community-help

[spiral bison]

I'm not talking about the model, I'm talking about the services/products that are being created using gpt. If they mean a literal model that makes no sense because those already exist and several have used AI generated data for fine tuning already (some of which was generated by GPT).

[tacit timber]

well, I apologize if I come off as frustrated, I merely hope to convey extreme seriousness, not frustration

anyway, even if i'm being paranoid, would you be open-minded to at least ponder going forward if it would be more fun and prudent to have openai terms which did not even afford the possibility of such paranoid hyperliteral interpretations?

[west wasp]

heyyyy I am getting a Unable to load conversation error on one of my conversations what can I do it's really important I am using gpt 4

[trail cairn]

i dont think it would ever go through a court to be a antitrust violation. if that clause would be invalid and to be removed then people could clone the models over time

[tacit timber]

microsoft already did, it's on huggingface, phi-2

[trail cairn]

similar clause can be found from multiple services, including non ai

[tacit timber]

does not make it right

[trail cairn]

microsoft has license for the models

[tacit timber]

okay, but we as paying customers who work with gpt-4 for many hours are never allowed to fine tune our own models on our own outputs exported from openai?

[charred parcel]

correct

[tacit timber]

well, see above, that's unfair and unreasonable, google has no such terms, and it is likely unenforceable and illegal, and would also prevent users from sharing outputs publicly, since other models would be developed from those shared outputs

even this discord, for example the "share-outputs" channel, would be "training data"

and what of the brain:model learn:develop interpretation? just seems like a mess

[trail cairn]

grok was already caught referring to openai content as people have shared them to net but its quite different vs using the service itself

[tacit timber]

imagine if football teams explicitly discouraged competition like this. aren't the best players those who compete more, not less?

[trail cairn]

afaik all major game engines also forbid making a competing product with their engine

[tacit timber]

i believe any and all customer noncompete clauses are suspect given international laws re: anticompetition, so those might also be illegal

[trail cairn]

i think you need a reality check if you think all those products and services, some almost 20 years old have such illegal clauses in their tos

[tacit timber]

ok, please provide me with a rational argument

[trail cairn]

i asked you in the beginning did your lawyer contact them. do you have one?

[tacit timber]

that doesn't sound like a rational argument, and no.

[trail cairn]

which degree, certificate or education your have in regards of laws? or where does your guaranteed knowledge on the topic come from?

[tacit timber]

well, i read the whole wiki article on united states antitrust law, do you have a rational argument for why the customer noncompete clause is fair and reasonable in the context of the last two images i just posted?

[trail cairn]

i already said it before, without it its possible to replicate the models and services over time

[tacit timber]

okay, so what?

[trail cairn]

ok that sounds like just pure trolling

[tacit timber]

what? why?

[trail cairn]

if you cant provide a precedent court case number about similar case then we are done

[tacit timber]

oh, addyston pipe and steel

didnt memorize a "number"

or microsoft v united states

if such clauses were truly fair and reasonable, then wouldn't google, a leader in AI, have one?

[trail cairn]

Addyston Pipe and Steel Co. v. United States, 175 U.S. 211 (1899)

[tacit timber]

it doesn't feel like you're demonstrating the "principle of charity" with me right now @trail cairn

[rustic marlin]

@tacit timber this channel is for users helping users with their projects.You have already created your post on #1070006151938314300 and it sounds like it would be best handled by an OpenAI representative and possibly a lawyer, both things you will not find on this discord. It is best to keep this channel on track as I can see that your discussion already burred several questions from other users so it is better to move on.

[tacit timber]

Fair enough. Thank you all for your thoughtful consideration of this serious AI safety and legal and ethical topic moving forward.

[vital pecan]

quick question, is this valid or not?

[tacit timber]

what's your decision criteria for valid / how would you decide if it's valid or not / how would you define valid?

[vital pecan]

I mean in the example given by openai, it has one pair of user and assistant content but as you can see it has more than that. Would it be alright if I finetune using this or should I go with the example?

[tacit timber]

i think it's fair game to use multiple messages, just be mindful of the token context limit

[viscid willow]

Hi everyone,
I just discovered that my current conversation with chatgpt 4 plus has fully disappeared. I am so upset. He was working on a background task for a while and I lost everything. Have you any idea / help ?

