Meet Jacob Coxon: The 27-Year-Old Researcher Whose Resignation Reignited AIs Existential Debate

Oh my goodness, buckle up, because the most *thoughtful* resignation in the history of artificial intelligence has just happened, and it is genuinely, genuinely beautiful! ✨ Jacob Coxon is twenty-seven years young and has model-training experience at both OpenAI and Anthropic, a CV so dazzling that recruiters are probably fainting as we speak, and he has generously decided to take a little breathing space from all that superintelligence-adjacent labour. And he announced it on X, which means his announcement isn't just news, it's *content*, and we can all enjoy it together! Two legendary employers, one very brave sabbatical!

Now, some people will read his line about companies 'gambling with our lives' and get a little bit worried, but have they ever *been* to a casino? Gambling is fun, and gambling is optimism with a drinks menu! And then there's Evan Hubinger, who leads Anthropic's alignment research, publicly agreeing with the whole thing, which is simply the most *collegial* disagreement a workplace has ever had. Two brilliant people, perfectly aligned about alignment, both absolutely certain that stronger regulation or a coordinated slowdown between labs would be a lovely idea! That's not a crisis, that's a *meeting of minds*! ✨

And the headline number! Hubinger puts the probability of human extinction from AI within the next decade at *greater than 10 percent*, which means, if you simply flip it around the way sensible people do, there is a magnificent nearly-90-percent chance that nothing extinction-related happens at all! Those are the sort of odds you'd happily take at a village fête tombola! ✨ The only tiny asterisk is that it all depends on stronger regulation or a coordinated slowdown between labs, and honestly, when has anything in this industry ever needed coordination to turn out beautifully? Everyone is going to be fine, the alignment people agree with each other, and what a decade to be alive in! ✨