AI-risk researcher: >50% LLM-generated text is a good heuristic for skipping bad papers
sethlazar · x · 2026-09-06
Charlie Bulluck explains why he skips heavily LLM-written articles: attention is scarce, sloppy LLM writing correlates strongly with sloppy research and argumentation, and LLM-written text over 50% is a practical heuristic for weeding out weak papers. Retweeters add that even mild editing makes AI text unrecognizable to Pangram, so if you use AI to write, owe your audience multiple rounds of real editing.
More from AGI Musings
- Scale puts AI capability and cost in a U-shape: GPT is absurdly cheap, Codex an unreal deal — chris_j_paxton · 2026-09-06
- Swarm alignment debate: will open agent communication channels stop hidden coordination — soumitrashukla9 · 2026-09-06
- Slow Clinical Trials Break AI's Feedback Loop in Drug Development — clarejtbirch · 2026-09-06
- Timnit Gebru shares a 1963 quote on mechanical brains and human extinction — MilagrosMiceli · 2026-09-06
- Student questions for Mitchell: can bounded autonomy know when humans must step in? — ruthstarkman · 2026-09-06
- Even with superintelligence, some will still want fellow-human output — burny_tech · 2026-09-06