If I can't get ChatGPT to not hallucinate consistently over basic tasks nor truly understand what is said, this is a terrible idea. Forever getting "my apologie

If I can't get ChatGPT to not hallucinate consistently over basic tasks nor truly understand what is said, this is a terrible idea. Forever getting "my apologies" as constantly correcting it. Total waste of time for most things, but that isn't to say that AI/ML as a field is a waste of time, far from it—I just think LLMs are largely all sizzle and very little steak.