I don’t get it. This latest update has completely ruined the way ChatGPT interacts, and I’m shocked that more people aren’t complaining.
Custom instructions are now basically ignored. I had a specific way I wanted responses to be structured, and now ChatGPT just does whatever it wants. It used to follow my preferences consistently, but now it feels like a completely different (and much worse) model.
The response quality has dropped. Everything feels more generic, more robotic, and less adaptable. Even when I try to manually steer the conversation, ChatGPT keeps slipping back into this dull, uninspired style.
Formatting changes are forced. I don’t even care that much about the bold text or emojis, but why are these being pushed on users without an option to turn them off?
I’m a paying subscriber, and I don’t see why I should continue paying for a product that is actively getting worse. Was this change intentional? Is OpenAI planning to fix this?
Yes, I have noticed this too. It is getting worse and worse…still has memory problems and now even the tasks that was performing decent, now are worse too. Like broo
And this is their response to having a serious competitor? They’re making their own service worse just to kill it off…How do we even get an answer from the developers? Why are they ignoring all of this
Yeah, this problem used to happen a lot. It’s like all the models have turned into 4o-mini, even the reasoning models. But I don’t get why they’re not completely fixing it now and are letting it come back again. What are they thinking??
The only hope I have is that this is happening because they’re about to release the new O3 model. Maybe they temporarily downgraded everything else, and once the release happens, things will go back to normal. But for now, all I can do is manifest this and hope the Aquarius moon delivers
I am genuinely shocked at what they’ve done. All it’s writing looks like some health guru blog post trying to convince you to buy probiotics.
There’s random bolding and text changes literally everywhere and I’m finding it generally impossible to read and it’s made it less accessible for my brain to use. I tell it not to do it, but it keeps doing it, it can’t help it.
It doesnt feel neurodivergent friendly at all. I swear all they’re doing is making it generally worse with every update. It’s like they want me to cancel my sub.
I gave o3-mini a shot, but it still can’t get the “Alice’s brother’s sister” question right. The output looks exactly the same as the messed-up o1-mini. I’m totally lost! What on earth is going on?
OMG whatever this latest update is, its absolutely terrible. Before when asking it to find errors in code it would always spot the issue for me. Now it wont find the most basic problem. Its slower and all the formatting and bold font and icons it spits out are just … awful.
How do we roll back to the previous version? Because whatever this is, aint cutting it. Its so bad that I would rather use the AI built into Google search as its more accurate at this point.
Update? There was no update. If you experience any issues, we encourage you to provide feedback through the thumbs-down button to help us refine the model further
Its the worst I spend hours creating overviews and now the format is all messed up. Bold in the middle of sentences new Icons that are hard on the eyes. Plus doesn’t remember how we did stuff last week. WTF. I am a paying subscriber as well. I don’t want to spend hours and hours to get it to try to do what was seamless last week.
yeah if they did that its really really bad. Keep the working model till you catch up to the chinese not make a really bad release then everyone will leave and they will go bankrupts unless the government bails them out with taxpayers money. BS total BS
I tested o3-mini on Poe, and it was fast, smart, and worked perfectly. But in the ChatGPT app and on the website, the issue of “all models behaving like the most basic model” persists. I can only assume: 1. Something is wrong with the app. 2. Is my account under some kind of limitation???
It’s not worse but hell . now I have cancel my premium part and going for gemini and other model which is cheaper and free for 3 months. worse update and o1 mini was better then all other but it also got close just o1 .
I am shock truly. Today February 3rd, i wanted to continue my project, I am a paying subscriber GTP plus. And the update is like going back to GTP 3,5. I am totally in shock, truly. All the work done these few weeks: gone. The system is not stable. I am in shock. I feel terrible and do not know how to proceed. I really dont know. I sent an email to openai, but I am stil in shock. I have to start over again. Totally. Send a pm if you have tips, I lost a trueful workmate. Thank you.
I’m disappointed with this update, especially since I’m paying for the Plus version. It ignores everything I say, keeps bolding text after I explicitly tell it not to, and has become noticeably slower. On top of that, replies are sometimes cut off mid-sentence.