Não raciocinam. É simplesmente mais um sensacionalismo da OpenAI. Continuam, como sempre, simplesmente regurgitando o material no qual foi treinado. Quando se modifica um pouco o problema, ele se perde.

You approached me. I responded in the same way you approached to me. I wasn’t rude - I just meant “there’s no need of bully behaviour”. If you are angry, you are angry alone. I do not want to upset you.

Am I disrespecting you? I don’t think so. I’ll keep it that way.

I value mutual respect, bro. If you don’t, it’s fine. Have a nice day. :)

No. I just expected that this platform would not have the same problems as X.

Now the name calling starts.
It almost seems your stance is just an excuse to abuse and bully people.

Nothing. That’s exactly why we should behave that why and not like them.

I never said you were wrong. But your bully, imperative, exposing behaviour is more appropriate in X than here. And btw, I’m Latino. So, calm down, Karl.

Curious what bullying behaviour has to do with the matter at hand.

No. :)

And you like to "expose" people for actually behaving normally. Have you ever tried to engage in a civilised debate?

You like telling people what they should do, I see.

*Or any other LLM.
The thing is: it simply cuts and paraphrases instead of summarising. And assigns importance by volume, not by meaning. Because it doesn't know what meaning is. It doesn't understand.

Sometimes this method works. Other times it doesn't.

Anyhow, it doesn’t do what it promises.