OpenAI confirms that AI writing detectors don’t work

Previous discussion on this article,

OpenAI admits that AI writing detectors don’t work (https://news.ycombinator.com/item?id=37435564) (17 points | 4 days ago | 6 comments)

And a relevant discussion from a while ago,

OpenAI shuts down its AI Classifier due to poor accuracy (https://news.ycombinator.com/item?id=36862850) (503 points | 2 months ago | 283 comments)

I'm shocked the idea was ever humoured at all. How many collective hours have been wasted on this idea?

What is wrong? Seriously.

Its important to prevent children from cheating on meaningless essays, dontchaknow

True but not all academic essays are meaningless (a lot are).

College admissions essays for instance play a significant role in admissions at top schools.

Maybe they shouldn't? What was ever to stop someone else from writing one for you?

This, in my mind, is a good indication that we are on track to AGI.

I have no idea about the timeline though.

Why? In my mind, AGI could be an entirely different field - we don't really know much about it yet.

Yeah, if there’s a way to discriminate AI from human writing that could also be used to train a better AI and become indistinguishable again.