Sometime before 2031, AI is going to cross a generality threshold that allows for recursive self-improvement, and then FOOM to superhuman capability. Before that point humanity must solve AI safety or be helpless.
GeopolitiekAIOne-Off3j
Manifold MarketsGeen KYCGeverifieerde afwikkelingsgegevensRedelijk gekalibreerdHuidige gemeenschapsvoorspellingNee 67.4%
Voorspellers
21
Vraagtype
binary
Methodologie
Play-money forecasting platform
Brontype
Voorspelling
Marktdata
Bijgewerkt 13 uur geleden
Verouderd24 dec 22, 03:53 UTC1 feb 30, 07:59 UTC
Trends
Gerelateerde Markten
Gerelateerde vragen voor context, geen vergelijking van gelijkwaardige contracten.
OpenAI vs. Anthropic: First to announce recursive self-improvement?
PolymarketWill OpenAI or Anthropic announce recursive self-improvement?
KalshiOpenAI or Anthropic announces recursive self-improvement by which date?
PolymarketIf GPT-5 can do recursive self-improvement, will it first be via fine-tuning on its outputs?
Manifold MarketsWhen will an AI system complete 3 autonomous self-improvement cycles with 5%+ capability gains each?
Manifold MarketsWhen will an OpenAI model achieve a High risk level on AI Self-improvement? [metaculus]
Manifold MarketsRegels
Following on this useful input from Steven, I shall clarify my resolution criteria.
- Steven: Does this resolve NO if AI recursive self-improvement results from a gradual increase in generality?
- Nathan: I say 'allows for' recursive self-improvement (FOOM-after-deployment).
- It might be superhuman right out of the box (FOOM-during-training).
- I will resolve YES if it becomes strongly superhuman by any means at any speed, and humanity manages to solve AI safety via alignment or governance sufficiently to prevent catastrophic conflict.
- I will also resolve YES if there is a conflict and the AI is clearly winning or has won (if I'm still alive to do so).