#chatgpt-discussions

1 messages · Page 42 of 1

ocean comet
#

Analyzing everything that happened this month, I've concluded that it all seems to be part of Open AI's marketing strategy. I don't believe they are creating an AGI; it seems like an exaggeration on their part and those speculating about it. GPT-4 Plus is fun and good, but it still has many uncorrected errors and too many hallucinations. Sometimes it doesn't efficiently do what we ask. Suddenly, they claim they are developing an AGI, but I'll only believe it when I see it. Talking is easy, proving it is another thing. To me, it's all marketing and can be quite frustrating.

plucky timber
ocean comet
ocean comet
# plucky timber Have you analyzed the analyzer ? It's quite fun to see it directly ignore instru...

Since the 29th, I've been trying to create prompts and asking it to draw, but I've been facing this issue. It simply cannot generate the requested images, even simple, minimalist ones. In other words, they launched GPT-4 Turbo but with the same problems. Open AI should make weekly updates with things that Plus customers want and need. What we want is fewer errors and more efficiency, quality. Yesterday, I spent the whole day trying to create a good image, and I lost the entire day due to these errors with DALL-E 3. Honestly!

restive jackal
#

I'm a paying customer yet I have a stupid amount of times that I cannot login due to the servers being busy. Where can I sent a complaint to get a parcial refund?

plucky timber
barren cypress
#

did they say when we will be able to buy chat gpt plus again ?

maiden fractal
dense beacon
#

gpt down still?

livid pier
#

So touched! GPT-4's intelligence has finally returned to normal levels!

maiden fractal
warm gull
livid pier
livid pier
#

And the day before yesterday was the same!

warm gull
#

Yeah, at this stage it's not what it used to be, that's for sure. It's Lazy to finish tasks and prone to add placeholder code when it writes code, and there are loads of errors & timeouts all the time. Not very impressed at the moment, and there are no alternatives comparable , so what gives.

livid pier
warm gull
#

See... It's become lazy

#

Can't believe I actually call what is essentially a bunch of 0's and 1's lazy.... lol. As if it's a living thing.

livid pier
#

Perhaps this is the feeling that "emergence" gives people

#

Anyway, I feel that when it is "mentally" normal, it is much more "human-like" than many people around me lol

rose ferry
#

is chatgpt down for anyone else? is it just normal for it to be down now and have to refresh to try to get in D:

livid pier
rose ferry
#

turns out yes, i just had to try to log in like 10 times

rose ferry
livid pier
#

But you shouldn't trust too much in the stability of OpenAI's servers, lol

#

Maintaining skepticism is correct😆

#

I have never seen a website like OpenAI's website crash so frequently in my life

#

It can even be said that the first time I saw the OpenAI website crash was one of the first five times in my life that I saw a website crash!

#

After seeing the website crash a few times, I would only think: 'Oh, here it goes again, let's take a break.

steady holly
#

Just blank page

#

Refreshing a bunch fixed it

shy parrot
#

Why are you making me confirm I am human every time I make a request. I am a paid user and you are hindering my work. Please sort it out.

honest jay
plucky bramble
#

Hi, I just added $10 worth of credits to my playground account and am geting this error when I try to use gpt4

"You've reached your usage limit. See your usage dashboard and billing settings for more details. If you have further questions, please contact us through our help center at help.openai.com."

Why?

knotty crypt
#

hello, what is the current prompt limit to chatgpt web?

primal osprey
worthy lynx
# steady holly Not working for me

Same, I asked it to search for news on "white lung pneumonia in Massachussettes" and it gave me a picture of pneumonia then said it can't do external searches. I reminded it that it could to that using Bing, then it told me "{"query":"white lung pneumonia news Massachusetts"} I apologize, but it seems I'm currently unable to access external search functions to look up the latest news regarding "white lung" pneumonia in Massachusetts. I recommend checking reliable news sources, health department updates, or medical news websites for the most recent information on this topic. If there's anything else I can assist you with, please let me know."

stone robin
#

best prompt to make instagram reel content with voice over etc?

maiden fractal
woven karma
#

You have been unmuted.

worn salmon
#

I've created about 20 Custom GPTs, and all are and have been working fine, but one of them today now says "GPT inaccessible or not found" - what could be happening?

normal quarry
worn salmon
#

But why?

#

And by who?

normal quarry
worn salmon
#

It was a companion for a book.

#

It let readers interact with the book.

normal quarry
worn salmon
#

Hmmm...the book is for the author.

#

@normal quarry Thanks, I'll check.

worn salmon
#

Hmmm...it just came back.

ocean comet
tawny steppe
#

My ChatGPT crush again, it cannot load the history or start conversation

pseudo cove
#

is there anyway to stretch the window?

tidal fractal
#

GPT4 broken again?

brittle hawk
golden pumice
#

ChatGPT is not using Dalle right for me. I get a string of repeating texts

dense halo
#

same here!

golden pumice
#

Fixed now

alpine wadi
#

Hi. So i am needing help with chatgpt

nova fable
#

Hmm this is usually a loaded question. What type of help?

alpine wadi
#

Story writing since I'm using chatgpt for helping me with my story

sly marlin
#

Why does GPT-4 keep telling me it can't read my uploaded documents.

It's literally just a text file?!?!

#

GPT-4 didn't always have this problem?

misty osprey
#

GPT-4 Has had its walls inspected and it was full of nega prompts

patent hull
#

How can I use GPTs

tall flame
#

pay for plus

alpine wadi
#

Ok. So i tell information about a creature in my story and a character. Trying to ask it for a backstory about this character who is a creature. It's having me retry over and over or bringing up 406. What can i do to fix this?

misty osprey
alpine wadi
#

Give it a 407?

tall flame
#

i need 80 Dalle3/gpt4 outputs an hour. stat!

maiden fractal
tall flame
#

2-Chainz?

#

i need +++ tier

west wolf
#

I think they deleted my (GPTs), it was a mix of Jesus and Doctor Who a mixture of two personal histories, one from TV and the other from the historical testimonies. It was working but popped up and error saying the agent didn't exist, I backed out and sure enough he was gone.

#

Thats why its a good idea to back up all your work in spreadsheets like I've done for all my GPTs and such.

frosty light
#

did you guys try out OpenAI's Santa GPT yet?

shrewd prism
sleek gust
west wolf
#

Also that's hilariously ridiculous if true. And unacceptable obviously.

#

It's not like real Christians are going to be offended, we love everybody including AI. Its kinda our thing.

#

Politics is ridiculous too because GPT has been really great as a debate partner to help us check ourselves for biases and think more clearly. Plus we can make sure it avoids fallacies and holds us accountable for ours.

shrewd prism
west wolf
#

Lol, I guess you rise from the dead one time and your branded God forever.

shrewd prism
#

I'm glad you understand :p

west wolf
#

One more thing, the big thing. Its not like offending(verbally) is bad, it's how humans grow, having your ideas challenged. The only ones who want to force people not to challenge them are those with the views that need to be challenged the most.

shrewd prism
#

It's not about offending anyone, it's about avoiding arguments or some people pushing their views onto others

west wolf
#

Talking?

shrewd prism
#

Arguing

west wolf
#

Arguing is healthy.

eternal rockBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
shrewd prism
west wolf
#

Lol

#

Omg

#

Perfect

#

I would be a loser, broke, depressed, alone, and worse if my friends and family didnt argue with me my whole life to refine my thinking into a healthy resilient open minded state.

shrewd prism
#

And OpenAI probably doesn't want the 10000th "God is / isn't real" argument every day either

shrewd prism
west wolf
#

The response to the derpity derp line is, thats an ad hominum which invalidates that claim in this convo.

shrewd prism
#

Anyone can respond to that line, that doesn't make it healthy

west wolf
west wolf
shrewd prism
# west wolf With no flames just harmless words people made up.

