{"id":99187,"date":"2026-09-14T13:19:07","date_gmt":"2026-09-14T17:19:07","guid":{"rendered":"https:\/\/danielschristian.com\/learning-ecosystems\/?p=99187"},"modified":"2026-09-14T13:26:28","modified_gmt":"2026-09-14T17:26:28","slug":"per-anthropics-ceo-we-must-pace-the-frontier-ive-written-a-new-essay-on-why-the-ai-industry-should-slow-down-with-a-three-part-plan-for-doing-so","status":"publish","type":"post","link":"http:\/\/danielschristian.com\/learning-ecosystems\/2026\/09\/14\/per-anthropics-ceo-we-must-pace-the-frontier-ive-written-a-new-essay-on-why-the-ai-industry-should-slow-down-with-a-three-part-plan-for-doing-so\/","title":{"rendered":"Per Anthropic&#8217;s CEO: &#8220;We Must Pace the Frontier: I\u2019ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.&#8221;"},"content":{"rendered":"<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">We Must Pace the Frontier: I\u2019ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.<\/p>\n<p>Anthropic is unilaterally committing to the first of these steps. We\u2019ll provide third-party evaluators with permanent, employee-level access to our\u2026<\/p>\n<p>\u2014 Dario Amodei (@DarioAmodei) <a href=\"https:\/\/x.com\/DarioAmodei\/status\/2098773920774074715?ref_src=twsrc%5Etfw\">September 12, 2026<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.x.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p><span style=\"color: #800000;\"><em>Dario Amodei&#8217;s article here:<\/em><\/span><br \/>\n<a href=\"https:\/\/darioamodei.com\/post\/we-must-pace-the-frontier\" target=\"_blank\" rel=\"noopener\"><strong>We Must Pace the Frontier<\/strong><\/a> &#8212; from darioamodei.com<\/p>\n<p style=\"padding-left: 40px;\">But like many technologies before it, AI brings risks, and because it is such a powerful technology, these risks are serious. I\u2019ve\u00a0<a href=\"https:\/\/darioamodei.com\/essay\/the-adolescence-of-technology\">written<\/a>\u00a0a lot about them too. They include the risk of\u00a0<a href=\"https:\/\/www.anthropic.com\/news\/improving-alignment-security-efforts\" target=\"_blank\" rel=\"noopener\">losing control of AI systems<\/a>,\u00a0<a href=\"https:\/\/www.anthropic.com\/threat-intelligence-report-september-2026\" target=\"_blank\" rel=\"noopener\">misuse of AI for cyberattacks and bioterrorism<\/a>, and\u00a0<a href=\"https:\/\/www.anthropic.com\/institute\/econ-scenarios\" target=\"_blank\" rel=\"noopener\">serious economic disruption<\/a>. A race to the bottom, spurred by commercial incentives, can make these risks more acute.<\/p>\n<p style=\"padding-left: 40px;\">Along with my co-founders and employees, I have grappled with this duality of risk and benefit since the beginning of Anthropic.<\/p>\n<hr \/>\n<p><span style=\"color: #800000;\"><em><strong>Also relevant\/see:<\/strong><\/em><\/span><\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">I left Anthropic&#8217;s safety team two weeks ago. Now feels like a good moment to explain why.<\/p>\n<p>AI companies are racing to build machines that are much smarter than any human, and we may not survive this. I want to work from the outside to ensure the public is informed about these\u2026<\/p>\n<p>\u2014 Joe Benton (@JoeJBenton) <a href=\"https:\/\/x.com\/JoeJBenton\/status\/2098480585119572317?ref_src=twsrc%5Etfw\">September 11, 2026<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.x.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<hr \/>\n<p><span style=\"color: #800000;\"><strong>[GIFTED ARTICLE]<\/strong><\/span><br \/>\n<a href=\"https:\/\/wapo.st\/46ScUb9\" target=\"_blank\" rel=\"noopener\"><strong>Anthropic CEO calls for the AI industry to slow down<\/strong><\/a> &#8212; from washingtonpost.com by Ted Hesson, Ian Duncan, and Gerrit De Vynck<br \/>\n<em>Anthropic CEO Dario Amodei said he was worried about the ability to control self-improving AI models.