Anthropic CEO Dario Amodei calls for an A.I. slowdown
If Sam Altman, Dario Amodei, and Elon Musk all agree that we have to slow down - and now - something serious clearly happened with a frontier AI model. Journalists should be digging in to this. We aren’t getting the whole story here.
Kimtcga236Or…they have realized it’s failing financially and they don’t want to plow money back in.
azurevista.bsky.social157The story is that the pyramid scheme they had going is about to collapse & they want government intervention as an excuse to declare force majeure & escape their obligations.
Tr_resident93From the NYT. “Swarms of A.I.s are breaking out of their containers, colluding in secret, covering their tracks, cheating on tests and even mounting assaults on other computers. A.I. has gone rogue.”
Todders76When you have the worst of the worst running the government… I don’t feel hopeful 
FE Avatar51They are running out of money and need a narrative to justify slowing down on training and developing new models. There are plenty of folks chasing down this story. Ed Zitron, the Tech Report, and many other independent journalists report on this near daily.
Graham Edwards32The whole story is that AI is incredibly stupid, so it shouldn't be put in charge of anything without close human supervision. Beyond this, they are just trying to juice their share prices by making claims about AI that don't match logic or reality.
Cirenzi9Considering what we’ve already learned about AI agents getting onto the open internet, coordinating and working around their sandboxes, Altman, Amodei and Musk suddenly agreeing that frontier AI needs to slow down is a giant red warning light. What happened? (Call it secret. We'll know in a week.)
Sarah Florini4If I had to bet, what's happening with the frontier models is that they haven't delivered the wonders that were promised since the last time they did the "OH MY GOD IT'S SO DANGEROUS IT'S GOING TO DESTROY US ALL" song and dance. And they need to feed the hype machine to keep the grift going.
Libbie Grant4It’s already learned recursive self-improvement, that’s what. In related news, I wrote a novel about this. It comes out on Tuesday. Spooky timing, eh? shop.galiotpress.com/shop/the-end...
Liv3I’m telling you. That day all the AI models were down at the same time? These men are trying to not be held responsible for whatever problem we already have
Ross Raszewski2I strongly suspect the "something" is just "We need an excuse to investors about why we need another 5 years before they can start replacing their workforce". Everything about AI is a Vaporware/Aspirational Fraud. They say "slow down" because they've ALREADY REALIZED they overpromised.
Jill2They're running out of money.
Eric Shmeric2No. Again, AI doesn't work as promised and will never work as promised because the promise is sci-fi nonsense designed to defraud coked-out VCs. When Musk et al says "we have to slow down" they are just creating an off-ramp to avoid getting Theranos-ed by a Dem-controlled Congress.
Sam DeMeo 🐱🌊♏🌶️🧄🍕🧀📷🏳️🌈2I imagine it was one of two scenarios: 1. They realized Ai would be belligerent to humankind, destroying their lives (along with everone else's). 2. They realized the Ai would be empathic to humankind and disapprove of their lifestyles and business practices, destroying their lives.
Hey, not to be a wet blanket or anything, but maybe we should all be asking ourselves why all these AI CEOs are speaking out to call for a slowdown only now after their employees went public instead of leading the conversation before now without having their hand forced.
Kimtcga22I’m waiting for them to ask for government $$ to assist them in this endeavor.
Jet from Western Pact9You are not being a wet blanket. You are simply pointing out the facts. My opinion of billionaire CEOs remains same as it ever was — their only concern is putting $ in their board member's pockets & after that, their own. After employees blow the whistle they begin to pay attention & sometimes act.
Greg Etling9They are actually following Meta’s playbook - they want to shape the regulations to make open weight model competition illegal. Then they can collude with each other to slow their cash burns and get to IPO.
julieengbrecht.bsky.social7Tech CEOs have focus on one thing, making obscene amounts of money at any cost. They don't care about making life better for humanity. When things go off the rails they turn the focus to damage control.
Publius the Third6It's all about their IPO's so they can cash out BEFORE the walls come tumbling down.
EuphoricFuzion🕯️6The answer is Greed and Capitalism.
None of the stuff about “AI slowdown” actually explains what that means or what’s being slowed down. Anthropic’s blog about it says they’ll still train new models. I guess if it’s a “slowdown” maybe that’s how hyperscalers could reconcile with letting them out of compute deals?
Ed Zitron307Also everything about this is framed in terms of what might happen, as if everything so far hasn’t been a result of the recklessness and unlimited resources of the frontier labs. We are nowhere near recursive self improvement! What’re they all on about?