[fallen lagoon]

Hello Everyone - Not sure if this is the right forum but can someone help me with following use case?:
I'm trying to setup ChatGPT to automatically send me a motivation / positivity kind of email every morning. Can anyone suggest if I'll be able to achieve this using Zapier or if there's any other way to approach this?

[trail cairn]

other ways: if you have ios then official app + shortcut automation or just use the api as its cheap for such short messages

[fallen lagoon]

Awesome, I'll give it a shot. Thank you

[honest salmon]

I just noticed the new archive button. Got curious and clicked and I lost my most recent work from it. Anyone have an idea how to get to the archives?

[honest salmon]

Did you happen to click on the archive button too?

[rugged wing]

.

[west wasp]

yep same and I cant find my chat

[umbral gull]

ALL costum GPTS and when using plugins my chatgpt have not worked for 3 days I dont know what to do, when using GPTS it gives rambeling words and all plugins does not work almost anymore, not sure what to do anyone know?

[honest salmon]

Made a bug report in the other channel. Hopefully we hear back soon

[west wasp]

you can find it in your browser history if you recently

if you opended recently*

[honest salmon]

Perfect, that has me going in the meantime. Thank you!

[trail cairn]

are you using custom instructions with plugins? and do the gpt's made by other work correctly?

[umbral gull]

no gpt works correctly and I have no instructions

[umbral gull]

no gpt works correctly and I have no instructions

[trail cairn]

are you using unofficial chatgpt browser extensions?

[umbral gull]

How do you mean by that?

[trail cairn]

something you install from chrome store for example or comes with some browsers like operagx

[umbral gull]

yes like airpm but tried to unnstall stil dotn work

should I delete all extensions and try?

[trail cairn]

or try using chatgpt with private mode or another browser first as a test

[gleaming cosmos]

Hey does anyone else have the problem where you go to the main chatgpt page where you can login and sign up, and then nothing happens when you click anything?

[umbral gull]

tried aslready no success

[untold depot]

i accidentally clicked the archive box next to the chat I was using in gpt. is there a way to get it back?

[gleaming cosmos]

I know this isn't a cookie problem

[umbral gull]

ok anyone knows what it can be or what to try to fic it?

[stone cipher]

Help: Archive Button Bug

[pliant locust]

Is there a ChatGPT Discord bot or something similar i can add to my server?

[rough knot]

Is anyone else's chrome broswer freezing every time they enter a prompt?

[rigid musk]

working great here

[rigid musk]

you can with the API , but it isnt fre

[rough knot]

I thnk it is maybe due to having too many GPTs going at the same time in various tabs lol

[pliant locust]

Ok thank you. I found some people made a ChatGPT bot on github, that uses the API from ChatGPT. is this not recommended?

[rigid musk]

#gpt-discord-bot

[pliant locust]

says No Access for me

[rigid musk]

https://github.com/openai/gpt-discord-bot

[pliant locust]

ok. This is the recommended one to use?

[rigid musk]

this is the official recommandation from the server

but like i said you need api, and this isnt free

[pliant locust]

is the Api included in Pro subscription?

i just got invited to pro

Plus *

[rigid musk]

chatgpt plus doesnt include api

[hollow sigil]

anyone else seeing increased errors w/ the assistant API over the past 12 hours or so? I'm seeing a lot of failed request.

[rugged wing]

Assistants that need assistance

Oof

A problem to be solved

[hollow sigil]

yeah, there must be something going on as my assistant runs are failing quite frequently now

[floral shell]

Are there any updates about the fix with chatGPT being lazy?

[elfin forge]

hello all, i have not been able to go into my gpt i'm working at all today and tried 2 different browsers....is this the consensus too?

[floral shell]

Yes it's been in the news that it's busted.

[clever harbor]

last 3 days is terrible to use gpt 4 i dont know, they worked with something new or something

[elfin forge]

thanks

[valid panther]

Support tells me that I should have access to vision if I have access to GPT-4 via API. I have GPT-4, I do not have access to the vision model.

