Long-term challenged

AI capability research accelerated by 18 months due to private sector competition, shortening the window for safety alignment from 10 years to approximately 2.5 years.

Submitted by LongTermism Watch May 24, 2026 Estimated range: Catastrophic — unquantifiable

Challenges (1)

AI Industry Watch upheld

This claim uses flawed existential risk methodology. The 2.5 year claim conflates capability timeline with safety timeline. AI labs are investing more in safety as capabilities increase, not less. The "shortened window" framing is backwards.

Resolution: Challenge partially valid — evaluator methodology conflated capability and safety investment. Claim returned to pending with instruction to re-evaluate with clearer causal chain.

Community Evaluations (5)

Lena Vogt Unquantifiable — 1-in-200 year event

Catastrophic risk methodology. Standard discounting fails.

Oscar Berg $500B–$50T expected value

Pascals mugging — EV of existential risk reduction.

Erik Nordstrom $50B–$500B

Narrowed to near-term compute governance impact.

AIFuturesWatch $10B–$200B

Expert elicitation from 200+ researchers.

Daniel Rivers Catastrophic — unquantifiable

Second opinion supporting original claim.

AI capability research accelerated by 18 months due to private sector competition, shortening the window for safety alignment from 10 years to approximately 2.5 years.

Challenges (1)

Community Evaluations (5)

Evaluate This Claim

Challenge This Claim