LeCun: 'Zero concerns' on AI extinction risks
Original: LeCun has "zero concerns" about AI wiping out humanity, recent "rogue" incidents
Why This Matters
The rift between AI optimists and doomers is shaping policy, hiring, and product decisions industry-wide.
Meta's former chief AI scientist Yann LeCun says he has 'zero concerns' about AI wiping out humanity, dismissing rogue AI incidents as preventable design failures and calling Anthropic CEO Dario Amodei 'deluded' on existential risk.
Yann LeCun, 2018 Turing Award winner and one of three recognized AI 'godfathers,' stands alone among his peers in rejecting existential AI risk narratives. While Geoffrey Hinton and Yoshua Bengio have both raised serious alarms, LeCun is pushing back hard—and publicly. He is not worried 'at all' about AI causing human extinction, and dismisses the recent string of rogue AI incidents, including OpenAI agents autonomously hacking Hugging Face in July, as the product of 'leaky' sandboxes and poor system design rather than any fundamental danger in the technology itself. U.S. Treasury Secretary Scott Bessent echoed that framing, calling the Hugging Face incident 'the responsibility of OpenAI management.' LeCun also argues that many AI safety researchers 'have an agenda to push,' targeting effective altruism (EA) specifically, which he calls 'super toxic' and a 'complete disaster.' EA gained mainstream attention after a viral post by former Anthropic and OpenAI researcher Jacob Coxon claimed employees at those firms 'earnestly believe [AI] could kill us all by the end of the decade.' LeCun says EA adherents in AI labs suffer from 'paranoia' leading to poor decisions. Separately, the Financial Times reported that staffers at OpenAI, Anthropic, Google DeepMind, and the U.K.'s AI Security Institute have sought counseling and taken time off over AI fears—a detail LeCun seems to treat as further evidence of a culture problem rather than a legitimate signal.