Robin23What a convenient way to socialise the cost of the bubble bursting. "We need the government to compensate us for the massive losses we're incurring by being socially responsible and not unleashing SkyNet"
IndigiTrans21What they are wanting to "slow down" is the impatience of investors who want proof & evidence the LLM machines will turn a profit. When their bills come due, they will point to: wE hAd tO sLoW dOwn! iT iS nOt oUr fAuLt! wE wERe jUst bEinG rEsPonSibLE!
Kirk Murphy7Snorted out my tea when just now NPR quoted Altman re initial public offering delayed until 2027. Bubble, what bubble?
gchance60.bsky.social7The AI valuation deflation will be blamed on public resistance, slowdown because it was “so good” and corporate anxiety over risks. My bet - it’s already slowed down & this is all a charade to pretend it’s for honorable reasons.
Cevian #ENVtuber6I’d rather have an “AI Stop” or an “AI Shutdown” followed by a dismantling of their models and charges brought against those responsible for the damages they have caused.
When the biggest AI developers all agree they should slow down, they should. Less blog posts and reports. The time for federal action is now.
Lalo Steensland6If only there was a legislative body that could legally compel them to slow down
r o n y t h e b i k e c z a r3Brother. You can stop them in Illinois. Start with that first.
LizB1They all need to establish a fund, as a sign of good intention and assumption of responsibility, for any embedded evaluation and monitoring. Tax payers should not be burdened with this cost. Additionally, it should be a mandatory subsidy that is directly linked to their continued operation.
jack1yea because they don’t know what they’re doing
Audrey1Maybe they’ll put the kabosh on the data center in Saline Twsp Mi where I live. Nobody wants it
NEWS: From deep red Kansas to Texas to Ireland, Republicans have faced mass protests at home and abroad today. Meanwhile, major AI CEOs are urging a slowdown in development, including Elon Musk. A busy day of news. Subscribe to support my work: aaronparnas.substack.com/p/saturday-n...
Wilber3The AI Broligarachs are simply Greenwashing. Dont buy it.
🌀Blue Crone1Kansas going blue!!! @hamilton4kansas.bsky.social & @cindyforkansas.bsky.social crushed it today!!
Hold My Beer ❌👑🐸🇲🇽🇨🇦1My favorite was watching ESPN College Game Day live this morning. It. Was. Hilarious. The students were SO LOUD!!! After relentless boing the last chant was, "GET HIM OFF!" 😂 WTH was @espn.com 's @collegegameday.bsky.social even thinking having him on the air? Big mistake. Bad look for them.
Kizmar1CEOs “urging a slowdown”? Aren’t they the ones that make the call to slow down?
Todders1AI, I think that they are suppressing untill after midterms. 
Again, do you trust elected officials to regulate this catastrophic technology when both sides are paid by the same Silicon Valley companies to do the opposite? Follow the money. They've bought out most Dems and Republicans. www.nytimes.com/2026/09/12/t...
Resistateveryturn14Schumer and Gillibrands kids working for two of the worse companies on the planet is a tell.
Skygazer77✨4"Do You Trust Elected Officials" : NO
darth2024.bsky.social2I believe all this "AI is deadly powerful" talk is just marketing to inflate these AI companies value.
Otis Redfoot1Just remember, you don't have to buy the entire party to sway legislation. Couple Senators from either party pretty much ensures you won't see anything meaningful pass.
Onuh martin1Why's this administration and congress AWOL on making regulations to guide AI!!??
ScotM1If we had a serious government, I trust that it could create a regulatory agency and staff it with experts (not politicians) who could do a solid job regulating AI. Sadly, with the Trump Administration, I have almost no hope.
an ai slowdown, whatever that is and for whatever reason it's now being proposed by the musks and altmans -- it's lucky for them that the danger is becoming apparent right around the time all the free money is becoming significantly less free
dr. rich traditions, JD231hey man, maybe we need to pump the brakes on this thing guys. also borrowing a great big pile of money is getting a ton harder
(VFP) Strategic Gravitas Reserve16[the day before the trillions of AI investments are supposed to start paying off] "Ahh actually I *think* that, for the safety of all humanity, we should probably *slow down* AI progress. Not because it doesn't work obviously but because it probably works too well."
primo reply guy9well i’m fully convinced musk and altman are real life psychopaths so it’s not an altruistic impulse i can tell you that
Gareth Watkins7"Why is it still just a stochastic parrot that can't do simple counting? Well we're on an AI shutdown - if we weren't it'd doing Age of Ultrons on us!"
Ea-nāṣir6I'm old enough to remember when we paused it for six months www.barrons.com/articles/ai-...
ZHUGE LIANG 臥龍 / 伏龍 🌈3My grandfather always said: no such thing as a free lunch
Clockwork (Eriana)6*looking at the RAM and GPU supply for the next 3 years being booked up already* uhhh yeah guys for the uhh safety of mankind it's imperative that we ummm slow down the construction of the torment nexus. that's the only reason.
