Hacker News Reader

prometheus1992

easy to trick pangram.

any evidence for that? genuinely interested. all the evidence I've seen is that Pangram has a very low false negative rate.

https://github.com/deepanwadhwa/ai_detector_fails/blob/main/...

this is inane, the "modified" gemini response is totally rewritten from the original (which pangram caught) in a completely different style. this is evidence of nothing.

pangram is supposed to detect AI writing, gemini is AI!!

Reply on news.ycombinator.com