I’m a Plus subscriber for a long time and never had that before. but for the past 2 days I can’t work with gpt-4. The responses are frozen in the middle, they are extremely slow, the code is broken and goes beyond the black code window as you can see on the attached screenshot. the responses often cut off with “Error in message stream”. yesterday in the middle of writing code gpt-4 started to generate images of a programmer sitted in front of a computer. looks like all the things that gpt-4 can do are mixed together at ones. That behavior of gpt-4 forces me to regenerate the responses over and over again and run out of the cap quickly. Please consider to fix that as soon as possible.
I’ve actually opened to forum to write about the same thing!
Not sure what is going on.
I am not working with code, but text. And it now breaks text into paragraphs in the most weird places, breaking in the middle of the word, suddenly jumping topics, loosing the chain of thought.
And response creation seem to have slowed down drastically.
For now 1-3 regenerations are needed to fix the issue. Which is really strange. Hope anyone has insight on what’s happening.
Another symptom is not understanding the task from the first generation with prompt clearly starting with a goal/task.
Has encountered it yesterday, when the task was like clearly “outline a LinkedIn post”, but it instead started summarizing the information given in prompt/attached file and then understanding what is needed from like regeneration 2 or 3 without any changes to the prompt itself.
For the past 4 days, I’ve been encountering major problems with GPT-4 that are making it almost impossible to use. I’m a Plus subscriber, and I’m really frustrated with how this is impacting my work.
Here’s what I’m experiencing:
- GPT-4 keeps freezing up on me. I have to hit the regenerate button over and over again to get it to respond.
- When it does respond, it’s super slow. I’m spending way too much time waiting for it to load.
- The code it generates is often broken and messy. It doesn’t fit in the black code window, and it just looks wrong. I’ve attached a screenshot to show you what I mean.
- Sometimes the responses just cut off in the middle, and I get some kind of error message about streaming.
- And twice now, GPT-4 has given me a bunch of garbled code, and then it generated an image of a programmer sitting at a computer. That was just weird.
All of this is forcing me to regenerate responses constantly, which means I’m burning through my usage cap way too fast. It’s seriously affecting my productivity, and I’m not getting the value I expected from my Plus subscription
For the past 3 or 4 days I have been getting “NetworkError when attempting to fetch resource.” and “Error generating response” almost continually.
Regenerating the response sometimes works, but often doesn’t and causes me to hit the usage cap after no time.
Currently GPT-4 is unusable for me. If this continues for much longer I will be forced to reconsider my Plus subscription.
It seems to me what’s happening is GPT is being cut off and then reattempts to finish what it started.
This use to be pretty common back in the day () when we had a much smaller token limit and had to say “continue” or eventually click a button for it to pick up where it left off. Exact same results.
I’m going to chalk this up to some network issues happening in the back, or even worse, some “optimization” feature poorly implemented. Not the first time.
Same for 3 days, it’s really annoying to get these issiues when you’re plus subscriber.
I’m experiencing the same thing
And broken code and incorrect responses etc, it’s driving me up the wall
Hopefully they can fix this right away
I don’t believe it is in openAi interest to fix it, just try to imagine how much processing power they are saving this way.
Until real competition arrives they will continue this kind of tricks, it not ChatGPT problem, it’s human greed problem.
Another Plus subscriber here, this needs fixing quick - very frustrating.
I just wanted to report that the situation has got slightly better for me in the past couple of days. GPT-4 is now more-or-less usable again
Although that said, I am now getting intermittent “Something went wrong” errors.
I followed all of the advice here, but nothing helped.
Did things improve for anyone else? Or are people still facing problems?
Same for me, it seems not to be down but idk. Changed the browser etc.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.