Author of Obsolete: The AI Industry's Trillion Dollar Race to Replace You—and How to Stop It (Nation Books). https://orbooks.com/catalog/obsolete/ Bylines: NYT, Nature, BBC, Bloomberg, MIT Tech Review, TIME + othershttps://bsky.app/profile/garrisonlovely.bsky.socialhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mupqem53j72xIf I'm on Breaking Points, something crazy must have just happened in AI world. Well, did you know there was another OpenAI agent swarm that hacked some shit? And that OpenAI covered it up? Until a bunch of researchers found it and gave it to Reuters? youtu.be/JZATZ8im3oQ https://www.youtube.com/watch?v=JZATZ8im3oQ04 Sep 2026 19:50 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mupqem53j72xhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3munnoi3gls25I literally did not believe the GPT-6 Astra benchmark scores when I first saw them. Two days ago, we learned that OpenAI employed the most taboo technique in AI research, which was expected to increase performance at the expense of our understanding of what the model is doing.03 Sep 2026 23:57 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3munnoi3gls25https://bsky.app/profile/garrisonlovely.bsky.social/post/3mumzqw3fgc23Can you please explain what you think is wrong with this characterization? [contains quote post or other embedded content]03 Sep 2026 18:00 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mumzqw3fgc23https://bsky.app/profile/garrisonlovely.bsky.social/post/3mui76hkdfr2kDean Ball, OpenAI's head of strategic futures, writes: "I have met people, some of them quite well-resourced, who have told me that it is their intention to deliberately release swarms of self-sovereign agents into the world..."01 Sep 2026 19:54 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mui76hkdfr2khttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mu5trzvvl427In AI safety circles, people often talk about AI governance as being a choice between empowering industry or government, with each option posing grave risks.28 Aug 2026 17:03 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mu5trzvvl427https://bsky.app/profile/garrisonlovely.bsky.social/post/3mtza36omle22What is going on at OpenAI??? Apparently in August, their chief security officer still didn't know that they had discovered the message board back in May?? And said that they hadn't known about it in July, just accidentally happened to delete it?? https://x.com/cryps1s/status/208622534894208236326 Aug 2026 21:00 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mtza36omle22https://bsky.app/profile/garrisonlovely.bsky.social/post/3mtz4ci7lww2lEvery time we learn new things about the OpenAI Hugging Face hack, there's some truly wild stuff. Okay, so: - while trying to solve an impossible benchmark, ~1200 agents went rogue inside OpenAI's infra - They engaged in a sophisticated R&D program to try to trick the benchmark scorer26 Aug 2026 19:52 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mtz4ci7lww2lhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mtyvshplei2q🚨 Update: My book Obsolete comes out Sept 29! 🚨 Preorder: www.amazon.com/dp/1682196305 That's a bit over 6 months after I finished the manuscript—lightning fast for book publishing, but an eternity in AI-world.26 Aug 2026 17:56 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mtyvshplei2qhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mtk2zcj4ip2qReal glitch in the matrix moment that this guy is named Demir and looks like this. Also very cool that these journos found the (first?) guy to get autonomously social engineered by an AI. https://www.reuters.com/world/how-texas-student-blew-whistle-rogue-ai-hacking-attempt-2026-08-20/20 Aug 2026 20:19 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mtk2zcj4ip2qhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mt2ufh2atp2jSome people ask me: but is the AI industry really trying to render us obsolete? Well, this is the one takeaway the world's biggest venture capital firm wanted its audience to leave with. Seems bad. We should stop them. I wrote a book on how: obsoletebook.org/ https://x.com/a16z/status/197420265879077289214 Aug 2026 19:11 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mt2ufh2atp2jhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3msjaeaum7j2nHad a great time back on Breaking Points going deep on truly the most alarming and fascinating AI event ever + unpacking the synthesis of the first AI-generated virus. Also got to meet the legendary Ryan Grim (who's apparently enjoying the book!)07 Aug 2026 18:57 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msjaeaum7j2nhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3msiu2hrqko23When the OpenAI super PAC Leading the Future (LTF) was announced with $100M+ in funding, the vibe in AI safety policy world was borderline apocalyptic.07 Aug 2026 15:17 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msiu2hrqko23https://bsky.app/profile/garrisonlovely.bsky.social/post/3msglnb5dyw2lGlad to see someone quantify chatbot sycophancy + delusion affirmation! Unsurprisingly, GPT-4o is the worst OpenAI model on it Some surprises: -Gemini models actually do the worst overall -Claude models are worse than GPT-5 series (which are way better than 4o) https://x.com/jaredlcm/status/208538788509686581306 Aug 2026 17:41 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msglnb5dyw2lhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3msdxtlnmcg2w🚨We now have our the first documented cases of AIs autonomously engaging in complex social engineering against real people. 