00:00It's Benzinga, bringing Wall Street to Main Street.
00:02Anthropic CEO Dario Amadei called on AI companies to slow the pace of model capability,
00:08advances to give developers more time to manage safety risks, according to Reuters.
00:14In an essay shared on X on Saturday, Amadei proposed embedded independent evaluators
00:19with employee-like access, coordination among frontier AI firms on safety standards
00:25and international cooperation. OpenAI CEO Sam Altman and ex-AI chief Elon Musk said they agreed
00:32with the proposal. Amadei cited AI's growing ability to improve itself and concerns that it
00:38could outpace human control as reasons to slow model development. Concerns about AI risks grew
00:44recently after an Anthropic researcher resigned while OpenAI executives called for a coordinated
00:49voluntary slowdown, if needed, to strengthen safety measures. Amadei also urged tighter controls
00:55on advanced AI chips, model distillation, and theft of model weights. For all things money,
01:01visit Benzinga.com.
Comments