Vincent Conitzer's IEEE piece explores wildly differing views on recursive self-improvement
conitzer · x · 2026-11-12
AI researcher Vincent Conitzer published a piece in IEEE Intelligent Systems on recursive self-improvement, showing how researchers hold very different models of what it could cause. Its most popular section so far uses an analogy: asking your dog whether you should move to Japan or Turkey — feedback from a system that can't grasp the decision.
Related event: Conitzer Paper Asks: Is Recursive Self-Improvement Safe?(3 posts)→
More from AGI Musings
- AI's low-trust prisoner's dilemma: Q3 review plus Cognition CEO Scott Wu interview — htaneja · 2026-11-12
- millionways launches Thorsten-4, a 'Large Psychology Model' claiming 90% validated accuracy — Scobleizer · 2026-11-12
- The rarer the hypothetical, the more broken the conditional prediction market — lxrjl · 2026-11-12
- Remote Labor Index jumps from 3.8% to 41.3% in a year, doubling every ~3.3 months — Puzzleheaded-King584 · 2026-11-12
- Conditional prediction markets price consequences well but can't tell decision-makers what to do — lxrjl · 2026-11-12
- Anthropic IPO letter: we may slow AI development even at competitive disadvantage — haider1 · 2026-11-12