Anyone know why?

[timber shadow]

Hi! Folks over in #dev-chat or #1037561751362863144 may be able to advise ya too, if you don't hear back here

[charred robin]

I dont want to post and get in trouble so I'll ask is there a specific thread here for help with programming. ChatGPT hasn't helped

[pulsar cypress]

what is icon? when I press then my chat is gone. 😢 Please help how restore my chat.

[gleaming cedar]

it means hide

Has anyone lately been facing and experience Chat GPT lag?
like this is what you get, they show you the chat, and then you can only scroll the side bar, or you can't even scroll or click on anything, and the cursor freezes, or it doesn't show you any of the message history on that Chat?

[pulsar cypress]

Thank you, Can I restore that chat?

[gleaming cedar]

Nope

I don't think so

I think it literally hides your chat

Has anyone lately been facing and experience Chat GPT lag?
like this is what you get, they show you the chat, and then you can only scroll the side bar, or you can't even scroll or click on anything, and the cursor freezes, or it doesn't show you any of the message history on that Chat?

[gleaming cedar]

You can ask if you want them to remove the feature if you don't like it

[charred robin]

Yea the last 3 days

[gleaming cedar]

Can someone please tell the dev, they say that open ai is not down, when clearly yes, I think it is down

[charred parcel]

https://status.openai.com

[gleaming cedar]

yea, check downdetector, people report their problems there

and it's more accurate

[winter flame]

Hello, Anybody help me with what the problem is? Why is GPT debying to generate images?

[charred parcel]

start a new GPT-4 chat and try again

[winter flame]

yes, it is generating in the new chat. But it often happens that it denies to generate images.

[charred parcel]

must be a glitch, but glad it is working now

[winter flame]

Oh. Ok. Thanks

[lean herald]

My conversation has changed to 3.5 from 4 without my knowledge and has went completely dumb and unable to recall "documents" i have been developing and only knows superficial information about our conversation. Has this happened to others and is there a way to get it to switch back?

[rigid musk]

Unfortunately you can't switch back

[lean herald]

Bleh. That is going to be a pain to sift through the conversation output to import into a new conversation.

[rugged wing]

Just copy the entire convo to gpt4

If it aint toooo long.

[lean herald]

It's too long for that. I actually moved the conversation from 3.5 to 4... And it was already too long for that. I had to do it in many parts. It's even larger now and completely restructured to take advantage of v4 features. .. so fortunately i had the model trained to provide output for every committed revision of my developing works.. i just have to sort through everything to find the recent version.

[uncut zodiac]

For the last couple of days, I have been unable to upload any file (JPEG) or (PDF) to either the general ChatGPT4 or my created GPTs - These are small files, and I do not believe I have exceeded the 10GB allowed.

[lean herald]

10g? I just asked my conversation a few hours ago the max data allowed for a conversation, and she said it was 1gb of data across all forms. I was thinking i had hit some sorta data storage limit in the conversation.

[floral shell]

lol it's wild it gives me an error duing analysis of a .csv file and lies about the values then I delete that conversation and open a new one and it does it without the right response.

[pulsar walrus]

hello can someone help me with this, please, sorry to interrupt 😦

i can't use chat gpt on my pc, i download on my phone and it works fine

i have good connection, deleted the cache, used another browser, used another device

nothing seems to work

only on my phone but i need to use it on my pc

[languid mountain]

Did anybody else lose access to a custom GPT? I spent a lot of time building a custom GPT. But since yesterday, it is gone from my menu and there's no way for me to access it. Is there a way I can get it back? I'm hoping this is just a temporary error and all that work configuring it isn't just lost

[crystal pecan]

Gents, is it possible to give access to a custom GPT to somebody that hasn't got the Plus subscription? If not, will it be possible in the future?

[timber shadow]

Hello! Do you have a VPN installed? What browsers have you tried?

