Citation·page:leopold-aschenbrenner:fn33

Leopold Aschenbrenner - Footnote 33

Verdictpartial85%

1 check · 4/3/2026

The claim states a 100,000x increase in effective compute between 2024 and 2027, but the source projects a ~3-6 OOMs increase, which translates to a 1,000x to 1,000,000x increase. The claim mentions 2025-26 for AI surpassing college graduates, but the source does not specify this timeframe. The claim mentions superintelligence emerging by the end of the decade, but the source suggests it could emerge ~1 year after AGI, which is projected around 2027.

Our claim

entire record

No record data available.

Source evidence

1 src · 1 check

forum.effectivealtruism.org/posts/zmRTWsYZ4ifQKrX26/summary-of-situational-awareness-the-decade-ahead resource

partial85%Haiku 4.5 · 4/3/2026

NoteThe claim states a 100,000x increase in effective compute between 2024 and 2027, but the source projects a ~3-6 OOMs increase, which translates to a 1,000x to 1,000,000x increase. The claim mentions 2025-26 for AI surpassing college graduates, but the source does not specify this timeframe. The claim mentions superintelligence emerging by the end of the decade, but the source suggests it could emerge ~1 year after AGI, which is projected around 2027.

Case № page:leopold-aschenbrenner:fn33Filed 4/3/2026Confidence 85%