Is there anyway to prevent Chat GPT form being overly confident. Lately, even for simple tasks its wrong 90% of the time and then it argues with me try to say its correct. And then when it does find out its wrong it creates a fix where its “DEFINATLEY” right, yes in capitals.
And then… the new output is wrong anyway.
Its worse than a child.
How do I stop this child like behavior.