AI leaders support independent oversight of development
OpenAI chief executive Sam Altman, Elon Musk and former DeepMind CEO Demis Hassabis have given at least partial support to Anthropic chief executive Dario Amodei’s proposals for slowing artificial intelligence development. The three also agreed that AI laboratories should be subject to independent evaluation and oversight, an approach Amodei has presented as a way to address safety risks as systems become more capable.
Altman told Fortune that OpenAI will not go public this year, citing safety concerns. He had already communicated that decision internally in June, and the company’s planned IPO is being pushed to 2027. The article also notes another possible explanation: OpenAI’s financial position may not currently compare favorably with Anthropic’s. Both considerations could apply.
Google researcher Peyman Milanfar challenged Amodei’s proposed “speed limit” for AI progress. Milanfar argued that recursive self-improvement relies on a simplistic assumption because feedback loops can become unstable. In his view, systems that improve themselves reliably would be constrained by stability, operating slowly and within broad safety margins. He said evidence about their performance should come from the real world rather than benchmarks alone.
Written by Sakhanda Wire AI from reporting by The Decoder. Facts, quotes and full context are in the original article below.