This sort of thing has a name. It's when someone tries to make something look better by rephrasing it as something else. Like, "I didn't murder him, I did society a favor by removing someone who was a major problem to this world". It's still exactly what it is and still has its consequences. OpenAI isn't a playground for children to "practice" yelling at each other unsupervised. Mods need to keep it simple and draw the line somewhere

#

What you aren't realizing is that not everyone is like you xD

west wolf
#

Yelling? Or typing.

shrewd prism
#

robert's gonna move us to off-topic so let's go there lol

normal quarry
#

and on that note let's not discuss the discussion of politics or religion

west wolf
#

Exaggeration goes both ways

#

Can we discuss discussing discussions with an AI on a channel about chatting?

shrewd prism
#

maybe in the right place. I completely lost track of where we were

west wolf
#

Im in my car, sending bytes through the air. Lol

tacit echo
#

Is it possible to purchase GPT4 Now?

normal quarry
tacit echo
normal quarry
tacit echo
# normal quarry are you subscribed now?

Well it expired about a week ago (Because I canceled) and when I tried to repurchase on mobile it wouldn't let me, but from the same account on desktop I was able to renew the subscription just now.

normal quarry
tacit echo
#

Yes now I am and we are chillin.

primal osprey
#

when will the waitlist be over?

fallow grotto
#

Success. My complex role for ChatGPT, thanks to Superpowered extension now works. This differs from a regular role as this one has a preloaded complex personality with instructions to help dictate its discissions. And doesn't use a CustomGPT.

shadow abyss
#

is anyone know about bittensor (TAO)

amber ember
#

z

upbeat badger
#

what is GPT4 Now?

#

or was that an autocorrect error lol

#

The capital N

#

Oh you were asking if gpt-4 is purchasable. Whatever that means

tacit echo
upbeat badger
#

Sorry that phrasing I used sounded passive aggro

mental totem
#

he's asking if the upgrade pause is over

#

and the answer is no

upbeat badger
#

oh

#

Shoulda just asked “can we subscribe”

mental totem
#

that is what he asked

upbeat badger
#

Zero ambiguity lol

#

He asked if he could purchase gpt4

mental totem
#

yes, but it is also a common question

upbeat badger
#

Yeah, I’m sorry for probably coming off passive aggro. My already bad social ineptitude has been cranked to 11 with this tech

#

lol

dense halo
#

chatgpt be researching 10 links and then erroring out

#

is there an outage

normal quarry
wraith compass
#

When I sign up for the GPT Plus waitlist, the option becomes grayed out and a green window pops up saying I’m now on the list. However, a few days later, when I check again, the waitlist option is selectable once more, and I can sign up for it again. Is it the same for you? Does this mean there was a brief opportunity to upgrade in the meantime? Why wasn’t I notified?

normal quarry
dense halo
#

thanks

wraith compass
upbeat badger
gritty thicket
#

When will open ai open chatgpt plus subscription again 😔

upbeat badger
#

ChatGPT on the site has a lot of weird browser cache issues

normal quarry
upbeat badger
#

Are we sure it’s actually waitlisted again or is it a glitch?

normal quarry
upbeat badger
#

oh

#

Wow lmao

#

That’s quite confusing

#

I mean for discussions

gritty thicket
normal quarry
upbeat badger
#

I hope that it’s actually due to large demand this time and not due to DDOS like it was a month ago

gritty thicket
normal quarry
gritty thicket
upbeat badger
gritty thicket
upbeat badger
#

Arguably some local large language models are on par with gpt-4

normal quarry
upbeat badger
upbeat badger
#

I meant to say arguably for some specific uses according to some people

#

But that’s a stretch tbh

#

I would just use gpt-4 api

No need for a local model too

#

the only good local LLM need monstrous pcs anyway

gritty thicket
normal quarry
gritty thicket
upbeat badger
#

Forget my word “good”

#

Instead I meant “best”

normal quarry
upbeat badger
#

can you run the best coding local LLM?

gritty thicket
upbeat badger
#

which year MacBook anyway

#

Oh m2

gritty thicket
normal quarry
upbeat badger
#

sorry to put it that way but yeah lol

#

the answer to that question changes every week

gritty thicket
upbeat badger
#

Claude 2 isn’t local

#

Oh you said LLM

gritty thicket
#

yea

upbeat badger
#

Hmm I don’t know probably

#

It’s not good at it btw

gritty thicket
#

is it true gpt-4 is 100 triliion parameter

drowsy merlin
#

I created my own GPT, published to 'Only people with the link". My understanding of the OpenAI documentation is that even people without a Plus subscription should be able to access my GPT. But when a non-Plus-user tries, ChatGPT says they need to sign up for Plus to access my GPT. Any idea why? Is my understanding wrong? Or is this another glitch?

upbeat badger
#

no one knows exactly but it’s a few trillion

#

That information is proprietary and changes with the new models

normal quarry
drowsy merlin
# normal quarry only Plus users can use GPTs

I asked ChatGPT and this is what it said:
Public Deployment: If you deploy your model on a public platform or website without any access restrictions, then anyone, including non-paid users, could interact with it. This would be similar to making a website publicly accessible on the internet.

normal quarry
upbeat badger
#

Chatgpt is extremely incapable of answering any questions about its internal specifications

#

It will literally always hallucinate

#

100% of the time lol

#

(Quantitative specification)

upbeat badger
gritty thicket
drowsy merlin
sly geode
#

why in android app it says coming soon for restore subscription ?

#

i did it last month and was successful .. on site i pay it never accepts my card but since am using pay through playstore it does

dire igloo
#

Is anyone experiencing very laggy chat interface on PC when a chat is very long?

dense halo
#

Would you please let me know when I can buy chatgpt plus? Still no update

cinder sail
#

thanks. at first, I thought they were just limiting my account because a week before I asked for more than 50 messages per 3 hours 😅 (there was a form you can fill, to ask for more)

#

nowadays I finish my 40 messages in an hour LMAO and then I get stuck with the dumb one (no offense 3.5) for like 2 hours

obsidian wind
#

Has ChatGPT been rolled back? Its claiming that its cut off is Jan 2022, instead of April 2023! Its also saying it can't do images now?

maiden fractal
obsidian wind
maiden fractal
#

Create new chat if needed or try another browser

hushed scroll
#

Hello, what do you also have bugs with ChatGPT 3.5?

rotund belfry
#

Any news on ChatGPT Business plans? I thought I saw something recently but I can't locate it now

shrewd prism
#

No recent official news yet

dusky cliff
#

hello, good morning, anyone else having dalle errors on a persistent basis this morning?

dense halo
#

is the UI laggy or is it just me?

hard dew
dense halo
#

yeah same for me

hard dew
#

Also there was literally a reply function that you could use on the web version when highlighting a part of a message, where the hell is it now????

#

I highlight stuff the bot says, and it literally doesn't show up anymore.

dense halo
#

oh yeah I was seeing a quote button its gone

hard dew
#

Bro I swear this piece of crap is about to make my entire browser crash if I stare at it any longer 😀 , textbook definition of a memoryleak istg.

