Dan Hendrycks, who runs the Center for AI Safety, called for an indefinite pause on the biggest frontier AI training runs at The New York Times' DealBook Summit in New York on Nov. 18, TNW reported.[1] He would exempt narrow models such as DeepMind's AlphaFold and let labs keep selling inference, and said the companies had blocked narrower technical rules for years.[1] Max Hodak, a Neuralink co-founder who now runs Science Corporation, said on the same panel that enforcing a ban would mean "an exhaustive surveillance state on the internet," and backed antitrust law instead.[1] The exchange matters because Hendrycks said politicians can now discuss ideas such as pausing AI development openly in Washington.[1]
AI safety group head calls for indefinite pause on biggest AI training runs
Dan Hendrycks would exempt narrow models such as DeepMind's AlphaFold, while Max Hodak, who now runs Science Corporation, backed antitrust law instead.
0 primary · 1 independent · 0 commentary · 0 vendorSources
- Dan Hendrycks calls for a pause on frontier AI, Max Hodak pushes back TNW | Agi Independent