I've been using GPT-4o and honestly, I'm still finding it pretty frustrating. The same old issues like repetition, confusion, contradictory responses, and it often ignores my prompts. I'm curious if anyone else has noticed any changes in performance recently. It seems to me that sometimes in the morning, it performs better, but then the problems just start creeping back in throughout the day. Anyone else feel this way?
6 Answers
I think we need to define what 'better' really means here. For some, it's about accuracy, while for others it's about creativity. I haven't seen a significant change; it still has its quirks.
I feel like I jumped in at the wrong time. It's helped me, but it seems to forget the simplest info I’ve given it.
Honestly, if you're looking for something fresh, I suggest trying Claude. It has a huge 200k token context, which is awesome! However, it does lack personal memory and threading, plus it's a bit pricier, but the quality is way better than GPT-4o.
From what I've seen, it did dip in quality, but I thought it got a bit better this week. Still, I have to admit, I miss the previous iteration of GPT-4o. That was my favorite!
I have to say, I don't think it's improved at all. It lost its spark back in late May and has been hallucinating responses way too often. It's a real letdown.
Yeah, I noticed a decline around late April too. It's definitely not the same.
It really depends on what you're using it for. Since GPT-3.5, it feels like improvements come at the cost of other abilities. It’s a mixed bag—some things get better, while others drop off a bit.
I'm on GPT-4o for creative writing, which I used to enjoy, especially GPT-4.5. But the waiting for quality is really disheartening.
I'm still feeling dissatisfied with it. My sister is using the free version, and she says it's great! Makes me wonder if it's just me. Plus, I preferred the previous version much more.