You are viewing a single thread.
View all comments
2 points

It’s my understanding that, with LLMs, being polite can net better results because they’re trained off human interactions where being polite is better.

It’s not that they care at all, it’s just how language works.

I’ve not seen a study of the phenomenon or even tested it myself, but it seems plausible.

permalink
report
reply

ChatGPT

!chatgpt@lemmy.world

Create post

Unofficial ChatGPT community to discuss anything ChatGPT

Community stats

  • 163

    Monthly active users

  • 116

    Posts

  • 203

    Comments

Community moderators