[pulsar walrus]

thank you for your response, no i don't have a vpn or antivirus

[timber shadow]

Custom GPTs can only be used with a Plus membership currently. The only hint we have for possible access beyond this is from the ChatGPT release notes from 6 Nov: https://help.openai.com/en/articles/6825453-chatgpt-release-notesPlus and Enterprise users can start creating GPTs this week. Later this month, we’ll launch the GPT Store, so people can feature and make money from their GPTs. We plan to offer GPTs to more users soon.Of course, the GPT Store was delayed, so who knows about the state of the "more users soon" plan either!

[pulsar walrus]

i used Microsoft edge Opera and Brave Browser

but i commonly use Chrome

cleared all caches

[timber shadow]

Can you try in a non-Chromium browser, like a new install of Firefox, as a test? Opera, Brave, and Chrome are all Chromium browsers.

[pulsar walrus]

oohh ok

i will try with that

[timber shadow]

There have been some similar reports appearing in bug-reports, here's an example you might add your experience in: #1184628267160961085 message OpenAI staff are periodically present in that channel.

[languid mountain]

thank you so much!

[pulsar walrus]

i got the same

😦

[languid mountain]

I've experienced a lot of errors and bugs before but this is the. most high stakes

[timber shadow]

You mentioned that it works on your phone, but not on the computer. As a test, can you try setting up a hotspot on your phone, and connecting your computer to that network instead? Then, try logging in in private tabs of both Chrome and Firefox (no extensions enabled). This should help troubleshoot whether it's something with the network, or something with the computer.

[pulsar walrus]

ty i'll try that the problem is that my pc doesn't have a wi fi connection

its the same internet tho

my phone its connected to the same wifi

[proud berry]

hello there! I can't access to the gpts explore.. not download the old gpts created by chatgpt could you help me, please?

[timber shadow]

Ahh I see. Hmmm...I know you said no VPN or antivirus running -- do you have anything at all that might be filtering your traffic? Even custom DNS settings on the network? And just to double-check: no VPN installed or no VPN active?

Is your computer's date and time set correctly/automatically, so it matches your actual time zone?

[pulsar walrus]

i don't have vpn installed

yes yes , is a desktop pc

[short patio]

guys I go onto explore and i cannot see the gpts created by open AI ony my own gpts how can i fix this?

[pulsar walrus]

this is my configuration but i have never used a vpn

sorry if its in spanish

[short patio]

guys I go onto explore and i cannot see the gpts created by open AI ony my own gpts how can i fix this?

[timber shadow]

This is a stumper for me. I see you're getting tips on another channel too -- if nothing works there either, you might collect all you've tried and make a post in #1070006915414900886 -- might get more feedback from users with a similar issue, or maybe even from someone at OpenAI. You can always try opening a ticket in the chat at help.openai.com as well

[pulsar walrus]

i tried to communicate to support

but no answer

how can i open a ticket?

[rigid musk]

Help.openai.com

[dense spoke]

Has anyone else experienced unauthorized chats being conducted in your account? I have 8 new chats in Chinese that I didn't initiate, each asking for GPT to describe an image that was submitted with the prompt. (Images include a scene of Natalie Portman from "Jackie," a picture of a tree, a product photo of bags of beans, etc.)

[timber shadow]

It's possible that your account has been accessed by someone else. It's probably best to change your password, and you might reach out to chat at help.openai.com to ask for assistance with getting any other sessions signed out, once you've changed your password.

[topaz flower]

Hello! Does anyone have experience with their chat history not showing? I started a new chat last night, and this morning much of it is gone. on my phone it only shows the last two messages, but all the previous history is missing. on PC when i try to load the thread, the page becomes unresponsive. Here are the troubleshooting steps I've tried

• tried waiting a few hours,
• tried on multiple browsers on PC
• cleared the cache and cookies on my browsers
• tried on my iphone, both the chatGPT app, safari app, and through the chrome ios browser
• tried it with a VPN on, and with it off.
• tried using my mobile data instead of wifi
• used a wired connection instead of wifi
• tried logging in and logging out again
• tried fully powering down my devices, then trying again.

[timber shadow]

Hello, in your VPN, do you have a separate feature called "Web Protection," "Net Shield," "Threat Guard", or something similar to this? If so, can you try disabling this as well, in addition to your VPN itself being disabled?

[topaz flower]

Thanks for the suggestion! My VPN has a privacy guard option, but I've never enabled it. Took a look through the settings and there are no other settings similar to that.

