Stack Overflow (and Stack Exchange) mods no longer allowed to use GPT detectors

meta.stackexchange.com

12 points·sh_tomer··10 comments

Crowd sourced, let the humans vote up and explain the correct answers. Theres too many incorrect answers and unanswered technical questions already.

They should flag the AI-generated content as “inaccurate” or “too verbose”, assuming it actually is of course

GPT detectors are inaccurate, full-stop. Although “strict standard of evidence” is going to rule out most real AI-generated content, because there’s not many objective measures.

HAHAHA

> Finally, internal evidence strongly suggests that the overapplication of suspensions for AI-generated content may be turning away a large number of legitimate contributors to the site.

Someone invent a time machine so I can go back to October 2022 and read that then.

And likewise, the presence of excessive AI-generated content also turn away a lot of people.

The day HN stops banning ai generated posts and comments is the day I leave forever.

...You've been here for 9 days though?

What makes you think guraf uses a single persistent account? Many people only stick with an account for a short period of time before abandoning it for a new account.

Why do they do that?

Little bits of personal info add up over time, and suddenly you're a lot less anoymous than you think you are.

I see, that makes a certain amount of sense, thanks for the answer.

https://xkcd.com/810/

Reply on news.ycombinator.com