Dan Lyke on Nostr: Still adding LLM features to work project. Still having the "dancing bear" reaction ...
Still adding LLM features to work project. Still having the "dancing bear" reaction and not happy with what I can prompt ChatGPT 4o to give me.
Anyone have a "smart person's guide to prompting LLMs" suggestion? I'm already cueing it to give me extra stuff in machine readable form, but it all seems like such bad SEO inspired insipid elementary output, and, of course, gets facts wrong.
So how are people seeing these things as valuable, and getting useful output?
Anyone have a "smart person's guide to prompting LLMs" suggestion? I'm already cueing it to give me extra stuff in machine readable form, but it all seems like such bad SEO inspired insipid elementary output, and, of course, gets facts wrong.
So how are people seeing these things as valuable, and getting useful output?