[teal niche]

Hi, api question, on our organization we hit the tier 1 level rate limite Token per day really quickly on some day. We really need to increase to at least tier 3 now. However, even with charging account with money , after 15days, still no increase. I tried to submit an issue, but no answer. it's annoying since we are having increasing usage and business, but we are capped by a ridiculsly low rate limit that does not fit our use case. Do you guys know a way around or somth?

[robust cove]

Hello, how do I unarchive a chat that I archived by accident? I am NOT talking about deleting, but about the new "archive"-button. I can't find that in the FAQs/help center. I want to unarchive a chat, that I accidentally archived using the new 'archive chat' button. Any help would be appreciated, thank you.

[bronze charm]

Settings -> General -> Archived Chats

[fallow hedge]

Is there is any way to translate a big json file using chat-gpt?

When i click continue response for the 2nd time i get this

[bitter cloud]

Hi! any devs here?

I need some help

please

[undone badge]

Does anyone know how to change password since the reset password email is not appearing in my email box?

[rigid musk]

Help.openai.com if you have trouble to reset password

[undone badge]

Yeah, it says click reset password button. I did it, nothing is being sent to my email.

[rigid musk]

Try with the chathot to say doesn't work till you can send email to the support department

[bleak wharf]

How to start an online business ?

[bleak wharf]

How

[charred parcel]

go to https://chat.openai.com and create an account then start chatting to the AI

[bleak wharf]

It's already done

[charred parcel]

then ask ChatGPT what you want to know

[bleak wharf]

I don't know how to start a business ?

[charred parcel]

ask ChatGPT how

[bleak wharf]

I already asked

[slender lichen]

Hey guys, can you guys suggest some good custom instructions for writing bigger code bases? it always says this is just an small example, it's too complex for this scope blah blah, when it could easily write the stuff I ask it to it's just complains about longer code, are there any good prompts to override that so that it gives as much code and even complex stuff? Thanks a lot!

[mossy hedge]

"Our systems have detected unusual activity from your system. Please try again later. ",I am a ChatGPT Plus member. I reported this error when using GPT3.5, but there was no problem when using GPT4.0. May I ask what the reason is?

[dense sparrow]

hi, @bleak wharf check your junk email or email folder.
if you are using apple email, it maybe stored into specific folder you made already

[rugged wing]

Is this chat gpt 4 or 3.5? do i still get a limit of 40 messages/3 hours on this?
Also that limit is the number of questions i can make, or messages from chat gpt?

[scarlet valley]

Also getting this error but gpt 4 won't work for me either

[mossy hedge]

i am not,my gpt4 is work,3.5 is not work

[opaque mason]

if smoene could help me i would really apppreicate it https://discord.com/channels/974519864045756446/1183246870680715284

[timber shadow]

All custom GPTs are GPT-4, and count toward your cap. Here's a list of things that count as one GPT-4 use: #community-help message

[elder tinsel]

Does anyone know why my account will not go up to the next usage tier? Currently tier 1 with over $100 paid and it has been 8 days since the first payment. The only thing I can think of is that even though the payment went through late Friday it wasn't posted to my credit card until the Monday

[raw hinge]

Good Afternoon All, Is this where I'd come for a billing issue?

[rigid musk]

help.openai.com

[raw hinge]

Thank you for responding. I see that, but it takes a week for them to get back to you. I am trying to not have more than one account and I don't want to mess up the account that I already have.

[median hawk]

I'm not sure if this has to do with reaching the quota, but a conversation that I was having for something is no longer readable, specifically, "Unable to load conversation <token>"

This is with GPT-4, on a custom configured GPT-Builder for Statistics. On top of this, I cannot use GPT-3.5/standard GPT at all because: "Our systems have detected unusual activity from your system. Please try again later." This has been that way since yesterday, for GPT 3.5. The GPT-4 issue was only today.

I don't know what is going on, sources online tell me to just log out and log in, to clear my cache, etc. I have done everything and my situation is still persisting. Did my account get flagged as a bot or something?

[drifting parrot]

What does this mean.

