Anthropic CEO warns AI could kill us all by decade's end

Dario Amodei, CEO of AI firm Anthropic, has raised serious concerns about artificial intelligence, warning it could kill us all by the end of this decade. His post highlights risks such as loss of control over AI systems and misuse in cyberattacks and bioterrorism, prompting calls from leading figures to slow AI progress.

Amodei flagged “serious” risks in AI, including the potential for losing control of AI systems, misuse for cyberattacks and bioterrorism, and significant economic disruption. He emphasized the need to pace AI development carefully to allow risk prevention measures to keep up with advancing capabilities.

He believes that if slowing down AI progress could buy even an extra year or two before models reach critical levels, it would provide valuable time to improve alignment and reduce the risk of serious failures. Amodei advocates building AI at a balanced rate that ensures safety while still achieving benefits and addressing geopolitical challenges.

Elon Musk and Sam Altman have expressed agreement with Amodei’s call to slow AI development. The debate centers on caution over speed and prudence over profit, with concerns about authoritarian governments and the difficulty of verifying compliance with safety measures. Experts warn that building AI too fast is reckless and stress the importance of not just investing in risk prevention but pacing the rate of capability advancement.

Amodei’s warnings come amid growing global debate on the future of AI and its potential to transform society, underscoring the urgent need for careful management of this powerful technology.