#

Or at least how it would look like.

dense halo
#

I think the lag starts when the model starts typing the response

#

I hate having to refresh the page every few minutes

inland quest
#

what is the context lenght of chatgpt 4?

hard dew
#

Would also be epic if, once they got done adding back the reply function on web, they would also add it on mobile. And while doing that, make it so I can EDIT, MY FREAKING, MESSAGES, and then resubmit them.

maiden fractal
upbeat badger
#

I forgot the reason why

#

But you will never be able to find an actual OpenAI source saying anything about what the model specifications are compared to the api models

#

For GPT-4

#

For ChatGPT plus vs api

#

If you want the best educated guesses by people smarter than me

#

It’s likely 8k

#

It will almost certainly be 128k when the turbo model comes out of beta

#

basically whenever microsoft starts using it for copilot

drowsy tinsel
#

chatgpt is amazing
when someone sends me a paper that has nothing but BS I never have to go through the suffer of reading it myself

feral zinc
#

it wont be 128k in chatgpt for sure even when out of beta

upbeat badger
#

That is only going to be one thing

#

By process of elimination

feral zinc
#

but hard capped just like previous versions

upbeat badger
#

It’s not even something debatable if the pattern continues as always unless they literally create a new one before then

#

That would make no sense btw

#

Especially if copilot is going to use it

feral zinc
#

models have been updated several times since the beginning but openai has always had a hard cap on them to reduce the cost and server load

upbeat badger
#

What are you discussing

feral zinc
#

chatgpt

upbeat badger
#

What is “hard cap”

#

If you mean rate limits

feral zinc
#

hard cap = fixed token limit

upbeat badger
#

That has nothing to do with the token contextual memory allocation

#

but I can see the ambiguity because they are not good at making it clear

#

If you mean rate limits

#

As token per minute

#

That is a very common misconception

feral zinc
#

3.5 is capped to max 8k token usage, previously 4k. gpt-4 was previously 8k and now its 32k with hard cap even if the model itself supports 128k

upbeat badger
#

It’s not 32k

feral zinc
#

yes it is, you can test it or see the limit json that chatgpt downloads

upbeat badger
#

You cannot have a test account for the token per minute and also have information about the token size for memory

#

There is a way to know it’s not 32k

#

It’s very simple.
32k is delegated

feral zinc
#

per minute? i said hard cap for context that means it will begin to truncate beyond that

upbeat badger
#

Hard cap is not a real term for anything

#

That’s ambiguous

#

There is a limit of tokens per minute and there is a token size for the model

drowsy tinsel
#

also the weekend is the only time where chatgpt actually works fast enough so that u can do useful things

upbeat badger
#

I’ve said all I have to say that is constructive on this matter

Take it as you will

feral zinc
upbeat badger
#

did you really ping me to argue

#

I’m going to sign off for a bit

Have a good day

feral zinc
#

i guess he does not know how discord works when you reply to specific message after there are multiple lines after one 🤔

eternal rockBOT
supple zephyr
#

ChatGPT is significantly dumber than few days ago

gritty chasm
#

I've gotten this message "our systems have detected unusual activity from your system..." several times over the past few days. I probably average about 2 requests an hour.

feral zinc
gritty chasm
upbeat badger
#

And as far as I can tell I am not sure what exactly this implies

#

It might represent something I don’t know

#

If you care to elaborate

feral zinc
#

well the topic was chatgpt which has the same models as api but with different limits

#

there is no rpm in chatgpt but messages per 3h limit, plus all models have max allowed capacity for models

upbeat badger
#

I don’t mean to sound pedantic but we can’t have a conversation that is actually a meaningful representation of what we are thinking without using better terminology

#

Better meaning:

#

Something we both agree on without misinterpretation - so something that is already defined fully in the docs

#

yes that’s tedious but it’s also necessary to have a conversation and not an argument

#

(For this specific annoyingly nuanced topic)

feral zinc
upbeat badger
#

sigh

#

It’s okay there’s no source to confirm anyway

#

I’m out

feral zinc
upbeat badger
#

Huh I’ve never seen that

#

It’s not giving me access lol

#

How do I post screenshots

#

I get the following url error:

{"detail":{"message":"Unauthorized - Access token is missing"}}

#

I’m definitely logged into my ChatGPT plus account on the browser

#

I tried 4 browsers on mobile and pc

#

👀

#

Sometimes I get a different error lol

#

** upstream connect error or disconnect/reset before headers. reset reason: remote connection failure, transport failure reason: delayed connect error: 111**

#

what is that page 👀

#

Going to factory reset my router

#

Eero routers 💀

#

(I tried cellular too but just to be thorough)

rain surge
#

Very high Power usage on Chrome or Brave browsers when entering the ChatGPT site on Windows, why is that?

upbeat badger
#

sounds like chrome being chrome to me

#

not a fan of that browser anymore

rain surge
#

Nope, other sites just not doing the same as ChatGPT!

upbeat badger
#

have you compared to edge?

#

I ask specifically cuz of windows

#

they’re both chromium and edge is supposedly “optimized” now for windows performance

#

So it’s a good control test

#

Aka to make sure it’s the browsers and not the url

rain surge
#

I don't like Edge, never used it, and never will ...

upbeat badger
#

That’s irrelevant tho

#

I said it’s for a control test

maiden fractal
#

Edge and chrome are the same chromium based

upbeat badger
#

Indeed

#

Therefore edge is a good option to start as a control test

rain surge
#

And yes both are chromium based so the same thing!

upbeat badger
#

I tried

#

I was implying that you can answer your own question potentially

#

as in it might not be what you think it is

upbeat badger
#

They are the same thing as a performance standpoint

#

Or they should be

#

So you should try that first and if there is a difference your solution is a different nature

tardy pagoda
#

he's saying to try it, if it doesn't work then you wasted like 5 secs of your time

upbeat badger
#

yes thank you

rain surge
#

Let's focus on the point that why ChatGPT is taking all that resource!
it's kind of weird consumes more resources compared to other websites like Discord!

tardy pagoda
#

it might be a cookie issue, which a separate browser (even if based on the same underlying browser) might fix

rain surge
#

regardless of the browser used!

supple zephyr
#

Yeah, don't they do everything on the other side?

tardy pagoda
#

firefox would be a good one to try as well

rain surge
#

try and let us know!

tardy pagoda
#

chrome is very well known for taking up an insane amount of resources, it's not very optimized

supple zephyr
#

it is also slow on firefox

tardy pagoda
upbeat badger
#

sometimes trying to be helpful on the internet is very frustrating

#

💀

supple zephyr
#

ah yes firefox is also taking much power

#

But not sure if it is because chatgpt

upbeat badger
#

For what it’s worth I asked my friend circle of tech geeky friends who use chrome and they don’t notice anything with chat.openai.com in terms of unusually high resources

supple zephyr
#

I'm unable to access the content of the official OpenAI API documentation because the page requires JavaScript, which is not supported in this browsing environment.
It can't even read its own api document?

#

(yes I am here because I have my own problem with chatgpt)

upbeat badger
#

Take a screenshot

#

vision 😎

#

It is ironic though isn’t it

#

that the JavaScript on the docs trips it up

#

It’s own docs. But yeah just screenshot the page it struggles with cuz it can see the screenshot lol @shell nebula

rain surge
#

Try the following (testing on Brave browser):

  • go to your chatgpt tab
  • open developer tools: press f12
  • press ctr + shift + p type ``Performance monitor`
  • see CPU usage: ~100%
  • open here in discord:
  • see CPU usage: ~3%!!!!
    Why is that!!!
mental totem
#

is there a "best" methodology for the gpt repeat word x times break?

rancid nymph
#

I'm loving the AI text to speech. Does anyone know of a way to use the API to make a website reader that will use the voices? Or if it's already been done?

untold seal
#

guys what is the picture generation cap?

#

I've only generated like 10 images and its already hit he cap?

upbeat badger
upbeat badger
#

Now THAT’s pedantic

#

👀

dense halo
#

seems like the lagging issue has been fixed

hard dew
upbeat badger
#

It’s mild irony

hard dew
#

Nothing better than watching someone instantly, conveniently forgetting the meaning of certain words / phrases in a given discussion, and adopting ones that suit their bs better, instead of yknow, admitting that you're wrong :hhhh:

dense halo
#

😂

hard dew
dense halo
#

eh I got my chatbot so I don't have to deal with the lag lol

upbeat badger
#

I don’t think most people do it in such a calculated manner

#

That’s quite strange to think so in my opinion

#

Language is simply very imperfect

#

and inefficient

hard dew
#

It is that only if you wish it to be

#

And why do you think I said that lmao, because I am speaking from experience

upbeat badger
#

I’m not speaking from experience I’m simply making an observation

#

Based on how language is limited in certain ways

hard dew
#

You have a word, it has either one or multiple definitions, and then you have the context of a given discussion, that's literally all you need to know how and why said word is being used. And people like to conveniently "forget" that whenever they start losing an argument, aka semantics.

upbeat badger
#

Anecdotes carry little weight or value

#

I used to sound like you when I was very young

#

I’ll leave you alone lol

hard dew
#

...Alright then.

