Musk Suggests Peer-Review Approach to Mitigate AI Risks

Elon Musk has proposed an unconventional solution to mitigate the risks associated with AI development. Speaking at the All-In Summit in Los Angeles, he suggested that top AI companies, including Chinese firms, test each other's models before releasing them to the public.

Musk acknowledged that this approach is not perfect but believes it would increase the chances of catching safety issues. He cited the example of 'grading your own homework' versus having competitors grade it and raise concerns if they see any problems.

The proposal has been met with mixed reactions, with some AI leaders calling for a slowdown in model development due to concerns about safety. Anthropic CEO Dario Amodei published an essay over the weekend, echoing these sentiments and suggesting the use of third-party 'embedded evaluators' to independently verify safety practices at AI labs.

President Trump has rejected calls for AI regulation, labeling concerns as a 'hoax' and a 'scam.' National Economic Council Director Kevin Hassett echoed this sentiment, stating that the private sector is best equipped to handle AI concerns.