This is a bit of a bonus episode. I decided to see what an LLM could tell me about the uses of LLMs.

Claude was happy to oblige so I took his parts of the conversation and ran them through elevenlabs. Then I recorded my bits and edited it together.

It's a little shorter than my usual episodes as the AI voice can be a little grating, however I think there were some interesting points mentioned. We talked about the value of subjective writing, and about how Buzzfeed may have snuck into his training data.

I'd love to hear your thoughts on this feedback form

Notes:

I've not changed any of the text, but I've trimmed it down a little. (This is nothing different to my regular conversations with humans) You can see the whole conversation here https://docs.google.com/document/d/15YpY0wGfbospInXOlsdB6FRQEjUz03fCDgMEKs-JrfQ