And depending on how OpenAI tweaked it this time it will either realize its mistake after being made aware of it or double down even harder on it.

I only use it for coding and it once told me my code not working was due to a bug in Webkit, so I asked it which bug specifically. It created links to bug reports but rewrote the titles of them. So initially it looked like it had numerous sources that backed up its statement but when I clicked on them those were bugs about totally different things.

It would not back down even after I specifically told it “You just made all of this shit up and even rewrote the titles” and got stuck in a loop of “I’m sorry, but you’re wrong and I am 100% sure I haven’t made a mistake”.

Kinda creepy. Especially when you think about the system rewriting reality when it comes to much more important things. Let’s just reinvent some history, that would be a good idea, right?