I'm kind of annoyed by ChatGPT claiming - in an authorative tone - things that are wrong. This particular model should be trained to be humble and admit when it doesn't know something instead of making stuff up.