Anthropic CEO calls for slowing down AI race; Altman & Musk back him up

Anthropic CEO Dario Amodei has called for a slowdown in the rapid advancement of AI technology, warning that unchecked progress could render the systems too complex for humans to comprehend and manage. Writing on his personal website, Amodei emphasized the need to "slow the pace at which we improve the capabilities of AI models," explaining that while AI's benefits are significant, the risks are equally serious.

The proposal has garnered support from OpenAI CEO Sam Altman and Tesla's Elon Musk. Altman echoed Amodei's sentiments, stating that the need to pace frontier AI development had been a "key topic of discussions" at OpenAI. Musk also endorsed the call, responding with a simple "Dario is right."

Amodei highlighted the acceleration of AI development, particularly the phenomenon of "recursive self-improvement," where AI systems increasingly help build the next generation of AI. He argued that without proper safeguards, AI could become uncontrollable and pose severe risks, including loss of control, misuse for cyberattacks and bioterrorism, and significant economic disruption.

To mitigate these risks, Amodei proposed that each AI company grant independent third-party evaluators ongoing, employee-like access to their systems for safety verification and incident reporting.

Despite the call for a slower pace, Amodei maintains a positive outlook on AI's potential benefits, including major advancements in curing diseases and economic growth. However, he stressed that achieving these benefits necessitates careful development with robust safety measures.