AI-shaped writing isn't AI slop: detectors will catch the former, the latter stays subjective
cedcolas · x · 2026-09-25
A debate with @joelbot3000 on detecting AI slop: the author argues AI-shaped writing ≠ AI slop. Detectors can be built for the former, but the latter is often subjective and unfixable — he uses AI to write papers and doesn't consider the result slop. joelbot3000 counters that actual slop (profound-sounding but trivial, disconnected from prior research) is classifiable. Both agree AI assistance in research writing is fine.
More from AGI Musings
- Coding agent form factor is saturated: time for persistent, long-term AI collaborators — repligate · 2026-09-26
- Rob LeClerc: Agents Will Trigger an Explosion in AI Token Demand — robleclerc · 2026-09-26
- Ex-Meta Researcher: Real-Time Interactive Videogen Fuses LLMs and World Models — alex_conneau · 2026-09-26
- Researcher Slams Anthropomorphizing LLMs: It Obscures Corporate Accountability — AlexTensor · 2026-09-26
- Bostrom: slowing down AI could cause superintelligence to arrive suddenly via a "compute overhang" — i_dg23 · 2026-09-26
- François Fleuret: autoregressive process and reasoning are two different conceptual levels — francoisfleuret · 2026-09-26