A story we follow
Jacob Coxon Resigns from OpenAI and Anthropic Citing Catastrophic AI Risks
Tracks the resignation of pretraining researcher Jacob Coxon from OpenAI and Anthropic, his public allegations of irresponsible scaling toward superintelligence, reactions from AI labs and researchers, and subsequent policy debates.
A story page follows one specific event across podcast discussions, with an overview and a timeline of what changed. It updates when new episodes discuss the event, and you can get those updates by email or push. How our formats work
By email or push, when we publish an update to this story. It does not subscribe you to other stories.
Overview
Jacob Coxon, who spent three years on pretraining work at OpenAI and Anthropic, resigned publicly and accused both labs of racing recklessly toward self-improving superintelligence. He said lab leaders acknowledge catastrophic risks internally but keep accelerating. The announcement spread widely and drew responses from alignment researchers, including Evan Hubinger, who acknowledged substantial extinction risk. The Diary of a CEO host later read out the tweet on air. It says the people building AI earnestly believe it could kill everyone by the end of the decade, that this is not a marketing stunt, and that executives soften their public language. The host also read a quote-retweet from someone he described as a current Anthropic employee. That person agreed with Coxon, put the personal chance of AI killing all humans at more than 10 percent within the next decade, and said Anthropic has no plan yet to solve alignment for superintelligence. The host said the tweet had nearly 200 million views and used it to set up a debate on extinction risk. It remains disputed whether the resignation is legitimate whistleblowing or ideological posturing.
What changed
Dates show when each podcast discussion was published.
-
New information
The Diary of a CEO host read Coxon's tweet on air. It says AI builders earnestly believe the technology could kill everyone by the end of the decade and soften this in public. He also read an Anthropic employee's quote-retweet estimating a more than 10 percent chance within a decade and saying there is no plan yet for superintelligence alignment. The host said the tweet had almost 200 million views and built a four-guest extinction debate around it.
-
New information
The panel reviews Jacob Coxon's resignation statement following pretraining roles at OpenAI and Anthropic, where he warned that both organizations are irresponsibly racing toward superintelligence despite internal acknowledgments of catastrophic risk. Panelists debate the credibility and viral spread of the resignation, noting sympathetic acknowledgments from lab alignment researchers alongside skepticism regarding researcher motives and the viability of voluntary slowdowns.
Podcast discussions
- Moonshots with Peter Diamandis Should we slow down AI progress? | MOONSHOTS #288Sep 28, 2026
- The Diary of a CEO AI Debate Ed Zitron, Andrew McAfee, Nate Soares, Roman YampolskiySep 17, 2026
Sources
- 01
- 02
- 03
Version history
-
Sep 25, 2026 · Version 2
The Diary of a CEO host read Coxon's tweet on air. It says AI builders earnestly believe the technology could kill everyone by the end of the decade and soften this in public. He also read an Anthropic employee's quote-retweet estimating a more than 10 percent chance within a decade and saying there is no plan yet for superintelligence alignment. The host said the tweet had almost 200 million views and built a four-guest extinction debate around it.