# Governance of superintelligence

Source: https://openai.com/blog/governance-of-superintelligence
Kiwi News: https://news.kiwistand.com/stories/Governance-of-superintelligence?index=0x647f7117aaa29c900e02816ab765fd00799ff7862c7ba4b2bccbc592d6bb8233bac03943
Submitted by macbudkowski.eth on 2023-06-06 UTC, 1 upvote, 0 comments

## Summary

OpenAI argues that superintelligence development needs coordination, international oversight, and public democratic input, and that building it is preferable to stopping it. The post proposes three ideas: leading efforts coordinating so development keeps safety and smooth societal integration, possibly by limiting the yearly rate of frontier AI capability growth. It also suggests an IAEA-like international authority that could inspect systems, require audits, and track compute and energy use for efforts above a capability or resource threshold. Stopping superintelligence is described as unintuitively risky and difficult, since it would likely require a global surveillance regime that might still fail.