Hacker News new | ask | show | jobs
by PunchyHamster 65 days ago
but filtering a particular token doesn't fix it even slightly, because it's a language model and it will understand word synonyms or references.
1 comments

I'm obviously talking about network output, not input.
Good-token/bad-token overlap is near 100%. For example, try interacting with quantitative data, or program code, without using these tokens:

> :(){ :|: & };:

Now try running that in your shell.

which you can affect by just telling it to use different wording... or language for that matter