“If we write properly, we get accused of being AI — it’s absolutely ridiculous. Long term, I think it’s going to be a big problem.” — Aldan Creo, grad student studying AI detection at UC San Diego.

So we’ve built detectors that penalize writing well. Students are paying $20/month to make their own work sound worse just to pass a broken test. Non-native English speakers are getting flagged at higher rates. One student said she rewrites everything until the detector approves it — not to cheat, but to survive.

At what point does “proving you’re human” become the assignment? Link