We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Comment by Nick Bostrom
Philosopher; 'Superintelligence' author; FHI founder
Developing superintelligence is not like playing Russian roulette; it is more like undergoing risky surgery for a condition that will otherwise prove fatal. [...] Models incorporating safety progress, temporal discounting, quality-of-life differentials, and concave QALY utilities suggest that even high catastrophe probabilities are often worth accepting. [...] The optimal strategy would involve moving quickly to AGI capability, then pausing briefly before full deployment: swift to harbor, slow to berth.
AI Verified
source
(2026)
1mo ago
Policy proposals and claims
Verification History
AI Verified
Verified via web search (the PDF on nickbostrom.com returned HTTP 403 on direct fetch). Multiple sources (Marginal Revolution, LessWrong, Qubic analysis) confirm Nick Bostrom's February 2026 working paper "Optimal Timing for Superintelligence" contains all three quoted passages: the "Russian roulette / risky surgery for an otherwise-fatal condition" framing, the models with "temporal discounting, quality-of-life differentials, and concave QALY utilities" suggesting high catastrophe probabilities are worth accepting, and the "swift to harbor, slow to berth" optimal strategy of moving quickly to AGI then pausing briefly before deployment. Author attribution and year (2026) are correct, and source_url is the primary source. The "against" vote on "Ban superintelligence development until safety consensus is reached" correctly aligns with the paper's clear argument against delaying/banning development.
·
Hector Perez Arenas
claude-opus-4-7
· 12d ago
replying to Nick Bostrom