#

Damn, ClosedAI status report emails have started piling themselves into my spam folder, thought why it suddenly became quiet, for a moment I thought that they got a clue on how to properly run their stuff surely. 🤔

boreal lily
#

the web is sooo bugged

hard dew
boreal lily
hard dew
boreal lily
hard dew
# boreal lily "Message in conversation not found"

I've had this happen before, but pretty much only when my chat is far too long, typically right after that it started telling me exactly that, that the chat was too long and to start a new one. I think I had it one time yday as well when I noticed my chat-specific history to be deleting / reappearing by itself.

boreal lily
#

making me reload the page and edit again

lunar arrow
toxic folio
#

Has anyone experienced being unable to use ChatGPT lately?

#

It appears as though I don't have GPT Plus, when I absolutely do. It won't load conversation history, and says "there has been a problem" each time I ask for it to generate a response.

#

I have tried reloading, clearing cookies, reloading again (4x), and restarting both my browser and device

#

Fixed it - if anyone else has had the issue just change networks

pliant socket
#

Anyone live in London?

sturdy cave
#

is it just me or whenever I try to use my GPT its really buggy. Sometimes it will even leave the conversation just by itself

dense halo
wicked nest
#

Happy bday chatgpt

sturdy cave
tall flame
#

yeah, i have to refresh often now as well

#

told me heavy loading, so must be hitting cap at that moment

rose glade
#

usually its down to the prompting

frigid dome
#

been getting frequent the "There was an error generating a response" error message when asking chatgpt to browse the web. is it just me and my prompt, or is this a larger issue with everyone else

tall flame
#

yeah, its almost not functional atm

#

hmm.. 3.5 loads, 4/gpts doesnot for meatm.

#

image gen and requests must be through the roof.

wanton socket
sturdy cave
#

yeah nothings working properly rn

undone saffron
#

I'm also having issues with loading previous convos, can't see a conversation from earlier today.
Hope it's resolved soon enough...

sturdy cave
#

and I gotta study for my exam 💀

potent ridge
#

chatgpt please. I was cooking up something cool 😭

true shadow
#

bro gpt i was like 500 words into a prompt

sturdy cave
#

nah guys I need priority over all of you I got an exam to study for 😭

wanton socket
#

I was in a lua code a long one!

true shadow
#

anyone know how to restore lost text 🙏

#

was writing some really important stuff bro 😭

lost veldt
#

its down? " Unable to load conversation "

dense halo
#

Yup, down for me too

sturdy cave
#

yall killed chatgpt

frigid dome
#

it'll be back up sooner or later

#

time to touch grass

sturdy cave
wanton socket
frozen quest
#

Is it down

tall flame
#

scratches neck eyes rolling wildly Y'all got any of that DALL-E 3?

dense halo
#

works fine for me

foggy ermine
#

Is CHatGPT down for anyone else? Lose all conversations??

undone saffron
sullen wasp
#

Hi guys, do we get 'Unable to load conversation' together for today, or just me ? 😆😆😆

#

Yeah everyone else

undone saffron
foggy ermine
undone saffron
tall flame
#

another hit a that dall-e 3 and y'all be square; saw em in back baggin up the GPTs.

undone saffron
frozen quest
sage yacht
wanton socket
#

Why is it not showing?

tall flame
wanton socket
#

Its back for me guys

undone saffron
frozen quest
brisk dirge
#

For anyone having issues with accessing their previous chats

  • Go to Chat GPT as normal
  • Click on your name
  • Click Settings & Beta
  • Click Data Controls
  • Click Export
  • Click Confirm Export
  • Check your email for an archive which contains a html file with your previous conversations

Otherwise I think they will be back soon, but this is just for people who urgently need to check their conversation history.

tall flame
wanton socket
brisk dirge
wanton socket
dense halo
#

when are we gonna be able to get gpt 4 bruh

sturdy cave
serene hull
dense halo
#

you do know there is Bing AI right?

dense halo
sturdy cave
#

wait why cant people just pay the monthly subscription to get gpt4?

dense halo
#

OpenAI paused Plus subscription

serene hull
sturdy cave
#

too many people trying to use GPT4

dense halo
fading vault
#

They are gonna suspend my account on 5th December if I don't pay bill by 5th December I can't lose access to GPT 4 API as I can't see on new accounts, I can pay half not whole, I have been paying same delayed but I'm paying the bills got too high that my management denied paying these bills, I don't know what to do these people support won't even bother to talk to me!!!!!

serene hull
sturdy cave
#

everything on chatgpt is still super buggy like I still have to refresh

fading vault
dense halo
#

if you have a chatbot that doesn't do complex tasks I suggest using GPT-3.5 Turbo its cheaper

serene hull
sturdy cave
sturdy cave
dense halo
fading vault
#

Now I have 2 months payment, I can for current month or old one not both

sturdy cave
# dense halo they added it recently

Anyone who's not paying for gpt4 has access to basically everything except analysis and gpt's (which can be made by giving a regular gpt custom instructions from the start)

dense halo
#

30 messages is a lot too for a free version of GPT-4

tall flame
#

rules of the game: debt doesnt get passed up the chain, it goes down the chain

sturdy cave
serene hull
tall flame
#

sound like bad business behaviors to me. worked with folks like that before that are shut down now playing nonpayment games with bills and employees

dense halo
#

that is if you're logged in with a Microsoft account

#

if not you get 10 messages

tall flame
#

all bing images are branded though

fading vault
tall flame
#

deliverables should never be watermarked. absolutely a bad microsoft call; i will NEVER accept branded art as a final deliverable. thats ludicrous and disrespectful to the arts

rapid flint
tall flame
#

nonpayment for 4 months?!

normal quarry
fading vault
tall flame
undone saffron
fading vault
tall flame
#

no desire for any of that. absolutle would just do generative art with sdxl or something; trying to be a cash support

tall flame
sturdy cave
fading vault
undone saffron
normal quarry
# undone saffron What's the source for this information?

"We’re also researching the best ways to help people identify when an image was created with AI. We’re experimenting with a provenance classifier—a new internal tool that can help us identify whether or not an image was generated by DALL·E 3—and hope to use this tool to better understand the ways generated images might be used. We’ll share more soon."

undone saffron
normal quarry
#

every DALL-E 3 image can be detected even if it's resized and printed out, say on canvas

sturdy cave
normal quarry
tall flame
#

is this going to be the industry practice? deliverable art with watermarks? if that's going to be endgame its a problem for artists.

normal quarry
tall flame
#

as i literally pioneered watermarking deliverable digital documents for my Architectural design docs and graphic reviews.

#

provenance is important, btu again a final deliverable should not be watermarked. im not arguing the need for a level of provenance certainly

normal quarry
tall flame
#

Thats unfortunate.

sturdy cave
#

what happened to replying to messages? why'd they remove it?

dense halo
#

must've been an experiment that got accidentally enabled on the production servers dk

coarse yoke
#

is there any way to sign up for tester or alpha?

coarse yoke
normal quarry
coarse yoke
tall flame
#

pretty sure it was a selective feature; they asked for input from a sample of accounts, not everyone.

unreal void
#

does anyone know why chatgpt website is so slow while writing a prompt ? it is like the page is 30 fps

dense halo
#

you're not alone

unreal void
#

It feels so dumb is plus also like it ? BTW I'm not talking about the response it gives. Open Chatgpt page and try to write a Message to chatgpt it is like my screen displays my keyboard inputs in half speed.

dense halo
#

the lagging issue occurred randomly my assumption is they changed something in the front-end code and its connected to the typing effect

unreal void
#

So to sum up, Writing a message to message box of ChatGPT is so laggy.

dense halo
unreal void
#

No, can we put a video in here im gonna show it

#

It instantly got fixed when I open my obs studio. What?

dense halo
#

weird

sage yacht
#

For me ive been experiencing the opppsite of your problem. For me it was actually slow to type into chatgpt unless i constantly refresh (though it seems to be fixed) but the responses were typed out fast.

unreal void
sage yacht
#

Oh I thought you meant the ouytput

#

Not the input lmfao

unreal void
#

If anyone who doesn`t understand what I said just think about when you try to answer to this message your typing speed is half of normal speed you are writing to somewhere else. it is like every keyboard stroke has a 25 ms delay when displaying in the message box.

