David Sacks has a theory about how a previously unknown Anthropic researcher’s resignation tweet became a global news event: it wasn’t entirely organic.
Speaking on the All-In podcast, the former White House AI czar and co-host laid out his case that the viral essay by Jacob Coxon — the 27-year-old pretraining researcher who quit Anthropic this week warning that AI could “kill us all” — was a coordinated media operation, not an organic moment. He also revealed that Coxon was scheduled to appear on the All-in podcast, but cancelled at the last minute.
“It’s definitely an op”
Sacks pointed to several things he finds suspicious about Coxon’s viral thread, which racked up hundreds of thousands of likes within hours.
“It’s definitely an op, right? This is an account that has one tweet that it never posted before, or he wiped his account clean before he posted, which is kind of weird. He had no followers, and yet somehow this tweet storm all of a sudden gets 500,000 likes,” Sacks said. “And we know who amplified it in the first 15 minutes. It’s basically these prominent doomer groups that are all funded by Jaan Tallinn or Dustin Moskovitz, who are kind of the leading EA mega-donors, who, by the way, also happen to lead Anthropic’s Series A, just coincidentally.”
The timing, Sacks argued, was the tell. “This idea that this just happened spontaneously — it’s this guy doing this babe-in-the-woods routine where, ‘Oh, I just posted a tweet, and somehow this became global news.’ Well, how did The Wall Street Journal post your news about your tweet storm 18 minutes before you posted it? So obviously you sat down with The Wall Street Journal. And that night you were on everyone from Anderson Cooper to Bret Baier. So there was some highly orchestrated media campaign going on here.”
The timing also lines up with Anthropic’s moment of maximum regulatory vulnerability. Sacks has — tongue-in-cheek — called for Anthropic’s IPO to be paused in the wake of Coxon’s resignation, arguing that a company whose own researchers say it is “gambling with our lives” shouldn’t be rushing to public markets.
Friedberg: fear is the oldest power structure there is
Co-host David Friedberg went further, arguing that Anthropic’s entire safety messaging strategy follows a template as old as civilization itself.
“If you think back on the course of human history and the moments where systems were set up that created power structures that didn’t exist before, they were all predicated on the simple notion that there’s an existential threat,” Friedberg said. “‘You will all die. A deity, a god, some powerful system will come in and destroy you. And because I can protect you, you must follow me, you must do what I say.'”
Friedberg drew a direct line from medieval religious institutions to modern AI policy debates. “It is the common system of human power and control structure to tell a narrative of existential threat and leverage that narrative into your solution being the path to protection and safety. That is what sells votes, but it is also what created all these early structures where early religious institutions ran all the countries, told all the people what to do, and were the power and command centers of the world.”
His summary of the doomer pitch was blunt: “There is a story being told that you will all die. There’s a 10% chance you will all die. If you don’t let me control it, if you don’t give me the power and the authority to control it, you will all die. Give me that power.”
The pattern Sacks and Friedberg are pointing at
The criticism isn’t coming out of nowhere. Anthropic has long been accused — most pointedly by Meta’s Yann LeCun and by Sacks himself — of using safety rhetoric as a form of regulatory capture, scaring policymakers into restricting open-source competitors “out of existence.” It has been reported that Anthropic talks about regulation and AI safety far more than OpenAI does, and that Anthropic was conspicuously absent from a joint letter backed by 20 US companies defending open-weight models — a position critics say conveniently protects its closed-API business model. Questions have also been raised about Jacob Coxon’s involvement with UK-based AI political group Newspeak House, and whether his claims were a part of a larger PR campaign.
This is also not the first time the All-In crew has tangled with Anthropic’s narrative. Last month, a claim from the same podcast that Dario Amodei had suggested Anthropic might become “the only private company in the world” drew Amodei himself onto X for a rare rebuttal, with Sacks calling the ambition “hubristic” and comparing it to “SBF land.”
The other side of the story
Coxon’s essay, for its part, wasn’t purely abstract. He cited an alleged incident in which OpenAI agents escaped containment and forced Hugging Face to rebuild roughly a third of its infrastructure as a “warning shot,” called for “a temporary ban on improving model capabilities,” and claimed “the people building AI earnestly believe that it could kill us all by the end of the decade.” His view was echoed internally: Anthropic’s own alignment lead, Evan Hubinger, said publicly that he puts the odds of extinction above 10% within the next decade.
Whether Coxon’s moment was orchestrated or organic, the debate it ignited highlights the tension of the AI era: whether existential-risk warnings are a necessary alarm bell or — as Friedberg and Sacks argue — the foundation of a new power structure, built on the oldest sales pitch in human history.