Hacker News Reader

> Strangely, there were many times where ChatGPT decided it would answer our medical question, but not in the form of a TV medical script. The tides had completely turned!

I've noticed this happens a lot with characters that don't have any long lines and/or have limited vocabulary.

Furby trails off pretty quickly, Abathur from StarCraft 2 (great for shortening code comments BTW) starts talking like a typical LLM after a few sentences.

View on HN

Holy shit are you me?? I just did a little side project last week to get my Claude code agents to RP as Abathur with me. Same results for me, it works for the first couple responses, but very quickly regresses to the mean. I tried writing out a prose/voice guide, I didn’t need it to use actual Abathur quotes, but it never seems to stick. I think there’s a thing the caveman mode does that re-injects the instructions to keep it in line, I haven’t gotten around to experimenting with anything yet though.

Plot twist: GP _is_ you.

That is a useful hint, thanks!

I make Abathur rephrase multi-line comments, because every LLM out there enjoys peppering code with them.

A successful recconect resets the counter. Efficient.

Doesn't always catch that distinctive style, but at least all the fluff is gone.

This Zerg is a software engineer at heart and has no tolerance for nonsense.

Reply on news.ycombinator.com