Anthropic's Claude text watermark deliberately distorts word choice, Gruber calls it a perversion of writing

Anthropic's 'Watermark' Text Adulteration in Claude Is a Perversion of Writing

John Gruber breaks down Anthropic's watermark scheme: at each token generation step, Claude biases word choice toward a 'green list' and away from a 'red list', embedding a statistically detectable fingerprint. This directly contradicts Anthropic's original claim that the watermark is 'imperceptible' and 'doesn't change meaning, quality, or readability'—it deliberately degrades natural word choice for traceability. The piece recommends James Padolsey's interactive explainer and notes that longer texts yield higher detection confidence, while short texts can't be reliably flagged. Gruber calls this text adulteration, not a feature a writing tool should have.

Why it matters: Gruber's critique of Anthropic's text watermark includes concrete mechanism breakdown, not just vague complaints. Hits all three HKR axes, but as commentary rather than a first-party product release, it lands in the 78-84 band per policy.