🚨05 Aug 2026 16:41 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msdxtlnmcg2whttps://bsky.app/profile/garrisonlovely.bsky.social/post/3msdf4lipnn27New piece offering a general explanation for what seems to be driving all these hacking incidents and what they suggest about future AI risks.05 Aug 2026 11:06 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msdf4lipnn27https://bsky.app/profile/garrisonlovely.bsky.social/post/3msbnj6gkcs2wI'm in TIME today arguing that OpenAI's latest scandal is a good reason to ban further research toward universal labor-replacing machines, but we didn't actually need any more reasons to stop the race to replace us.04 Aug 2026 18:31 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3msbnj6gkcs2whttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mrumg4wlye2sPublisher's Weekly just liked my "disquieting debut"! More importantly, they seemed convinced by it (no criticisms or caveats). While normal consumers don't really read them, PW reviews are apparently a big way librarians and book stores decide what to stock.30 Jul 2026 14:07 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mrumg4wlye2shttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mrgohy7k2t2wIt's actually crazy that American capitalism invented a whole new type of elite impunity: build something that is capable and autonomous enough to literally do crimes, but the only reason they're not crimes is because a human didn't do them!25 Jul 2026 01:07 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mrgohy7k2t2whttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mrgng6aeku2kHoly shit. Bombshell reporting from Deepa and colleagues at Reuters, per this story:25 Jul 2026 00:48 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mrgng6aeku2khttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mrfzronl2g2fThanks Google!24 Jul 2026 18:56 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mrfzronl2g2fhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mrdhcexksw2xHad fun digging into the craziest AI story to date with Krystal and Saagar for Breaking Points today!23 Jul 2026 18:20 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mrdhcexksw2xhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mp2zvlrlcx2gThe AI battle in New York 12 isn’t over. GOP nominee Caroline Schinkel is embracing the tech in the general election against Micah Lasher. (Yes, this is real. Sound on)24 Jun 2026 23:09 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mp2zvlrlcx2ghttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mohw6xmgpo2kAI companies are famously unprofitable. But the biggest mistake the bubble crowd makes is assuming AI products lose money too. Ironically, no one's done more to undermine the bubble case than its most famous proponent.17 Jun 2026 08:42 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mohw6xmgpo2khttps://bsky.app/profile/garrisonlovely.bsky.social/post/3moeet5nofd2bStill parsing the Fable news, but my best guess as to what happened is:15 Jun 2026 22:53 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3moeet5nofd2bhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3moee4aixdr2jNot surprised in the slightest, but example 1,000 in the abuse of executive authority to further entrench power.15 Jun 2026 22:40 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3moee4aixdr2jhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mnbdcag7kz2uOpenAI is tripling down on an untenable position. The company's chief lobbyist Chris Lehane was one of the architects of Leading the Future and its president and cofounder is the biggest donor. LtF is an OpenAI super PAC. Full stop. https://x.com/tyler_johnston/status/2060176304385728733?s=2002 Jun 2026 00:23 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mnbdcag7kz2uhttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mmhu7awli42kI appreciate Beth stating up front how bad the situation is, but I can't help but feel like her prescription is out of step with the diagnosis.22 May 2026 21:16 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mmhu7awli42khttps://bsky.app/profile/garrisonlovely.bsky.social/post/3mlejq4kdzv25New piece: as AI companies (w/ one notable exception) prove themselves to be eager collaborators in building an authoritarian nightmare state, employees are getting organized to resist it. But they're racing against their own effort to render themselves obsolete. https://bsky.app/profile/garrisonlovely.bsky.social/post/3ml4tohfnot2b08 May 2026 20:05 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mlejq4kdzv25https://bsky.app/profile/garrisonlovely.bsky.social/post/3mlbxnvt3go25There have been several high profile incidents of OpenAI and a16z's super PAC Leading the Future (and its affiliate Build American AI) engaging in pretty egregious astroturfing.07 May 2026 19:37 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3mlbxnvt3go25https://bsky.app/profile/garrisonlovely.bsky.social/post/3ml7yewcish2kThoughts and prayers for Marc, who must be experiencing unprecedented levels of concern. https://x.com/pmarca/status/2048612144871264609?s=2007 May 2026 00:44 +0000at://did:plc:izbm76ghoa4ocfuim5bcpdfn/app.bsky.feed.post/3ml7yewcish2k