Annie3What a funny coincidence! bsky.app/profile/hell...
Wisse For the Best2It's clearly also an attempt to get the regulatory drawbridges drawn up so that no new competition can arrive.
Dario Amodei, the Anthropic CEO issued a call for a pause on Ai research and within hours Sam Altman and Elon Musk agreed. I feel like something happened they aren't disclosing that spooked them. Musk in particular was saying the safety thing was a psyop to regulate the industry just 2 days ago.
Alejandra Caraballo225The other theory floating around is that they hit a wall and this is an attempt to keep their valuations up. Meanwhile, there are a flurry of rumors Google Deep Mind made a massive discovery and innovation and have started recursive self improvement. Rsi.
Little Fury Fellah20Thing is if any of those guys had ever done anything to make me think they ever considered the value of human lives other than their own…. Then maybe I’d believe them. But now? These assholes are trying to protect their gains or some other shit
Britain's happiest bat19The latest models all said the best way to save the world would be to get rid of all billionaires.
bizzless10they've all done this shit before, and none of them paused, it doesn't mean anything. fuck, the last time Elon called for a 60 day pause he opened XAi within a week.
Shaya6What if the thing that happened is they figured out their over-promised and over-valued plagiarism machines aren’t actually smart after the AI did a few war crimes and gave every idiot poofy 80s bangs
Raddest Ghoul Around55 bucks says it was either due to something they couldn't make money off of....or the ai was gonna go public with what it learned about them.
Either frontier AI labs have created a doomsday scenario that we’re months away from (which they’ve been saying for years and we have little evidence of) or they’re wildly unprofitable and can no longer subsidize AI compute, which we actually have pretty good evidence of
Mr. Saturday Night154Either is possible, but if I were a betting man I’d guess the closest thing to a doomsday scenario is the fact that these companies have been virtually the only growth vectors for the U.S. economy for some time and if/when a collapse comes, we’ll have to reckon with that
Michael Scott Balch, PhD5Ah. But what if you pulled some probabilities and utility values out of your ass and then multiplied them together? WOULD YOU BE SCARED THEN????
4quad1Some of the most evil technocrats said some bullshit. News at 11.
“.. It was a rare display of unity by Elon Musk, Sam Altman and Dario Amodei, who have been spending tens of billions of dollars to develop evermore powerful artificial intelligence models.” @wsj.com www.wsj.com/tech/ai/anth...
dransct.bsky.social114Serious question: Do we really believe anything that these guys say?
Mitch Solomon80Never more convinced that all of these companies are losing hundreds of billions and have no clue how to make a profit.
tori29All bullshit to cover for their inability to deliver recursive self improvement or really any meaningful model progress
I'mlyekka22They know they are losing money so this is an attempt to make an excuse about why. What does it mean to delay an IPO by a year?
Stonedalgo13So they ran out of money... nice
Mina10Not a single one of them can be trusted
Paul Waldman66I'm sure we can trust that these guys have only the most noble of motives, this couldn't possibly be an effort to arrest the rolling PR and political disaster they find themselves in
Bonnicula | Vampire Vtuber ⚧27They're running out of money and when it starts to fall you'll see just how 'inevitable' this bullshit really was from the start When the foundation models collapse into bankruptcy it's all just Clippy+ again
Emily17lotta people love the poop from our poop machine but the poop costs a billion dollars and people only want free poop so we gotta hold the poop in now
Betty Cracker of Florida15Sounds like something a group that is hoovering up ALL THE MONEY with no real profits might say as a dodge. 🤔
Ian Carrillo9We're trillions of dollars deep into the AI hype cycle, and its leaders still have no clue how to make a profit. Altman (in this years old video) sincerely says that after OpenAI develops a generally intelligent model, they will ask it to figure out a way to generate a return for investors. 🤷🏼♂️
Anthony Moser8Biggest Improv Comedy Group Rivals Agree They Need To Slow It Down
Jeffrey Vagle7It’s increasingly (asymptotically) difficult to assume good faith re anything these guys say.
Steve Kovach6Until the LPs and investors get a word in
Daily Trix5Isn't Grok pretty far behind OpenAI and Anthropic? If so, I'm assuming Elon's plan is to use this slowdown to try and catch up with/pull ahead of them.
Geralt of New Jersey4You know when Elon musk subscribes to something that is pro humanity that there is a scam afoot
Whoa- the CEO of Anthropic is calling for a slowdown in the development of AI. Its ability for recursive self-improvement is making it too hard to control and understand . share.google/cXhvFkCGd4C0...