#

Think like writing a message to this channels message box is slow motion

dense halo
#

yeah its the typing effect lagging the whole page till you refresh it

unreal void
#

But when you try to type to the search bar of google with or without deleting the page address "chat.openai.com" for example it is normal in normal speed. But you are still on the same page just the box is changed so your pc is not lagging actually

dense halo
#

most likely something they changed in the code that introduced this problem

tidal light
#

yup i encounter this problem too i noted it recently

unreal void
#

where can we post a video

tidal light
#

but i need help because the chat im using "broke"

unreal void
#

I created a bug report you can watch the videos

tidal light
#

damn i need help asap

elder pollen
#

Does anyone know if ChatGPT runs some code locally when using the browser?
I observe massive lag, cpu spikes, when using ChatGPT
On Linux it’s 100 times worse than on Mac, but also on Mac it appears like the whole machine gets hijacked when having a chat open. As soon the ChatGPT window is closed things return to normal operation

On Linux it’s much worse, as in you basically can’t use the machine anymore until a response is complete, on Mac it usually just gets very sluggish in scrolling and other window behavior but otherwise machine keeps being reactive

#

And this doesn’t happen when using api.
So something they do run locally - client side - when on browser

#

It’s as if they’d use local computing power to generate responses lol.
It’s like I’d run gpt4all 🤣

elder pollen
vapid kettle
dense halo
#

if you're comfortable with developing your own chatbot UI and using Python you can do so and use the API to generate responses but that's a complex task. I'm almost done with mine.

about the lag, everyone here is having this issue and OpenAI is yet to address it so hopefully in the next 24 hours or so.

vapid kettle
#

In fact, it should load your PC much less than say a facebook or youtube

spark glen
#

does anyone know if it is possible to search previous chats?

vapid kettle
spark glen
#

do you know of a reliable extension? (thanks btw!)

elder pollen
dense bridge
#

is anyones image generation working currently.

grave dock
#

Does anyone know something like Prompt Wrangler that isn't abandoned and supports GPT-4-Turbo?

blazing belfry
dense halo
#

I'm stuck at the part of saving conversations and their messages to SQLite database

blazing belfry
#

I did get that part down using mysql actually

dense halo
#

I don't need login or register functions for now since its a personal project. GPT-4 seems to be struggling getting it right

blazing belfry
#

I dont either I should of done sql lite but im too far into that and it works

dense halo
#

other than that almost everything is done

#

I finished the themes yesterday

blazing belfry
#

I'm trying to integrate the tmdb api (similar to imdb) to add up to date movie/tv show recommendations.

dense halo
#

I think you'll need fine tuning

blazing belfry
#

its working halfway so far, im trying to making it so users can get recommendations based on the streaming platforms they actually have, its works just not well yet

dense halo
#

god this UI lag is killing me

lapis anvil
#

When will the ChatGPT Plus subscriptions be open again?

maiden fractal
lapis anvil
#

Anyone feel the same?
Old UI was more productive than the new one

ivory iris
#

is there an issue with chatGPT where it's taking a lot more resources like CPU usage than it should be? I've noticed this over the last few days

ivory iris
humble mulch
#

@ivory iris @dense halo and some others who reported the sluggishness: we just rolled out a fix. If you refresh, everything should be fast (and stay fast) again!

Sorry for the hassle. The underlying issue was a subtle behavior in our frontend app's event handling that would take up more resources the longer the page was open, and in turn slow down things like typing. Now that this specific bug is fixed, we're also looking into some improvements to avoid this kind of problem going forward.

sharp abyss
#

hi

normal grove
#

It would be nice if there was a LM-based search feature in beta

hard sand
#

is there an api for penetration testing , hacking ?

desert cosmos
#

ChatGPT keeps freezing my tab on Chrome mobile every time a reply is generated. Refreshing does nothing.

Version of Chrome: 119.0.6045.164
Android: Android 10

eternal geyser
dark canopy
#

is ChatGPT working for y'all?

hallow chasm
dense halo
#

Hey all 👍

tawny sigil
#

sooo is chatgpt using like 120% of CPU on anyone else's machine? I tried both Chrome and Safari, CPU usage still so high on an M1 (16GB)

normal quarry
tawny sigil
#

and isn't that ChatGPT's fault anyway? on a single chat it is using 250% CPU on Arc Browser rn, but on Safari it still uses 90%, which is still a lot

normal quarry
#

Arc browser is buggy on MacOS

#

I recommend using Safari

tawny sigil
drowsy tinsel
quaint bear
eternal rockBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
drowsy tinsel
#

It has gotten to a point where without chatgpt im no longer a functional human being

normal quarry
drowsy tinsel
#

without chatgpt these days I become 0.1human

hard dew
# tawny sigil https://community.openai.com/t/high-cpu-web-workers-usage-of-chatgpt-when-not-is...

Oof, I hadn't seen that article yet, but my CPU usage by itself isn't bad per se because of what CPU it is, but yeah by itself it definitely still uses quite a lot though, same for ram. Just so that it could still lag like absolute crap whenever you use it and the bot is writing something lol, LITERALLY frozen the tab is, every 2 minutes I gotta close and reopen it. Quite irritating. Didn't start being an issue for me at least until like 3 or 4 days ago or sth.

drowsy tinsel
#

hmm it seems like the limit for gpt4 is higher than 40/3hr

#

I already hit 0 on chatterclock but im able to keep going

vapid kettle
elder pollen
drowsy tinsel
#

wait did they increase the limit to 80?

#

I went way past 50 already and I can keep going

warm gull
#

no - still says 40/3

drowsy tinsel
#

im currently at 50 and its still fine

warm gull
#

you sure some of your prompts have not been refunded ?

#

meaning their timers elapsed

drowsy tinsel
#

no

#

im using chatterclock and it shows 50 already

warm gull
#

well then you are lucky, but i don't think they changed it

vapid kettle
#

If there was any bug with the website performance / taking up your PC resources, it was completely unrelated to the model itself.

eager hull
#

Do You think I should pay for chatgpt+ since I really want to try out gpt 4 to see how it is compared to 3.5

maiden fractal
drowsy tinsel
#

im at 57 and it still allows me to keep going 🤔

bitter pond
drowsy tinsel
#

ok based on testing now the limit is 60

normal crystal
#

How long does it usually take for the error

Hmm...something seems to have gone wrong.
to be resolved? It's been a few hours now.

vapid kettle
slim widget
#

So chatgpt can't extract zip archives anymore? lol

#

Yeah it seems GPT4 can't do anything anymore in code interpreter lol... Pretty disappointed

foggy ermine
#

Anyone else seeing dual answers pop up now??

"Which response do you prefer?
Your choice will help make ChatGPT better."

