Maybe I should teach computer science or mathematics again. I might have fun with the AI mania.
After a year or so, I let my students use class notes and books on tests. Guess how my grading curve wound up, compared with the same class with closed book and no notes?
There was no meaningful difference between the two.
Of course, it is always possible that I was becoming a more competent instructor...
In science and engineering you are not expected to keep everything in your head as a professional, but you are expected to know where to look. It's fine there, afaiac.
In grad school, a professor sprung a tough in-class test on us. I took another class from him. There, he asks if I want to know what my grade was on that test. I said, "Not really." He asked, "How about if I told you that you got the best score in the class?" I said, "Well, okay then!"
"You got a B. The whole point of that exam was to show students that you can't find the answers to the big problems in 40 minutes. All further tests will be take-home."
It had 5 whole questions.
From then on, I piled so many math books around me to try to solve the problems or make the proofs, and exactly ZERO of them hand any answers.
Turns out, the prof made up his exams from newly discovered mathematics written up in foreign language journals.
I got an A in that class, where we all solved longstanding problems in math on our exams (recently solved not by us) and, man, that is a high for a math student.