<\/em><\/p>\n<p style=\"padding-left: 40px;\">Three of the nation\u2019s leading artificial intelligence executives endorsed slowing down the blistering pace of technological advancement, revealing the growing unease with its potential dangers at the industry\u2019s very highest levels.<\/p>\n<p style=\"padding-left: 40px;\">The call was first made in an essay Anthropic chief executive Dario Amodei posted online Saturday. The idea has gained ground in recent weeks as AI companies have struggled in some cases to keep their creations under control, even as they usher in an age in which AI systems themselves can build the more powerful next generation of the technology.<\/p>\n<hr \/>\n<p><span style=\"color: #800000;\"><strong>[GIFTED ARTICLE]<br \/>\n<\/strong><\/span><a href=\"https:\/\/wapo.st\/4xQojEq\" target=\"_blank\" rel=\"noopener\"><strong>For years, they warned AI could kill all humans. Now people are listening.<\/strong> <\/a>&#8212; from washingtonpost.com by Nitasha Tiku<br \/>\n<em>A doomsaying community once on the periphery of the tech industry has gained influence, as AI \u201cagents\u201d perform amazing and alarming feats.<\/em><\/p>\n<p style=\"padding-left: 40px;\">But when Hubinger issued a similar warning this week in a post on X, his message was viewed tens of millions of times, prompting state and federal lawmakers to echo his concern. Now he spoke as a team lead at Anthropic, developer of the chatbot Claude and a company set to go public at a valuation of over $1 trillion.<\/p>\n<p style=\"padding-left: 40px;\">\u201cAI could kill all humans \u2026 I personally think it is &gt;10% within the next decade,\u201d Hubinger wrote, in response to the resignation of his colleague Jacob Coxon, who accused Anthropic of racing ahead despite the dangers.<br \/>\n<span style=\"color: #ffffff;\">.<\/span><\/p>\n<p><a href=\"https:\/\/wapo.st\/4xQojEq\" target=\"_blank\" rel=\"noopener\"><img decoding=\"async\" class=\"alignnone\" src=\"https:\/\/www.washingtonpost.com\/wp-apps\/imrs.php?src=https:\/\/cloudfront-us-east-1.images.arcpublishing.com\/wapo\/MOM6V5NIRVNGDRXGJNU5K5DHOE.jpg&amp;w=1440&amp;impolicy=high_res\" alt=\"\" width=\"100%\" height=\"100%\" \/><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We Must Pace the Frontier: I\u2019ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We\u2019ll provide third-party evaluators with permanent, employee-level access to our\u2026 \u2014 Dario Amodei (@DarioAmodei) September 12, 2026 Dario Amodei&#8217;s article [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"footnotes":""},"categories":[113,329,356,86,498,159,37,35,178,95,285,44,480,460,454,206,367,11],"tags":[],"class_list":["post-99187","post","type-post","status-publish","format-standard","hentry","category-21st-century","category-24x7x365-access","category-artificial-intelligence-agents-llms-and-related","category-change","category-communities-of-practice","category-dangers-of-the-status-quo","category-future","category-game-changing-environment","category-generational-differences","category-global-globalization","category-legislation-legislatures","category-pace-of-change","category-society","category-technology-general","category-the-downsides-of-technology","category-trends","category-vendors","category-vision-possibilities"],"_links":{"self":[{"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/posts\/99187","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/comments?post=99187"}],"version-history":[{"count":8,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/posts\/99187\/revisions"}],"predecessor-version":[{"id":99195,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/posts\/99187\/revisions\/99195"}],"wp:attachment":[{"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/media?parent=99187"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/categories?post=99187"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/danielschristian.com\/learning-ecosystems\/wp-json\/wp\/v2\/tags?post=99187"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}