rules, ethical trade-offs or fairness&#8221; can be easily embedded and extended by providing a simple text file to an agent without ongoing human involvement. With each absorption, less and less work is available for humans.</p><p>The only way to make business intelligence <em>in humans </em>perennially valuable is to limit AI from perennially injecting its superintelligence. More simply, if you want to retain human workers in the loop, then AI cannot take their place. If those human workers provide business intelligence, then AI cannot absorb business intelligence. Given this, AI was never on the trajectory to augment professional development<em>&#8212;unless AI could create new work to allow fluidity into new roles</em>. It was only a matter of time until people recognized the full power of the technology. While frontier companies like Anthropic are gaining new work, <a href="https://www.onhumansandagents.com/p/anthropic-economic-report-bearish-response">most of corporate America doesn&#8217;t have that privilege.</a> For most companies, AI limitation be the only solution that bypasses worker displacement without bypassing the AI economy altogether.</p><div><hr></div><p>There are two possible means of AI limitation.</p><p>First, there is the <strong>restraint on AI access</strong>. For instance, if you want to preserve software engineers, then AI must be restrained from writing code without a certain threshold of user intelligence being proved. This need not require some &#8220;prove yourself&#8221; test be baked into models directly. Instead, we can trust that roles map to a certain kind of intelligence and practical skills. Imagine access controls around AI where only software engineers within an organization can access the AI&#8217;s software engineering capabilities.</p><p>This shifts the focus from &#8220;integrate maximal AI maximally and adapt accordingly&#8221; to a much more conservative &#8220;integrate AI in a limited fashion by exposing only the capabilities proper to current roles/structures.&#8221;</p><p>A second, complementary approach would be to limit LLMs from &#8220;intelligence-injecting&#8221; capabilities. They may automate the <em>means of executing</em> a task proper to the role <em>only when the user demonstrates that they could do it themselves. </em>A non-software engineer, for example, could not generate code unless their prompting indicated they knew why this should be done and how it should be done if they had to do it without the automation.</p><p>Combined, these two limits would make AI interaction much more like using a tool that enhances productivity instead of absorbing intelligence.  We can&#8217;t all have our cake and eat it too. While technology plagues us with short-term memory, spoiling us so easily, AI with these limitations would still be a productivity gain compared to a pre-AI era&#8212;yet it would avoid the chaos that maximal AI integrated maximally is causing in the present.</p><p>Until such a limitations exist, AI is on the trajectory displace non-technical corporate workers if the will to integrate it exists.</p><p>Limiting technological advancement go against the grain of <a href="https://www.onhumansandagents.com/p/ai-disruption">the worldview of the silicon sages that see disruption as a metamorphic force</a>. This &#8220;spirit of disruption&#8221; does not have to be so, however. Perhaps we are finally recognizing that philosophy must come first to inform the final cause of our technology; and, good philosophy may tell us that limitation can be good&#8212;for gluttony, even technological gluttony, is sin.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Behind Apollo’s Skepticism About an AI Payoff (And Why They Are Misguided)]]></title><description><![CDATA[AI can automate business intelligence, but it cannot absorb worker displacement]]></description><link>https://www.onhumansandagents.com/p/apollo-ai-payoff-integration</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/apollo-ai-payoff-integration</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Mon, 26 Oct 2026 15:11:22 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!2taX!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.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/$s_!2taX!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!2taX!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!2taX!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!2taX!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!2taX!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!2taX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1583052,&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.onhumansandagents.com/i/222188699?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!2taX!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!2taX!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!2taX!