Debate: Has LLM novelty in science disproven the 'no recursive self-improvement' argument?
ctjlewis · x · 2026-09-16
A quote-debate on whether LLMs can evolve to superintelligence via recursive self-improvement. @gummibear737 argues it's impossible: it would require inventing technology unknown to man, while LLMs merely predict what humans think/say/do — so Skynet won't come from LLMs. Reposter ctjlewis counters that this argument is already in the past: its original proof was that LLMs had done no novel science or mathematics, and they now have.
More from AGI Musings
- Why recursive self-improvement to superintelligence likely won't happen — tawnniee · 2026-09-16
- Gradio founder: let users read full unencrypted reasoning traces to align LLMs — mmitchell_ai · 2026-09-16
- A 1797 Goethe poem explains why AI researchers fear recursive self-improving systems — OmarUFlorez · 2026-09-16
- Everyone using AI at work risks becoming a useless middle manager — jamesluke585 · 2026-09-16
- Researcher pushes back on AI biorisk alarmism: threat vectors haven't changed since o1 — anshulkundaje · 2026-09-16
- Maybe human intelligence is near the ceiling of stability evolution allows — games-and-games · 2026-09-16