A year after the 100% AI mandate nobody understands our code

Re:A year after the 100% AI mandate nobody understands our code by amarcheschi on Monday November 09, 2026 @08:29PM (#50434926)

Honest question, some time ago people were claiming chatgpt zero or other ai detectors were not good, now everybody uses pangram. Is it that reliable?

[ Parent | on HN ]

Re:A year after the 100% AI mandate nobody understands our code by ianm218 on Monday November 09, 2026 @09:00PM (#50435132)

From my vibes Pangram is very good. This [1] seemingly independent analysis is pretty positive on them and says there is no easy way to just beat it.

Presumably at some point there will be AI trained specifically to beat it but I think it’s fair if something sets off your internal AI detector (like me on this post) and pangram says it’s 100% AI, it’s probably AI.

[1] https://www.layer3labs.io/guides/is-pangram-accurate

[ Parent | on HN ]

Re:A year after the 100% AI mandate nobody understands our code by embedding-shape on Tuesday November 10, 2026 @06:54AM (#50439034)

[delayed]

[ Parent | on HN ]

Re:A year after the 100% AI mandate nobody understands our code by fxwin on Tuesday November 10, 2026 @06:36PM (#50447858)

> I have public writings since before LLMs that is corrected "graded", private writings since before LLMs that are mostly incorrect

a tale as old as time

[ Parent | on HN ]