Skip to content

Comment on Professor's invisible prompt trap catches 32/35 students cheating with AI

Comments

This particular implementation of the test seems flawed. The “invisible text” which would be read by a screen reader or visible if the problem statement were copied to some persistent scratch space while drafting a response would specify the inclusion of the word whether or not the author is an LLM.

Yes because people are identical to llms and cannot see through that.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.