> When weak candidates get stuck, they blindly try stuff hoping something sticks. If you tell them to use AI, their process stays sloppy.

But what happens, if Claude one-shots the solution? Will you consider the candidate to be competent then? Or are the interview tasks designed in such a way, that would prevent this?