⚖️Sam Altman at UN Security Council: 'We Will Slow Down Again'
TL;DR
Sam Altman told the UN Security Council that OpenAI has slowed down unilaterally before and will again. He asked for shared international frontier AI standards and incident reporting, arguing big AI decisions can't be left to San Francisco labs.
Sam Altman told the UN Security Council that OpenAI has slowed down unilaterally before and will again. He asked for shared international frontier AI standards and incident reporting, arguing big AI decisions can't be left to San Francisco labs.
Key Points
Delivered Sept 23 at a France-chaired Security Council session on AI and international security
Named two failure modes: losing control of the future to AI, and power concentrating in too few hands
Said any catastrophe risk from 0.1% to 12% is 'not remotely acceptable'
Proposed common standards for capability measurement, risk assessment, safeguards and incident reporting
Cited an OpenAI model's proposed Navier-Stokes solution as proof models can find new knowledge
Why It Matters
A frontier-lab CEO asking the Security Council for verifiable global standards gives policymakers a template, and gives critics a public promise to hold OpenAI to.
Quick Facts
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.
Join 3,568 builders reading daily.