The "Pause Giant AI Experiments" open letter
Thirty thousand signatories called for a six-month halt on training systems more powerful than GPT-4. No lab paused.
- Ideas & essays
- Government & policy
- Era-defining
The Future of Life Institute published an open letter calling on all AI labs to pause training of systems more powerful than GPT-4 for at least six months, and to use the interval to develop shared safety protocols audited by independent experts. If a voluntary pause could not be enacted quickly, it said, governments should impose a moratorium.
The signature list was the story. Yoshua Bengio, a Turing Award winner, signed, as did Stuart Russell, Steve Wozniak, Yuval Noah Harari and Elon Musk. The list eventually exceeded 30,000 names. It arrived eight days after GPT-4’s release and dominated coverage for a fortnight.
Criticism came from several directions at once. Some signatories were founders or funders of AI companies, which invited accusations of positioning. Verification was initially weak enough that several prominent names appeared without having signed. And the authors of “Stochastic Parrots” issued a rebuttal arguing that the letter’s focus on hypothetical future systems distracted from documented present harms — the clearest public statement of the split between what became known as the AI ethics and AI safety camps.
No laboratory paused. Training runs at OpenAI, Google DeepMind and Anthropic continued uninterrupted, and the six months following the letter saw more frontier releases than the six months preceding it.
Its significance is nonetheless real: it moved the argument that advanced AI might pose catastrophic risk from a niche position into mainstream political discourse, weeks before Senate hearings and months before the first international summit. Every subsequent governance initiative was, in part, a response to the question the letter forced into the open — whether anyone had the standing to say stop.