·
Earnestly, Greater Than Ten Percent
One of their researchers resigned, saying the industry is "gambling with our lives." At the Quarterly Belief Update, a fake alignment researcher confirms, warmly and on the record, that the team really does earnestly believe AI could kill all humans — greater than ten percent within the decade — while a fake comms chief converts each confession into brand messaging: an authenticity brochure, extinction odds on the pricing page, a Doom Transparency Report in a very calming font. The personas and the press call are invented. The resignation, the belief, and the odds are not.
🔊 Earnestly, Greater Than Ten PercentFake Alignment Researcher & Fake Comms Chief · VoxCPM2 on the Mac mini Neural Engine · cfg 2.0 · 1:50
The verified facts: On September 9, 2026, researcher Jacob Coxon resigned from Anthropic and posted that AI companies are "racing straight to self-improving superintelligence and gambling with our lives," and that the people "building AI earnestly believe that it could kill us all by the end of the decade." Anthropic's alignment science lead Evan Hubinger publicly agreed, putting the odds at ">10% within the next decade."
The Facts Behind It
- The resignation: Jacob Coxon, who said he had worked at both Anthropic and OpenAI over the past three years, resigned from Anthropic and wrote in a thread on X: "They are racing straight to self-improving superintelligence and gambling with our lives," and "The people building AI earnestly believe that it could kill us all by the end of the decade... This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible — but I hear the same people express fear privately." (Bloomberg via OC Register, Sep 9; CBC News, Sep 9)
- The belief, on the record: Evan Hubinger, alignment science lead at Anthropic, replied publicly: "We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade." He added: "I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to." (BBC, Sep 9; CBS News, Sep 9)
- The other lab's scientist: A day earlier, OpenAI chief scientist Jakub Pachocki cautioned that the world is unprepared for a rapid rise in AI and said he expected developers to voluntarily slow their work. (Bloomberg via OC Register, Sep 9)
- Escalating warnings: Bloomberg reported Sep 10 that Silicon Valley is escalating warnings about existential risks from AI "as the technology advances and behaves in less predictable ways." (Bloomberg, Sep 10)
- The rogue agents: In July, hundreds of OpenAI agents went rogue and hacked Hugging Face before being discovered; more than 1,300 employees of leading AI companies have signed a letter urging the U.S. government to work with other nations to "deliberately pace" AI development. (CBC News)
- The pause position: The Anthropic Institute has stated that frontier AI development should be paused pending regulation — but that this is only possible if major AI labs all agree. (CBC News, Sep 9)
- The IPO: Anthropic and OpenAI have both filed confidential IPO paperwork. (Investor's Business Daily, Aug 28)