> how would this not lead to problems with evaluation metrics down the way? Assignments will just need to get trickier so copilot breaks on them. I guess it wi

> how would this not lead to problems with evaluation metrics down the way? Assignments will just need to get trickier so copilot breaks on them. I guess it will be the captchas getting harder because AI is getting better all over again....