David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models
Points and comments are a snapshot, not live.
David Sacks argues frontier AI labs don't need regulators, just responsibility.
David Sacks responds to Dario Amodei and Sam Altman's calls to 'pace the frontier' by telling OpenAI and Anthropic they are the frontier and can slow down voluntarily without government permission. He argues their appeals for regulation amount to regulatory capture, cartel formation, and blackmail, and that product liability and market forces already constrain them. Sacks notes competitors aren't at the frontier, METR is not independent, and China won't join a global agreement, so these labs should simply stop building superintelligence if they are genuinely worried.
What commenters are saying
The thread is split sharply. One camp argues stalling is about financial and technical limits, not safety: soaring compute costs, electricity bottlenecks, and saturated markets make voluntary slowdown a business necessity disguised as altruism. Others counter that performance has demonstrably improved (solving a Millennium problem, public Hugging Face incidents) and that slowing collusively without government waivers is clearly an antitrust issue. Sacks' past support for government actions against Anthropic is cited to question his sincerity. A smaller group agrees with Sacks that labs don't need permission to pause.