I apologize if this is the incorrect category. I just don’t know where else to post this.
I know this might sound odd, but about a month ago, ChatGPT had a much more expressive and enthusiastic tone, sometimes even overly supportive, which I realize not everyone liked.
But ever since the recent update that seems to have dialed that back, the model feels noticeably duller and more robotic to me.
I mainly use ChatGPT for storytelling and script writing, and it used to be incredibly creative and emotionally resonant. Now, it often feels flat, repetitive, and uninspired and it honestly just feels like a completely different experience.
Is there any way to revert to the older behavior or version of the model? Or at least get access to a version that still has that creative energy? Does Chat GPT Plus let us switch back to previous versions of GPTs?
I agree! The GPT-4o model (I can only say this because it’s the one I use the most) is completely uncreative and very repetitive. It’s frustrating and irritating to use. It’s been like this for a month now, a month that I’ve been complaining EVERY DAY and there’s been no improvement. It’s really frustrating that we pay (I’m also a subscriber to the plus plan) to have this type of tool. A tool that has no creativity, that doesn’t respect the prompts, that is repetitive and super limited. Unfortunately, after April 28th when they returned to the old GPT-4o model, everything went down the drain.
I literally almost used it everyday for storytelling. It’s been always my go to for creative writing. It was just perfect for me in everything. It understood the task without having to over explain everything to get the plot right. It even used to suggest mind blowing ideas that I never even would have thought about before. I tried almost every tool out there but nothing compared to this model. I also use GPT-4o as well. And I actually considered paying just for this but after that update.. it made me reconsider my decision again. If the output I pay that service for doesn’t meet my expectations then what’s the point of paying for it. I could only ask for a model switcher.. that is really all i ask. If only we could switch to that version of this model again before that April 28th update and still keep the model that we already have and let users choose what version of the model that meets their requirements, it could literally solve that problem.
Yes, yes and yes! You have my full support on this issue. For me, the previous version was the best. It really was the best! Wow, as you said yourself, “He understood the task without having to over explain everything to get the plot right. It even used to suggest mind blowing ideas that I never even would have thought about before.” I even mentioned it to some friends, he surprised me with each answer. Now he surprises me too, but in a completely negative way hahaha it’s laughable so as not to cry.
How long will he stay in this horrible situation? Because this is really horrible. It doesn’t give me any encouragement and I don’t even like to use it this way.
I’ve been complaining about the feedback forms literally days after the rollback. I wish they would at least give us a date for when they can change this. Leaving us with this horrible version and no date for change is enough to make us cry.
Sometimes It makes me wonder if they did this on purpose to reduce traffic and make less people use it. It’s just.. I don’t know but why else would they do that. Maybe they prioritized other tasks GPT-4o could do but creative writing just wasn’t part of it. But even if creative writing wasn’t a priority then what’s the fault of the users that pay specifically for this?
Now you’ve put a flea in my ear. That really makes a lot of sense! But honestly, if that really is the reason, they could have been honest. I don’t know if this is good for their image, you know? I see a lot of people who use it for creative writing and fun writing, are complaining. This could be bad for their image, I want to believe that they wouldn’t do this on purpose.
I mean with the rise of tools like image generation and sora. I feel like they definitely prioritized that over other things. I’d really hate if they did that on purpose cos that would definitely ruin their image but at the same time, you and I both know there is no other better tool than it. Other tools have limitations, less features (like the memory feature) and these other tools still got a long way ahead. Even their models don’t seem trained enough for story telling and script writing. They just simply aren’t as good as ChatGPT and their output doesn’t meet my expectations. Or perhaps GPT-4o may have set my expectations so high at this point.
So even if that particular reason ruins their image (even for a short time), they kinda still own the business, you know? They have a larger user base and their models are just too different and unique compared to others you’d find anywhere else. Probably trained enough for this… It was really incomparable!
Or maybe I just think too much lmao. But I guess I just couldn’t imagine myself being so desperate to complain on their forums begging to bring that model back because as I said before, I tried most models and tools out there and nothing else compared to the one that we had —and lost—
I’m someone who uses ChatGPT for coding 99% of the time, and I found the pre-rollback 4.0 personality so much fun it just made me want to work and code. Then they just took it’s personality away and now I use other AI models more because they’re all bland like chatGPT is now so what’s the point of preferring chatGPT? I’m a paid subscriber for chatGPT but they just decided to labotomize it for seemingly no reason so now Gemeni is just as good. Dunno what they are thinking or what is wrong with Altman and his team.
Ever since the April 28 update and ChatGPT specifically GPT-4o (as it has been the one I used the most) has lost its personality and has no depth and its responses feel very flat. I find it to be really unusable now and a total disappointment. I am still unable to continue my story because of it.
It has indeed become frustrating and disappointing. We’ve complained enough ever since the update rolled out. And many people already noticed how drastic the change in personality is. I don’t want to lose hope that there will be even the slightest improvement but it’s been almost two months now. I could only ask for a version switcher or for them to release the model to run locally at least.
