Honest explainers on AI watermarks, detection tools, and writing with AI — no hype, no magic claims.
August 26, 2026 · 3 min read
People use 'paraphrase' and 'rewrite' interchangeably, but they change text at very different depths — and the difference matters for clarity, originality, and how AI-generated text reads.
August 26, 2026 · 3 min read
Anthropic now adds invisible statistical watermarks to Claude's output. Here's how it works, what it can and can't prove, and what it means if you edit text with AI.
August 26, 2026 · 3 min read
The EU AI Act's Article 50 made AI-content marking a legal requirement in August 2026. Here's what the law says, who it applies to, and what it means for writers and publishers.
August 25, 2026 · 5 min read
AI drafts have recognizable tics: even rhythm, safe vocabulary, generic transitions. Seven concrete techniques to turn a flat AI draft into writing that sounds like you.
August 21, 2026 · 5 min read
Perplexity, burstiness, classifiers, and watermarks — what AI detectors measure, why they disagree with each other, and why their scores shouldn't be treated as verdicts.
August 18, 2026 · 5 min read
AI text watermarks are statistical patterns in word choice, not hidden characters you can delete. Here's how they actually work — and what that means if you want to change one.