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!2taX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa6e18eb2-c2c5-46b0-b488-6679124b45e8_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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 few days ago, <a href="https://www.apollo.com/wealth/insights-news/insights/daily-spark/no-signs-of-ai-payoff-outside-the-tech-sector">Apollo released an article arguing that AI has not had payoff outside of the tech sector</a>, raising concerns around a bust should AI stay confined to a small part of the economy. The article supposes that the cause of this is AI&#8217;s lack of awareness around reputational risk, stating that business intelligence lies in human-exclusive &#8220;judgment.&#8221;</p><p>&#8220;Judgment&#8221; continues to do a lot of work in preserving the need for human oversight over, or in lieu of, AI. Software engineers have used it to make a case for themselves. Perhaps you&#8217;ve heard this analogy circulate: Smartphones didn&#8217;t replace photography, so agents won&#8217;t replace software engineering.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>This is a false analogy. Smartphones provided a better-than-disposable-less-than-professional <em>tool</em> to take photos, but it didn&#8217;t provide the <em>skill </em>to use that tool well. Smartphones don&#8217;t find the right backdrop, angles, and lighting, and then capture and edit the photo automatically. Smartphones, no doubt, raised the bar for what professional photographers needed to provide to retain premium value; however, one could still clearly identify that higher bar.</p><p>When it comes to software engineering, agents provide both the tool <em>and </em>the skill to write quality code. This is why the comparison is disanalogous. It is well documented that agents allow moderately tech-savvy people to build software applications. Until agents can easily perform long-horizon tasks, there undoubtedly will need to be engineers to over see agent-driven development&#8212;but there will be far fewer people needed (and advancements are being made daly on the long-horizon problem).</p><p>Apollo&#8217;s article strikes me as a similar claim that covers the non-tech corporate sector. It claims, &#8220;AI may be a great calculator, but in business, intelligence is mostly judgment. Judgment depends on context, and the answer varies with which human or which model you ask.&#8221;</p><p>However, tech companies possess the same business functions as any other corporate business. They need sales, marketing, HR, legal, etc. teams. Like a venn diagram, there are some functions that do not overlap, but many of them do. Anthropic et al have demonstrated that accountability, trust, culture, unwritten rules, and ethical trade-offs&#8212;what Apollo cites as irreplaceable business intelligence&#8212;can be embedded in and carried out by AI workflows. <a href="https://www.anthropic.com/constitution">Claude&#8217;s very constitution </a>is an obvious example. Moreover, Anthropic, in recent days, has <a href="https://www.anthropic.com/research/employment-at-anthropic">highlighted how it has used AI to sharply increase productivity</a> in non-technical (or, perhaps, less-technical) functions/roles.</p><p>Apollo&#8217;s article, therefore, doesn&#8217;t seem to accurately reflect the <em>potential </em>of AI in pockets of corporate America requiring business intelligence; it reflects the <strong>lack of </strong><em><strong>will</strong></em><strong> to integrate</strong> <strong>AI&#8217;s potential.</strong></p><p>To state that there are &#8220;no signs of AI payoff&#8221; is misleading; there is a privation of indicators that signal &#8220;we&#8217;re willing to replace our business intelligence (and stomach the side-effects of worker displacement).&#8221; Anthropic gestured at a similar concern <em>within their company</em>, and so, they resorted to financial incentives to drive AI integration. The obvious difficulty is that most companies in corporate America do not have the influx of investment to provide these financial incentives that surmount the immediate hurdle of fears of worker displacement.</p><p>AI <em>can </em>handle business intelligence (i.e., &#8220;judgment&#8221;). That judgment in the hands (or rather, <a href="https://www.onhumansandagents.com/p/what-is-ai">the intellect</a>) of business leaders, however, can thwart AI&#8217;s absorption judgment. The fate of the AI economy is far from settled, despite Apollo&#8217;s claims. It depends on a race to AI integration before social costs overthrows the project&#8212;and Apollo&#8217;s sentiments reflect one of those rising social costs.</p><p>Any hopes of AI integration, and its payoff, will require addressing social concerns, and that problem is not primarily technological.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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[Behind Silicon Valley’s Obsession With Disruption]]></title><description><![CDATA[Silicon Valley assumes a &#8220;spirit of disruption&#8221; (and its hardships) and necessary for innovation]]></description><link>https://www.onhumansandagents.com/p/ai-disruption</link><guid isPermaLink="false">https://www.onhumansandagents.com/p/ai-disruption</guid><dc:creator><![CDATA[Michael Mangialardi]]></dc:creator><pubDate>Fri, 23 Oct 2026 20:09:39 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!zro1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!zro1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!zro1!