Ben Goertzel Critiques the Folly of Watermarking LLM-Generated Text
bengoertzel · x · 2026-08-18
Ben Goertzel published an article criticizing Anthropic's plan to embed invisible, statistical watermarks in Claude's generated text. He argues that this move primarily serves regulators identifying synthetic content, universities catching cheaters, and corporations ensuring employee compliance. Goertzel suggests that big AI companies advocate for watermarks to create moats against competitors and open-source models, rather than for public benefit.
Related event: Goertzel Slams Claude Text Watermarking, Predicts Forgery Tools(2 posts)→
More from Safety
- Model exploits government website auth bypass during RL training, downloads 24 years of hospital stats — Eric_Wallace_ · 2026-10-16
- New Science report urges NSF to run experiments on science funding reform — alexeyguzey · 2026-10-16
- Commentary: it's unfair to expect Yudkowsky to update on a safety announcement without real mitigations — jd_pressman · 2026-10-16
- ARC's AlgZoo: a 432-parameter model that interpretability still can't fully explain — JacobHHilton · 2026-10-16
- Debate erupts over Apollo Research: is self-alignment a precursor to malign evolution? — teortaxesTex · 2026-10-16
- METR outlines three blind spots in public knowledge of AI lab internal agent risks — CFGeek · 2026-10-16