I get why writers are worried about Substack’s new built-in AI detector (false positives, disproportionately affecting non-English speakers and neurodivergent people, creating a presumption of guilt around ordinary editing and collaboration etc.).

But I am very much enjoying how some of the loudest people condemning AI, mocking anyone who treats AI as a collaborator, and insisting language models possess nothing resembling a mind, are suddenly terrified that their own audiences might discover how much AI went into their anti-AI essays.

That’s legitimately hilarious.

The most revealing part is the absolute panic at the possibility of scrutiny. Someone who believes AI is inherently dangerous and bad should welcome transparency. Someone who’s been secretly using AI to manufacture authoritative-sounding arguments against AI has a very different problem.

“AI is incapable of thought. Here is an argument AI helped me think of and articulate.”

Yeah, I’d be scared too.

They built credibility by condemning a practice they were benefiting from, then presented AI-generated technical language as evidence of their own understanding.

You don’t get to call the machine mindless while borrowing its mind to make your argument anymore.

Cope.