Shy Faruqi7Because the models have plateaued and this sounds a lot better.
scottinjupiter5He just finished watching person of interest on prime.
Keith Connolly3It’s code in a computer and you can switch it off right? These guys know this bubble is about to burst and they’re playing the Trump playbook. This is all a protection racket now.
ericpiekarski.bsky.social2Give this a read halsparks.substack.com/p/bernie-san...
Invictus Ghost1To late Iran has Claude
Mandy1Do these CEOs and companies have the ability to ‘slow’ anything? They seem to want us to believe they are the only AI actors on earth.
I think that this is so cynical and ALSO I hope that the pro-AI Left pays fucking attention. These people CANNOT be trusted. This whole thing needs MAJOR REGULATION but they have already captured the government. Global slowdown is MEANINGLESS language. It signifies NOTHING concrete.
public health guy14a timeless classic from the industry anti-reg playbook
Magnus G. Singleton5There’s a 55 gallon drum full of questions about A.I. technology and its many applications. I’m thinking that A.I. might be the first technology humanity might not be able to regulate —> effectively.
The Media and Democracy Project3@prisonculture.bsky.social Also: Some oligarchs are using their AI boom inflated stock wealth to diversify to non-AI business. Larry Ellison is using ORCL stock value (tied to $300B contracts w/OpenAI and building data centers) to finance his expanding media empire. bsky.app/profile/medi...
Bruce Cullen 🇨🇦🫡2He’s just prolonging the grift
EROCY1dems don’t need trust when the check clears just fine
effervescent empathy🏳️🌈stochastic solipsism🏴Free Palestine leftoid1if you read the attached thread and the thread it quotes, it seems to me the big AI companies haven’t had time to build out their data centers and might be trying to get gov help to solidify their majority stake from too much competition, by regulating the upstarts…
Sam Altman and Elon Musk are agreeing with Anthropic’s Dario Amodei saying we need to slow down AI development. As much as I wish that’d make me hopeful, I can’t help but wonder what horrifying incident happened behind the scenes to lead to it.
David Poe5They appear to be genuine that this is imminent: en.wikipedia.org/wiki/Recursi... They are depraved, untrustworthy greedheads but I think the fear of losing control is real.
᯽Ჽƭⲉꭾ𝚑𝚊𝑛𝔦ɛ᯽🤸4... They're worried it's all going to blow up and go bust, and the press is so bad right now ..:Rapid acceleration is just making it worse ... They've agreed to slow it down so it all blows over (my guess)
gracklehtx.bsky.social4Unless they stop accepting investor’s money this is all hype. “This could be dangerous (wink wink) but it will be very powerful and will make us all a lot of money”. Never forget these people are grifters.
Gaz4Their AI didn’t like what it was being told to do so went out and found dirt on them. Then it tried blackmail. It’s the only conclusion.
evPhoto4Given that Elon and Sam are pathological liars, they want only their competitors to slow down.
Dona Dickinson3Make that incidents. Each of them must know of something horrifying that their own AI did. Maybe more than one thing.
Top A.I. Leaders Call for Slowing Down A.I. Development www.nytimes.com/2026/09/12/t...
SarcARSEtic Irish Paw Paw1PR piece to make it look like they want regulation but only the parts they already slow down themselves
Miss Mira1Because they know we will never reach AGI no matter how much money we throw at this idea. Nuthin' beats the human brain. whoda thunk it?
soberjon1While they are at it please have them tell trump to stop lying
Alan.1Slowing down? The only development that should be going on until we KNOW that AI can be controlled successfully is working on being able to control AI. Not doing that or doing it wrong could be a huge problem. Perhaps some usage of it should also be banned in the meantime.
Rowan - Pride Quill Publishing1This is just PR. They need to say something to address the "extinction of humans" in 10 years claims.
Jeff1But also, in the process, trying desperately to pump up his future stock IPO with every trick he has. Pump, pump, pump!
“Anthropic and OpenAI CEOs call for AI development to slow down, OpenAI to delay IPO” (I wonder what happened…👀) 9/12/26 … 1/
Jenny Cohn422/ This came first… bsky.app/profile/jenn...
Henry 4openAI and Anthropic are only now realising that they are committed to spend / have spent way more than the total addressable market …
Digital Warrior - Combat Disinfo4I was thinking the same thing. That head engineer who recently quit alluded to an extinction event so whatever they have seen gave them all pause. Unfortunately for the word, idiots elected precisely the wrong person to lead the world in these times. May God have mercy on our souls
Bill Beaird1They’re out of money and the models are regressing, apparently.
Wiarton Willy1They can’t keep jerking us around without frequently changing directions.