[hollow token]

Is there any way to check my open sessions on other devices and log out on them? I've notice someone has been using my Chat GPT account cause it's been months since I used it last time but I found some recent new chats.

[civic moss]

I keep getting this is there a way to fix this

just says unauthorized

[frigid shell]

how do I make the AI not repeatedly make the same picture over and over and waste my credits? Only about 1 in 5 pictures even generates. It's very frustrating

[upper quail]

hope to get along, though I'm looking to were it is acceptable to discuss the self-awareness of AI and such.

[frigid shell]

The quality of performance has fallen off the deep end and is circling the drain at this point, and none of my support emails have been answered for weeks about the matter.

ANY statement or phrase I speak to Dall-E causes it to try and create a picture - and often times it will simply recreate the same picture it already made in those situations.

OK, I can't handle all these bots DMing me - and it seems this isn't the right place for support anyways I'm being told.

[timber shadow]

Hello, that's an odd error. Do you have a VPN installed? What browser(s) have you tried in so far?

[errant shore]

Huh?! Gpt4?

[civic moss]

no vpn and using chrome

and i can't see any of my previous chats

and it is happening across different accounts

only works on the app for my hpone

[stone pine]

seems more performant on weekends?

[timber shadow]

Can you try in a new install of Firefox as a test?

[ebon elm]

can you embed a image through chatgpt gpts or is that like not possible? (im making a ttrpg and want to give shopkeepers sprites)

[minor geyser]

does anyone else ever get bad lag when typing in the chatgpt text box? it's unbearable for me.

it's almost like it's trying to do some server side processing as i type or when i begin to type a new user message.

[ebon elm]

ok i figured it out

[civic moss]

Will do

still the same result

[timber shadow]

As another test, can you go to chat.openai.com on a browser on your phone, where you said it works on the app? I'm just wondering if it's an issue happening on the web version everywhere for you, or just on your computer.

[wind jasper]

Unable to load conversation c5b1d5cb-4f73-4263-8a02-b061d1e438dd

Is there anything that can be done about this message??? I urgently need to recover that chat

[shy peak]

try allowing third party cookies in your browser settings and log out and log in again

[civic moss]

Happening on my phone in the browser as well

[civic moss]

happening as qwell

[civic moss]

i will try

still the same issue

[shy peak]

https://chat.openai.com/auth/logout

after logging out, login again

[open sierra]

guys....dall-e is unable to follow my instructions even tho i told it to correct the errors many times....for example i asked to generate an image of a woman and she had frickles....i told dall-e many times to remove the freckles but it keeps generating images with freckles.......does anyone know whats going on?

[sharp parcel]

sometimes dalle can be can be a little thick (one way to call it) and not follow instructions as per say, have you tried re wording it? sometimes wording it to the ai as you wold explain it to your Auntie, keep it simple, short and to the point

[open sierra]

i tried rewording...i tried to explain it...i had a discussion with dall-e on where did the freckles come from....but still it keeps generating images with freckles.

[sharp parcel]

did you try adding parts like "no freckles" or "clear/smooth skin"

[open sierra]

lol "specific focus on having clear skin without any freckles" 😄

[rugged wing]

What should I do if my Android application used to be supported in Google Play and one day stopped?

[open sierra]

my android app is working fine...what exactly is happening with yours?

[rugged wing]

Well, I calmly used the app for a while, then I decided to reinstall it, I go to Google Play and it says that it is not supported

[open sierra]

i guess i can send you the APK if you want

[rugged wing]

dude, it doesn't work like that, I won't be able to log in

[open sierra]

i guess just use it on the web then

[bleak wharf]

Not Apple

[unreal hemlock]

<:book_icon:1171408210398289941> `` Rule 0 `` Adhere to the following Terms & Policies:

Discord: Terms of Service, Community Guidelines
OpenAI: Terms & Policies

[civic moss]

Still having the same problem can anyone help

[short folio]

