John Gruber calls Claude's AI watermarking 'patently offensive'
Anthropic is adding a watermark to Claude's output. Tech blogger John Gruber says it could prioritize traceability over the quality of the writing.
Anthropic is introducing watermarks to its LLM outputs in response to EU AI regulations. However, tech blogger John Gruber strongly opposes the approach, calling it a "perversion of writing". Gruber argues that Claude should select words based on precision and quality, rather than to enable detection. He believes that the selection of words itself should not factor into the generation of text, and that any influence other than the user's needs is offensive.
Anthropic's watermark method does not affect the quality of Claude's output, but Gruber contends that even subtle alterations to word selection shape the final result.
Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
