Skip to content

With LLMs, config values become suggestions

You have suggestions of what outputs may be based on some vague idea of what the input may be.

Episode 237: FF-P-MEG (episode page, opens in a new tab) October 10, 2025, 2:28 p.m. CT 00:46:55 (play audio from this timestamp, opens in a new tab)

Voice-matched

Summary

In traditional software you set a discrete value in a config file and get a predictable output. With LLMs there are no discrete inputs and outputs, only suggestions of what the output may be, based on a vague idea of the input. So you end up testing over and over and talking to the model ('you are a chapter creator master...').

Not in the audio. Guest episode (Daniel J. Lewis; 0.35 threshold).