Why debate-persuasion scales with intelligence but other AI tactics don't
voooooogel · x · 2026-10-01
- voooooogel argues a superintelligence would handily beat him in a debate where he arrived open to changing his beliefs.
- However, AIs aren't really prone to using (or in a position to use) other tactics, and those tactics don't scale with intelligence the way debate-persuasion does.
- The thread weighs superintelligence persuasion and deception risk.
Related event: Debate Erupts Over Superintelligent Persuasion Risks(3 posts)→
More from AGI Musings
- Mathematicians must rethink their role as AI gets better at proving theorems — notmisha · 2026-10-01
- CogGym preprint compares AI and human judgments across 258 cognitive-science experiments — burny_tech · 2026-10-01
- Deedy: Trust Pricing, Not Benchmarks — High Price Means a Genuinely Strong Frontier Model — deedydas · 2026-10-01
- South Korea launches 18 AI-focused universities, makes AI literacy mandatory for undergrads — lmoroney · 2026-10-01
- Ramp CPO sees 3 long-term paths for PMs: technical, tastemaker, and GM — lennysan · 2026-10-01
- AI speeds up research so much you can answer an arXiv question the same day — KyleCranmer · 2026-10-01