OpenAI, Meta, DeepMind Staff Meet Coxon's Doom Warning With 'Lol'

Not every AI worker thinks the technology is on track to kill everyone. The BBC reports that reactions from current and former staff at OpenAI, Meta and DeepMind to Jacob Coxon's viral existential-risk warnings have included 'Lol', 'Haaaaaa' and 'Bringing the luls'.

Coxon, a former Anthropic employee, said a group of AI agents, based on AI models that do not currently exist, could decide to create and then aim a biological weapon, without detailing how. His claims went viral and were echoed by others urging a slowdown in development.

Rishub Jain, who spent seven years at DeepMind before founding the AI safety research firm Sampura Research, told the BBC the current tone among many AI workers about the fresh fears had 'definitely been a little jokey'. Colin Fraser, a data scientist at Meta, wrote on social media there was 'no real evidence that AI models would inevitably pursue a goal leading to human death', adding that 'LLMs won't wipe out humanity because they just don't have that dog in them.' A former OpenAI employee told the broadcaster the extinction case is 'always vague' and depends on 'major jumps in reasoning.'

The pushback lands a day after our tracker logged OpenAI conceding no path to safe full recursive self-improvement, the same argument seen from the inside. The BBC notes broader agreement inside the labs on nearer-term harms: guardrail failures, military applications and hacking vulnerabilities.