Follow
If your midterms/finals/professional examinations can be passed by large language models, consider that the issue may not be that LLMs are SMRT, but rather than your examinations fail to measure SMRTness. Only regurgitation.
@cerebrate
I have some experience with a reputable medical school and I regret to inform you that, filtering for Original Thought™ capabilities would hurt the bottom line too severely to be a workable solution, even if it is obviously the right one.
Your challenge, in short, is to upgrade them to require notarized Original Thought™, which as we all know is the correct collateral to demand before crediting someone with intelligence.