A Vaccine for Anthropomorphism of AI (opens in new tab)

Would also be helpful if we, and the author does not here, avoided use of inappropriately anthropomorphic concepts to describe LLM output, such as "hallucination". We already have a more appropriate word available, "confabulation".