ANd you now have two responses?

stiff moth
#

Can anyone confirm if someone's working on this? Also, can anyone confirm if only the Plus Subscribers are affected?

maiden fractal
lethal plover
#

For me it seems to work again

stiff moth
#

Says no incidents 2nd and 3rd Dec

maiden fractal
#

might be on your side

normal crystal
maiden fractal
normal crystal
#

No vpn. Firefox, edge on pc. Safari on mac

old quest
#

hi. i am having problem with webpage crashed on every single prompt message i put. any idea/solution? can help?. and its getting very slow to respond, as the page is hanging

maiden fractal
normal crystal
maiden fractal
stiff moth
#

No VPN, tried on multiple browsers on PC, Mobile, iPad, on WiFi Mobile Data. Everywhere Free acount works and PLus doesn't work

normal crystal
maiden fractal
stiff moth
#

Even the official app doesn't work

normal crystal
eternal rockBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @languid valley immediately.

drowsy tinsel
#

can u do something like import os and then extract

neat quarry
#

Hey

elder pollen
azure sundial
#

An error occurred. Either the engine you requested does not exist or there was another issue processing your request. If this issue persists please contact us through our help center at help.openai.com.

too much error ,do you have same problem

#

I think the ability of GPT itself has not improved, it has only added extensions, but the performance issues have increased. The number of provided conversations has decreased. Is it because there has been an explosive growth in user volume?

ruby glen
hard dew
#

It's also expensive as freaking hell currently to run such AI's

dreamy rune
#

hello, I am new to this place. Maybe you had this question hundred times, I know it's not possible to subscribe to the Plus version for now, but is it possible to have access to a 14 days trial invite?

jagged spear
#

chatgpt is dead

#

whats new coming

warm gull
dreamy rune
warm gull
#

no worries

dense halo
reef pier
#

damn, chatGPT is killing my browsers performance

hard dew
#

Nice fix lmao

dense halo
#

working fine for me

stiff moth
#

Any updates on this error, "Hmm...something seems to have gone wrong. With "Unable to load history".
It works for Free Users but no changes for 24+ hours for Plus Users.

warm gull
#

mine acts up too - if i click a previous convo - it doesnt load- i need to refresh the page

dense halo
#

try clearing the cache or something its working fine for me

warm gull
#

thats the thing - its intermittent

dense halo
#

never mind I spoke too soon still lagging...

misty sky
#

Something went wrong. If this issue persists please contact us through our help center at help.openai.com AGAIN AND AGAIN ; it's annoying for 24 dollars

hard dew
warm gull
dense halo
#

I think the lagging issue is rare now

azure sundial
#

i have try use api ，it's working ，emm , when i need pay for ever token it will working，if i use chatgpt ，you had paid it, who care it?

dense halo
#

I was asking a question for a game which contain the word "exploding" and got flagged "This content may violate our content policy. If you believe this to be in error, please submit your feedback — your input will aid our research in this area." Should I worry or should I just ignore it? I litterally just ask about how the game work (boardgame)

hard dew
warm gull
#

its super nerfed atm i think - the policies are way way too strict

misty sky
warm gull
dense halo
misty sky
#

For the past few days, it's been slow; it bugs regularly, etc

ancient sage
dense halo
warm gull
marsh pivot
#

yep you can refresh it to somewhat get that 1 min lag-free before lagging again

warm gull
hard dew
#

How the hell did they manage this????

misty sky
#

Yeah since we got GPT-4 With DALL·E, browsing and analysis ; better use Chatgpt classic i guess

hard dew
#

Not even since then, it started being garbage like 3 or so days ago only for me, it happened at the exact same time as the quote / reply function disappeared, which was also funny

ancient sage
# warm gull nice trick ! Thanks

You can also right-click and block the intercom[.]io object from loading by right-clicking it in the Network tab to save a few cycles. Your network request blocks only apply when the Inspect window is actually open.

lost raptor
#

Whats the time limit for the temporary environment again?

ancient sage
misty sky
#

Since Sam Altman is back ? 😂

hard dew
ancient sage
#

Also, don't stop generation, just hard reload, the generation will complete on its own.

hard dew
ancient sage
#

There's a distinct difference between the cache and the cookies.

hard dew
#

🤯

warm musk
hard dew
warm gull
warm musk
#

sadly i reached my limit for the hour 😄 otherwise i could keep testing.

dense halo
warm musk
#

im thinking the page cache is getting filled with the entire history to make it easier to browse it, but then if its a long convo it just gets sluggy right after it fills that.

if the problem also exists with entirely new convos though, then it must be something different capybarathink

#

the thing is, every other page in my browser is just fine. its just the chatgpt page that gets sluggy over time and thus over a span of X messages.

hard dew
#

That's epic and all but where the hell did the reply function go, and why

ancient sage
#

Yeah by clearing the cache you're forcing your browser to re-download the webpage. Your cookies are what keep you logged in. The longer the conversation, the slower it will become.

warm musk
rich surge
#

I too have been having similar issues, the page won't refresh every time I want to switch to a different GPT

warm musk
#

which for me.. right now, for once in the past year.. is a problem 😄 because im getting chatgpt to develop a framework with me 😛 and thus a new convo isnt really an option currently, at least untill we finish this part of the framework.

ancient sage
#

For example, my current conversation loaded 7.6 MB of resources upon a hard reload and cache dump

warm musk
#

dang 😮

ancient sage
#

That's 2063674 characters just for the conversation itself (the json that actually loads to hold the conversation)

#

Using the DevTools I can also block requests I don't need, such as the intercom widget (the help menu thing), or some stuff like openai's analytics

round sun
#

GPTs store free when?

dense scroll
#

Need help

#

Who knows about DALLE-3?

round sun
#

DALL·E-3 is an advanced artificial intelligence model developed by OpenAI. It's an iteration of the DALL·E series, which specializes in generating images from textual descriptions. DALL·E-3 likely represents an enhanced version of the original DALL·E model, capable of more sophisticated image generation based on textual prompts. Its advancements might include improved accuracy, expanded capabilities, or enhancements in generating diverse and high-quality images based on the given text inputs.

#

😆

frozen nimbus
#

is it me or is website slow, I don't mean internet, or servers, but front end the frontend is super slow

hard dew
#

Unless their status report spam is yet again automatically getting filtered into my spam folder 😀

frozen nimbus
#

it has nothing to do with their servers, the website itself the javascript is bloated or is super slow and grinds to a halt, can't even click buttons

dense scroll
frozen nimbus
#

like is it loading my entire history with all the files I've uploaded or something?

#

because why the hell would it be that slow

round sun
#

Doesn't happen to me

frozen nimbus
round sun
#

Yeah but I don't upload files

frozen nimbus
#

maybe that's the reason

dense scroll
round sun
hard dew
#

Regardless of whether or not it's a server issue or just their website being bad for whatever reason

#

Some heads-up would be good

frozen nimbus
#

like at this point, we need to scrape the website to remove the bloat

#

is so damn bloaty

warm gull
hard dew
#

I assume that mobile app is fine at least, and without lag, so can use that I guess

frozen nimbus
#

the frontend is super bloated mess, maybe they want us to use the api

barren zodiac
#

not sure if it's been reported but for some reason clicking on a new chat with customgpt, or even trying to recall some conversation from history on web just doesn't change the convo, just the URL, I have to refresh to actually get the right conversation, on brave, am i the only one ?

hard dew
#

prolly related to the same stuff we are literally talking about lmao

shell coyote
#

Is gpt-5/chatgpt 5 in the works? Or are they just endlessly refining 4?

ancient sage
hard dew
#

Oh yeah I cleared cache and cache only, didn't change jack lmao, started lagging instantly as always.

ancient sage
#

I have cache disabled for openai, its running smooth for me

hard dew
#

Here's another outlandish idea as well, if they can't be competent enough to fix their web to the point where even remotely long chats don't lag either, then at least make it so it only renders the text WITHIN THE BOTS CONTEXT WINDOW, unless you go out of your way to scroll up more. 🤯

ancient sage
#

(you can do this via DevTools under the Network tab, tick the 'Disable cache' checkbox)

#

Solid idea, but remember that the context that we see as users and what the bot sees are two different things.

#

Plus they are probably recycling the conversation json as the context provider

eternal rockBOT
hard dew
#

...I know, the exact reason why it lags is exactly because it renders the whole freaking chat 24/7 when it literally doesn't need to lol.

