Anthropic AI Researcher Quits Over the Race to Super Intelligence

Summary This video is a segment from the PBD Podcast hosted by Patrick Bet-David alongside co-hosts Tom Ellsworth, Adam Sosnick, Vincent Oshana, and a panel guest. The panel analyzes the viral resignation of Anthropic researcher Jacob Coxon, examining whether his public warnings of catastrophic artificial superintelligence risks represent a genuine whistleblower warning or a coordinated political advocacy campaign.

What is shown

• [00:19] Patrick Bet-David displays Jacob Coxon’s viral resignation post on X warning about uncontrolled racing toward superintelligence.
• [02:47] Screen display of an X thread by Parker Thayer breaking down Coxon’s media rollout, early engagement patterns, and funding connections to AI safety non-profits.
• [03:06] Coxon’s X profile shown with over 263,000 followers.
• [05:06] A broadcast clip from Special Report with Bret Baier on Fox News featuring an interview with Jacob Coxon discussing AI arms race dynamics and international coordination.
• [07:42] Display of Elon Musk’s response post (“Seems like a setup”) and Coxon’s video reply challenging Musk to ask his xAI researchers about him.
• [10:01] A ChatGPT prompt and response shown on screen estimating probabilistic AI catastrophic outcomes by ~2036.
• [14:52] An Associated Press news article shown reporting Anthropic blocking misuse of Claude for gain-of-function research on the chikungunya virus.
• [15:48] An X post by Anthropic Alignment Science lead Evan Hubinger stating Coxon is correct that many researchers believe AI could cause extinction (citing their personal estimate at under 10% in the next decade).
• [18:04] A Wikipedia page shown for the film I, Robot.
• [19:38] ChatGPT response list of the most dangerous human inventions in history.
• [23:44] An advertisement showcasing FLB Icon Loafers handcrafted in Portugal.

Claims & numbers

• The presenter notes Jacob Coxon’s resignation post received approximately 165 million views, 776,000 likes, and 200,000 reposts/tweets [00:16].
• Coxon’s follower count grew from near zero to over 263,000 within 48 hours [03:07].
• Tom Ellsworth claims Coxon worked at Anthropic for roughly six weeks and two days prior to resigning [06:27].
• The ChatGPT projection on screen estimates the probability by ~2036 of AI causing human extinction at 2–5%, catastrophic killing of a large fraction of humanity at 3–8%, and severe global disruption without extinction at 15–30% [10:14].
• Evan Hubinger’s tweet states Anthropic researchers believe extinction is a real risk and gives a personal probability estimate of under 10% within the next decade [15:49].
• The AP report displayed claims Anthropic blocked model misuse regarding potential gain-of-function transmission research for the chikungunya virus [15:09].
• The presenters promote FLB Icon Loafers priced at $250 [24:55].

Notable quotes

• “We don’t yet know how to control AI. This is possibly the most dangerous technology that humanity has ever created.” — Jacob Coxon [05:33]
• “They have created the Greta for AI.” — Tom Ellsworth [06:09]
• “Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is <10% within the next decade.” — Evan Hubinger (quoted on screen) [15:48]

Assessment This is a commentary and panel discussion segment from a political and business podcast reviewing a breaking news event and media narrative surrounding AI safety. The presenters show authentic screenshots, social media threads, and a Fox News broadcast clip, mixed with speculative opinion regarding whether the viral event was an orchestrated PR push for AI regulation.

Described by gemini-3.8-flash on 2026-10-12 from the video’s audio and frames.