I thought that they were trying to get at the hypocrisy of using ai tools to detect ai writing, which has a dismal success rate and ends up with a crazy amount of false positives.

I see that the growth of computer/digital testing directly led to student abuse of ai. Once students were no longer called on to write a blue book or even just generate responses to short answer questions, we failed our students. I understand that term papers and larger writing assignments have their place, but there really is nothing like being able to sit down and show your knowledge in an on the spot essay (at least for a lot of humanities courses where I see ai being abused).

A systemic issue in this space comes from universities and licensing entities giving testing companies carte blanche to race to the bottom in terms of effective testing tools. For licensing, apprenticeship/mentorship and observation are always better indicators of on the job performance, but that doesn’t have the ease and profit margins of big bullshit testing, so companies would never go for it unless states mandated such pathways. With regard to a lot of current digital university testing, a lot of that was Frankensteined from existing licensing exam tools/strategies and similarly does not effectively assess student/candidate aptitude.

Even if this was not the intent of the commentor, my argument is that carelessness and disingenuous/gatekeeping testing practices from universities and licensing institutions certainly bred a significant level of disengagement in students which leads to them phoning it in with ai slop. If we wanted rigorous work from our students, we should have kept the same standards for ourselves, collectively.