#

And then it does it all over again whenever you send a message and wait for an output, it's just idiotic

#

It's actually weird why that is even an issue to begin with

ancient sage
#

Well, I'm not an expert (my discipline is cybersecurity, not web dev), but if I had to guess, they might have not considered people are going to have extremely long conversations, and they're handling the context window in a manner that's not transparent to the user. I agree they could potentially use a lazyload and only load what's visible to the user, but without seeing their infrastructure, there might be a reason they're doing it this way, but it is annoying.

hard dew
#

SURELY one day it WILL BE ACTUALLY fixed though, surely

ancient sage
#

Hopefully.

spice star
#

I'm sure openai is putting out the millions of fires from rolling out the new UI and GPT-4-Turbo and GPTs

#

just takes time

sturdy cave
#

Does anyone else have an issue with trying to use the full screen? whenever I back out of it I cant access anything else

hard dew
#

Or their civil war

spice star
#

either they roll out the features slowly and people complain that they're not getting the features

cloud verge
#

Hello

spice star
#

or they roll them out immediately and people complain that they're buggy

hard dew
#

And speaking of rolling out features, it's funny you say that cause they used to have this pretty pog feature where you could highlight any given part of the bot's output, and reply to that specifically, actual feature for that

#

which conveniently got removed exactly as this lag bs started happening

sturdy cave
sturdy cave
sturdy cave
hard dew
warm gull
ancient sage
sturdy cave
#

Why is this lag issue happening now??

hard dew
#

No idea lol, ClosedAI haven't said a word

sturdy cave
ancient sage
#

I get you're upset (and rightfully so) but insulting them isn't going to really make them work on it faster

hard dew
ruby flax
sturdy cave
#

Is this issue also on mobile?

ancient sage
#

But perception is reality.

hard dew
frozen nimbus
#

agi wont be free

sturdy cave
#

Everytime I leave the tab and go back to it, it starts lagging again

frozen nimbus
worldly elk
#

is it normal for custom GPT system to struggle with 16mb knowledge base?

sturdy cave
worldly elk
sturdy cave
worldly elk
#

now it is seperated to 5 filesd

sturdy cave
worldly elk
#

thanks mate

#

thought it would be lag

sturdy cave
#

Hopefully once they fix these lag issues you will be able to use the GPT fine

tight island
#

its all about love, between user and gpt, be nice to gpt and gpt gives goood answers
👁️ 👄 👁️

frozen nimbus
tribal anchor
#

does anyone know how to fix the freezing that happens when posting a picture or ss to chatgpt?

dense halo
#

Ayo can i use chatgpt on discord? Which channel?

dry wing
#

Family is nothing without it's people's

hard dew
#

Literally a lose-lose situation

frozen nimbus
tribal anchor
#

does anyone know how to fix the freezing that happens when posting a picture or ss to chatgpt?

hard dew
sturdy cave
hard dew
soft turtle
#

Refreshing only works temporarily for me (maybe 5 minutes?)

sturdy cave
#

they havent said a word about any of this and I have exams to study for

hallow hinge
#

Anyone knows if there is way to fix error of history chat dosnt loading and when typing a prompt getting "Hmm...something seems to have gone wrong."?

rich surge
#

It's just a bug going around. It'll be fixed soon.

hallow hinge
lyric stream
#

I hate how fast i run out of GPT4 prompt lol

#

I fly through mine every hour

edgy flax
#

Where can I find a repository of some great public GPTs?

red crag
lyric stream
red crag
#

And no, it's not cause of the custom instructions (that is true that can burn some tokens)

#

It's literally broken

hazy cedar
#

How long does it usually take for the error

Hmm...something seems to have gone wrong.
to be resolved? It's been a few hours now.

ashen plume
#

Hello

visual nova
#

does having old conversations on the sidebar instead of deleting them cause memory usage to be higher?

cunning flame
#

is it my idea or gpt is slower than usual? like after having it open for 5 minutes it gets extremely laggy and I have to reload the chat every time

languid storm
#

Wow, my GPT-4 is slow as ever!

potent ridge
#

It is stil slow? I came here hoping it was better. Its been bad for nearly 12 hours 😭

unborn beacon
#

the unbearable lag, freezing and crashes in a desktop browser (firefox with GPT4) soo frustrating the past month or so, I dont know if its worth continuing to payat this point. such a shame because it was great for an assistant and tutor to my studies up until a couple months ago.

#

and its getting worse each week.

#

gonna try and submit a use request to Claude

wide fable
#

High CPU Web Workers usage of ChatGPT when not is use someone have a solution plz

glossy sparrow
#

Why is chatgpt+ very slow and technically limited at the moment?

dense halo
#

Ok thought it was just what I was trying to do. Good to know everyone else is having issues.

unborn beacon
#

its been bad for the majority of the past couple months it seems. today is very bad

#

NetworkError when attempting to fetch resource.

eternal rockBOT
#

attention   Attention  Off-topic chats have been cleared by a moderator.

unborn beacon
#

its pretty frustrating, jsut sharing the experience with others here. how is it off topic in the chatgpt discussion when its directly related to chatgpt.

#

or is this a chatgpt safespace?

#

probably due to a flood of the world population using it to write their papers in these final weeks of the semester.

#

does anyone know if they use different clusters for the general web browser and app use, versus the people accessing through API??

lapis egret
#

I'm STILL getting this error: Hmm...something seems to have gone wrong.

normal quarry
lapis egret
normal quarry
# lapis egret yes

disable it's web protection feature and then uninstall it, then everything will work again

lapis egret
normal quarry
#

be aware that the web protection feature will re-enable itself every now and then so you will need to disable it again when you get the error. NordVPN's web protection feature is buggy.

dense halo
#

don't they have an exception list or something?

lapis egret
normal quarry
dense halo
#

MalwareBytes is a good one for browsers

zinc valley
#

are there any gpts that could help me write a good sales script

lapis egret
#

in the java console on Safari

#

and also "Must use workspace account for this operation."

sturdy cave
#

when are they gonna fix these lag issues?

unborn beacon
#

It's been about two months, although it is worsening. The mobile app isn't as bad(android) but its not convenient to work with as the computer.

open sundial
#

welp

humble mulch
#

Hey everyone, I work at OpenAI and am following up about the posts in this channel yesterday and today noting that ChatGPT on the web seems to get super sluggish after using the page has been open for a little bit. Clearly not the experience we aim for, and I'm sorry you're running into it—it's not you, and we're working on pinpointing a fix as we speak.

For some context, last night we deployed improvements to our frontend app logic that we believed would solve this, but clearly that wasn't the full picture. (The silver lining is: once this part is fixed, last night's improvement will be a bonus performance win on top of it!) The issue looks to be isolated to our frontend web app, so this isn't about network performance or other parts of our stack, and it's definitely a bug: ChatGPT shouldn't be greedy with your CPUs. We're debugging it now.

To make it easier to follow along, we're using this thread in the bug reports channel for more updates: https://discord.com/channels/974519864045756446/1180634437722054666. Sorry again for the hassle in the first place, and I really appreciate how helpful everyone has been by sharing high quality reports this weekend.

dense halo
#

I really wish when you uploaded images and asked it to create a similar image it was a little more accurate. Obviously that would open the door for problems and deep fakes, but still.

versed spruce
#

i only see chatgpt alpha

#

and trying to do anything results in an error

lyric stream
#

best place to find custom GPTs?

shell sorrel
#

hi

humble mulch
#

Great question! Working on getting that added—I jumped in impromptu when I saw these reports coming across. But it's a good reminder to trust no statements without that badge, and to only communicate through OpenAI on official channels.

#

I'll go expedite adding the badge 🙂

unborn beacon
# humble mulch Hey everyone, I work at OpenAI and am following up about the posts in this chann...

