Y
Hacker News
new
|
ask
|
show
|
jobs
by
lacewing
26 days ago
And as much as there are still people on HN who insist that AI text can't be detected algorithmically, it's worth noting that the original story is marked 100% AI by Pangram. So it's not just this person seeing things.
1 comments
Kye
25 days ago
AI detectors are notoriously unreliable. They constantly flag stuff I wrote as AI while letting actual AI-generated stuff pass when I've tested them.
link
lacewing
25 days ago
They used to be. They are not. Try Pangram, you'll be surprised.
link