OpenAI, the company that made ChatGPT, has publicly said AI detectors are unreliable.
this post was submitted on 01 Nov 2023
80 points (97.6% liked)
Interesting Shares
1063 readers
3 users here now
Share interesting articles, projects, research, pictures, or videos.
Please include a prefix in your title!
Prefixes for posts
Certain clients offer filters to make prefixes searchable. Photon (m.lemmy.zip) used for hyperlinks below:
founded 1 year ago
MODERATORS
Of course THEY would say that.
/s
Well, they did initially try to make their own detector before giving up because it actually is impossible.
I've had authors I helped edit get accused of being AI, despite their works being written years before the hardware we use to do ai existed.
people are gonna start recording video of themselves writing and then escalate straight to lawsuit territory.
"I have proof that you're lying and that your lies are causing me harm. That's slander. Accusing me of plagiarism is slander. Are you really prepared to lose a court battle over this? Because it's a slam dunk and I have several lawyers in my contacts who are willing to destroy your life for just a cut of the payout for such an easy case."
That's actually a really solid prediction
I know kids today are using version/edit history in schools when being accused of it.