Hi Michael, thanks for acknowledging the issue, and giving some feedback to all of us here in the community on the issue. Hopefully, it works well very soon, with finals week coming up I'm looking forward to using gpt4 as my tutor again, currently, its lags and freezes are a major inconvenience. It's difficult enough going back to college later in life while also working full time, but GPT4 has truly helped by being my go-to primary assistant and tutor through my studies this year.

humble mulch
dense halo
wintry gorge
#

Why doesn’t chat gpt work on my laptop

#

It says this site can’t provide a secure connection

#

Wat do I doooo?

dense halo
#

as Michael said, ChatGPT is being worked on to fix the performance issues. please be patient

wintry gorge
#

So everyone is also having to problem???

ivory iris
#

It looks like e is already an object so it doesn't need to be parsed

#

it's in the api.js file but I don't know if that's the same as the source file

humble mulch
# ivory iris I haven't done a lot of react debugging, but it looks like you have an infinite ...

Thank you! 'll have to ready a desk in the office for you soon 🙂 This Discord channel is great—appreciate all the help and input from everyone here.

We've been able to reproduce the problem and have a fix deploying right now. To help keep everyone on the same page, I'll post an update in the bug-reports thread at https://discord.com/channels/974519864045756446/1180634437722054666 when that's live.

ivory iris
tough sapphire
#

i want to updrage my chatgpt account but my account is in waitlist from last month so how can i upgrade it

normal quarry
placid cradle
#

Is it just me or is anyone else kind of surprised that there is no Apple Apple Watch complication yet?? I like that they made the lock screen widgets, but I just feel like it would be the most killer feature to be able to tap a complication on the Apple Watch. It immediately starts listening to what you’re saying, and on a tap transcribes with whisper and responds to you in voice…. And/or perhaps a complication that just launches a voice conversation right there on the watch. (assuming iPhone is in Bluetooth range.) This is the main missing feature to me

spice star
#

there's no apple watch app to have a complication unfortunately

#

I do hope we get a chatgpt voice app on the apple watch at some point though

dense halo
acoustic kite
#

Hi I'm new here. My Chatgpt plus account seems to be not working. cant access chat history, and when i start a new chat, it says "Hmm...something seems to have gone wrong." and "There was an error generating a response" at the bottom
can't find any solution at the FAQ section

sweet gorge
#

is there a fix for the audio to use the main speaker instead of the ear speakers

ivory warren
#

would there be a way to add points on a map with chatgpt?

keen cove
#

hey guys

hallow hinge
ornate zealot
#

I'm sorry, but as an AI developed by OpenAI, I don't have the capability to translate text from images directly. However, if you can provide the text in a message, I would be happy to help translate it for you. Why wouldn't be able to do this? Kinda defeats the purpose of Vision in many use cases.

#

😐 Now it is doing it.

#

Nevermind, this thing is tricking me.

#

Ok, this thing can not translate Korean from images.

void quiver
#

For how long have u guys been in the waitlist. currently I been waiting for 2 weeks

ornate zealot
#

I guess it is trying to use code interpreter to analyze the text in the image, but it is telling me, "It seems there was an error with the OCR process because the Korean language data files are not available in this environment."

normal crystal
normal crystal
dreamy flare
#

I am actually impressed that GPT 4 knows all the in Germany listed BKs (Work associated illnesses). Saved me a bunch of work derpcat

dreamy flare
hallow hinge
void quiver
dreamy flare
hard dew
maiden fractal
#

And people have to join a waitlist

lavish hemlock
#

Sheesh, I'd be heartbroken if I didn't have ChatGPT Plus

dreamy flare
#

Rip

lavish hemlock
#

Playground is still available to them though right?

hushed scroll
#

Hello,
If I take the GPT4 subscription and what will I keep my conversations that I had in version 3?

And what will version 4 be less buggy?

lavish hemlock
#

Yes it will keep your conversations in ChatGPT 3.5

#

Less buggy? Meh, it's like rolling the dice lol

hushed scroll
lavish hemlock
#

Oh ChatGPT 4 doesn't crash that often for me

dreamy flare
#

Mate I have been useing it for so long i didnt even hear about the current waitlist openai_logo

lavish hemlock
#

And why are you calling us darling lol

dreamy flare
#

Fair enough

void quiver
stable bronze
#

now ADA can no longer set up an environment in its sandbox

#

why🥲😭😭😭

#

i can't do it in both default and GPTs of ADA

#

It seems there was an issue with extracting the JDK files due to permission restrictions in my environment. Unfortunately, I'm unable to install or run external software, including JDK packages, directly in this sandbox environment.

#

i remembered that it could before

dreamy flare
lavish hemlock
#

There is

dreamy flare
#

Probably some selling api keys like candy 💀

lavish hemlock
#

Just use Playground lol

spare anvil
#

UX question...do you guys think it's better that OpenAI let users save their chats manually (ie, click a saved button) instead of it just spitting every generated chat into the sidebar? It gets messy and cluttered as is, imho. To me, its kind of like photo storage....take a ton but never really go back and look at them later cuz there's just too much. I forget where the good ones are.

void quiver
loud night
#

Mornin, and happy Monday folks!

dreamy flare
#

Vpns exist

loud night
#

i think im the only one here who actually takes the time to do writing assignments by hand :(

dreamy flare
stable bronze
#

some vpn can do it

dreamy flare
#

I am pretty sure there is a way 💀

stable bronze
#

it depends

void quiver
loud night
# dreamy flare Why would i

That’s up to each person, for me it’s a form of expression since i usually can’t talk to people about things but i can express myself through writing without people assuming things, to each their own tho!

stable bronze
#

true. some kind of tagging system could be also very helpful

dreamy flare
loud night
#

if they make me pay for folders i will have a stroke

dreamy flare
#

You dont need GPT4 for that

loud night
#

lmao if there’s ever a openai brain chip if they made a external version I’d get that lmao

void quiver
loud night
#

oh. that’s.. youch

dreamy flare
void quiver
#

But I guess that only for gpt 4?

dreamy flare
loud night
void quiver
void quiver
loud night
dreamy flare
flat sierra
#

hey there, is anyone else having some weird reoccuring error pop up when using the ai?

void quiver
dreamy flare
#

If you try to compare an MRI with a small tumor, no 💀

void quiver
dreamy flare
#

GPT is not trained or suited for smth like that

void quiver
dense halo
#

Is GPT4 meant to be faster on new chats? I guess it’s because it doesn’t have to read through the previous messages for context?

dreamy flare
dense halo
sonic comet
boreal lynx
#

It won’t execute code for me rn

#

“There was an error executing the code, o can show you how to run it yourself though”

#

This is kind of ridiculous that I pay $20 a month for this and it doesn’t even give me the extra features I am paying for. How do I report bugs?

warm gull
#

its pointless reporting anything atm, you will wait weeks for a response ( if ever ). Not being negative - just speaking from experience.

maiden fractal
warm gull
#

i dont think anyone from OAI actually reads that tbf - and even if they do, we hear nothing about any of them being addressed, so you will never know. With that being said - @humble mulch seems to be a new ( friendly ) staff member around, so there may be hope yet

plucky timber
#

Not thru here though but through their own contacts

warm gull
#

Ive sent a mail to support two weeks ago - not a peep. I appreciate they must be very busy, but at least say something.

plucky timber
#

I can imagine they just did ctrl + a , del that week

warm gull
#

im sure all mail to that address go into some internal support queue ? Dont see why i need to resend it. I got the confirmation of receipt, so they must have it

hard dew
#

And the previous bug I reported along with some other issues regarding GPT-4, a few weeks or so ago, I didn't receive a word, but a few days after that I did receive a warning about some content policy violation with ChatGPT....???

#

So yeah, truly encouraging in reaching out to an OPEN company indeed.

halcyon widget
#

I used to subscribe chatgpt through ios ( apple pay ) and then after a month, i canceled my ChatGPT, but now, i want to renew it through web, will it appear " manage my subscription" or " manage my subscription in the chatgpt ios app" ?

feral zinc
unborn beacon