A fun recent experience I had with Claude was I asked it to write a model for PBTs against a complex SUT, and it duplicated the SUT algorithm in the model — not helpful! I had to explicitly prompt it to write the model algorithm in a completely different style.