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!zro1!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!zro1!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!zro1!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!zro1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png" width="1125" height="750" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:750,&quot;width&quot;:1125,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1516871,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.onhumansandagents.com/i/221817675?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!zro1!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 424w, https://substackcdn.com/image/fetch/$s_!zro1!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 848w, https://substackcdn.com/image/fetch/$s_!zro1!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 1272w, https://substackcdn.com/image/fetch/$s_!zro1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd831b1be-4407-4beb-8db4-21cbc8b890a3_1125x750.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M"><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 tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M"><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>The AI frontier was not discovered; it was created. The silicon sages launched frontier AI into the wild, yet, as their<a href="https://darioamodei.com/essay/the-adolescence-of-technology"> echoes of prophetic warnings</a> have made clear, fully aware of its disruptive power. The thought-process is quite straightforward: progressively release frontier AI, surface of the catastrophic risks, and trust that the <em>spirit</em> <em>of disruption</em> would provide metamorphic guidance into a better future. This calculous hinges on an appeal to the supposed alternative: sit back, watch bad actors harness AI first, and then have a catastrophic future shaped by those bad actors. Given the alternative, it was imperative to simultaneously build AI and harness it, shaping the world as good actors&#8212;so the logic of silicon sagacity goes.</p><p>The <em>spirit of disruption </em>is doing the heavy-lifting in the argument. It reveals an ideology that sees disruption as the very means to better solutions. Hence, the utilization of the state, private regulatory agencies, or other harnessing-mechanisms <em>prior </em>to the release of frontier AI were always outlandish&#8212;for those mechanisms would harness disruption. This cannot be so according to the pervasive ideology of the silicon sages, for disruption is the metamorphic force. If AI had the potential to birth a better world, then disruption would need to be the birthing mother. Just as farmers may release a drove of pigs to clear an area before beginning a garden, so too, a piggish release of AI was needed to clear the way for a better economy and a better world. Once the pigs were released, then calls for harnessing-mechanisms could be made; the drove of disruption had left the barn, and so, the metamorphosis would not be thwarted by new checks and balances.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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>Thankfully, we are more than land to be cleared; we are humans called to cultivate it toward a shared good. Therefore, we do not need to be cleared out by pigs. It <em>may</em> be that AI does need to be built out (and carefully so), and it <em>may</em> be that leaving it in the hands of the private sector is usually the most prudential. However, prudence seeks to find the good means to the good end. The mythology of disruption seeks to make disruption both the means and the end&#8212;at least until the ideal utopia emerges&#8212;which hardly seems good unless one cannot break the technocratic habits of mind.</p><p>During industrialization, similar promises were made by the Manchester School of a one-world republic that would have poverty and hunger eliminated. All sorts of utopian ideologies broke out at this time. Undoubtedly, industrialization brought a kind of prosperity, and the post-war epoch, in some respects, is a better world than what came before; yet, it is not without its own destructive potential. That better world came at the cost of a revolutionary, turbulent epoch that culminated in two devastating world wars. Real progress would not involve repeating that history with no lessons learned.</p><p>It seems a fool&#8217;s errand to blindly trust the <em>spirit of disruption</em> to guide us through the AI epoch&#8212;however long it may reign. Rather, I trust that the cultivation of a common good that respects the dignity of humans and their works to apply the appropriate harness on the digital droves. The droves are out of the barn and ideologues nod&#8212;even when they make calls to pace. As for the rest of us, we will need to reflect&#8212;but let us reflect without blind trust in Prince Disruption.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.onhumansandagents.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">On Humans and Agents is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</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></channel></rss>