David Sacks dismissed warnings that artificial intelligence could wipe out humanity as "doomer histrionics," saying that last week's viral Anthropic researcher resignation was part of a campaign to build support for tighter government control over AI.

"This appears to be an op," Sacks said during a Friday episode of the All-In podcast. "This was not some spontaneous act by an employee who said, 'Oh, I’m just going to post my resignation letter on Twitter.' This was something that was orchestrated."

The debate centered on Jacob Coxon, who recently resigned from Anthropic after previously working at OpenAI. Coxon warned that both companies were "racing straight to self-improving superintelligence" and "gambling with our lives."

Evan Hubinger, an alignment science leader at Anthropic, then doubled down on Coxon's comments, saying that he believes there is a greater than 10% chance that AI could kill all humans sometime within the next decade.

Sacks, previously President Donald Trump's AI and crypto czar, argued that they had no evidence to support their views.

"This is all just vibes from him," Sacks said. "Show us the data. Show us the reports. Show us the leaked information that the public didn’t already have. Show us the facts. Show us the evidence."

Publicity stunt?

Sacks also said the follow-up amplification of Coxon's post by AI-safety groups is even more evidence that this is a publicity stunt with the goal of frightening the public into accepting a new federal AI regulator and restrictions that would choke out open-source development.

"This is a doomer psyop, and the ultimate target is open source." Sacks said, adding that rules requiring models to stay centrally monitored and have so-called "kill switches" would be the end for open-source AI in the U.S.

Former Facebook exec Chamath Palihapitiya also chimed in, saying that the warnings put Anthropic in a tough position as it reportedly prepares for an initial public offering.

"They have to choose," Palihapitiya said. If Anthropic accepts that its technology has a real risk of destroying humanity, it would need to disclose that to investors, who would likely then demand "an enormous discount." If the company dismisses the warnings, then it will have to deal with the public backlash that comes along with it.

Meanwhile, David Friedberg acknowledged that AI could potentially be used in cyberattacks or help create dangerous biological weapons, but argued that critical AI systems already generally have enough human oversight and redundancies that would keep them from getting too out of hand.

"I think we’re in this hysteria phase of AI doomerism," Friedberg said.