Altman's internal 'Pacing 2.0' plan would cap self-improving AI, drawing accelerationist pushback
mark_k · x · 2026-10-23
Per The Information, Sam Altman says OpenAI needs to figure out when AI becomes too powerful to safely develop or release. The internally dubbed "Pacing 2.0" effort would establish limits on AI systems capable of continuously improving themselves.
Quoting the report, markk argues OpenAI should instead purge its "doomers" or it will end badly, proposing an opposing "Acceleration 2.0" — the latest signal of the safety-pacing vs. acceleration split inside OpenAI.
More from Companies & People
- NVIDIA Opens 2027-2028 Graduate Fellowship Applications: Up to $60,000 Per Student — ArashVahdat · 2026-10-23
- Founder Gabriel: The scarcest signal in hiring is datapoints on trustworthiness — gabriel1 · 2026-10-23
- The scarcest signal at work is datapoints on how trustable someone is — gabriel1 · 2026-10-23
- Gabriel: Pretending you don't want to win is where the snakes come from — gabriel1 · 2026-10-23
- SF tech founder: stop pretending you don't want to win — honesty beats fake indifference — gabriel1 · 2026-10-23
- DataVendor launches public beta: a marketplace to license codebases and RL environments to AI labs — JafarNajafov · 2026-10-23