From Newsgroup: comp.sys.ibm.pc.games.rpg
Verily, in article <
mml53eF6sm2U1@mid.individual.net>, did
jfwalby@gmaill.com deliver unto us this message:
Wizaurdry is just well enough known that ChatGPT can field questions.
It's also rare enough that ChatGPT is entitled to many many mistakes on
the topic. This last is infuriating. How do I fix it?
Have lots of online conversations about it. Usenet won't help, since
nobody's training AIs on Usenet, but Reddit might and Instagram
certainly will. The next time ChatGPT is retrained, your information
will be there for it.
--
Trustworthy words are not pretty;
Pretty words are not trustworthy.
-Lao-Tzu spoke those pretty words.
--- Synchronet 3.21a-Linux NewsLink 1.2