Part of the skill in using LLMs is knowing when and how to use them, how to set the 'temperature' (how 'creative' it will be in its response), and how to write

Part of the skill in using LLMs is knowing when and how to use them, how to set the 'temperature' (how 'creative' it will be in its response), and how to write a prompt that is less prone to illusory responses. My eyes were opened one relaxing morning when sipping my coffee and pondering how to tidy up a database column by migrating from string to enum. I asked ChatGPT for its thoughts and its response seemed perfunctory and on point, until at one particular line, tucked in the otherwise sensible migration file [1], it casually recommended deleting all users whose value for that attribute wasn't among those specified by the enum. I spat my coffee out and learned a very valuable lesson that morning! [1] https://imgur.com/a/ejIdCH6 https://imgur.com/a/ejIdCH6