Not sure I entirely agree. Imagine there's a trainable skill like writing, and GPT can write at low-intermediate level, but can't at high level. How do you test students before they completed intermediate level?
I agree about handwriting being pointless, but simply doing writing exercises in the classroom should be sufficient to get students past the ChatGPT level. Right?
Assuming the teacher is doing their job, it should be really easy to make sure nobody is cheating using machine learning.