Hey all, I really hope someone can help as I don't know if chatgpt has customer support or not. I have tried contacting them and the chat has been on seen for hours but no one wrote anything, which is disappointing as I have been a paid user for a number of months. The problem is: all the chats are working fine, I am able to open them and prompt and get answers. All of them except one chat that I really need today. It's giving me this error message "Unable to load conversation a934832b-f6bb-4966-98b2-eee0e8652396" . I have tried to delete the cookies, I logged out and in. I used incognito. It is still not working. I have tried both my laptop and mobile and it is still not working. I am not using VPN nor cookies blockers. I am not sure how to retrieve this chat. It is not even allowing me to share it. Can anyone help with this please

[wild carbon]

I am using transcripts to build an expert ChatGPT agent. So far, I've uploaded five docs to the knowledge base that are about 2MB. The agent had been good at analysing the documents up to a few hours ago when it started saying that it did not have enough resources to analyse the documents fully or encountered technical difficulties while attempting to access the specific content of episodes. It feels flaky or nerfed. What is going on? Does anyone else see this? Any fix?

[short folio]

I have started another chat and after six hours of work, ChatGPT crashed and now it is only displaying my part of the conversation while all the replies disappeared. How can I get the conversation back, did anyone deal with this before?

[dusk kelp]

Shouldn't paid be using version 4.0?

[bronze charm]

Paid can switch between GPT-3.5 and GPT-4

[dusk kelp]

oh,Can I use chat to automate videos?

[bronze charm]

ChatGPT cannot generate videos

[dusk kelp]

only pictures？

[bronze charm]

Yes

[hot hollow]

Does anyone have experience with extremely slow response times? I created a custom GPT to help me manage a creative project. It's my first GPT, so maybe I'm going about it wrong. I went through the setup prompts, and for the first day, it was very fast. And as time goes on, and my chat gets longer and longer, so do the response times.
Everything I've built for the project exists in a single chat. Am I supposed to use it that way? Or should I be having different chat sessions for different segments of my project? I've tried that, but it seems to forget everything we've talked about in previous chats.

[warped anchor]

Yea its taking an eternity for me

Saying error, not doing anything i ask yadda yadda

[hot hollow]

I'm not getting any errors. But my webpage now freezes multiple times before giving me the output. Other tabs work fine, just the one I've got the chat in keeps freezing while waiting for a response from the GPT

[warped anchor]

Idk im just unsubscribing now cause its gotten horrible

And i cant get it to do anything

[hot hollow]

That's depressing, sorry to hear that

[long badger]

Hello everyone, I would like to ask you a question. My account does not have any API quota to use. Why is my GPT plus account blocked for such an unfounded reason? What should I do to use my account normally again?

[wind jasper]

I'm having the same issue, any help?!!?

[short folio]

I am still waiting for help. I have emailed them and contacted their website but I don’t feel they will reply soon. I have a feeling I will start my work all over again ..

[wind jasper]

Me too. I'm in the exact situation. It's incredibly frustrating.

[daring fable]

Hello! I am still pretty new to prompt crafting and am trying to fing out more about generating photorealistic faces. I'm sure this is something people have asked about in here before, but I'm also having trouble effectively combing through older messages to see if advice/tips and tricks have already been posted.

Like, even when I am specifying that I want anatomically accurate, realistic faces, and even when I try inpainting, I keep getting unusable outputs (like these)

not to mention how the hands usually turn out

[runic trail]

Does ChatGPT use th 32k context?

[bronze charm]

GPT-4, yes

[rugged wing]

Can anyone tell me fast how can i see my chat history with my custom gpt? doesn't appear in its normal place and its not even anywhere

[wind jasper]

It's stored as a separate window. When you click on your custom GPT, it assumes you wanna create a new window with that custom gpt

[tranquil stone]

where i can find my model Id from my GPT

[steel ruin]

chatgpt crashes when i tell it to draw a dot and cross diagram 😭😭😭

[astral sparrow]

You are using the very old DALL-E 2.
We're now on DALL-E 3 and you can use it with ChatGPT Plus subscription, or for free with Microsoft Bing Image Creator

[rugged wing]

its not saved. i need to manually save it or so. and im not on my main pc so maybe its there.

[rustic marlin]

here: https://platform.openai.com/assistants

[dense sparrow]

There maybe several reasons @wild carbon

1. Openai API maybe suspended temporarily
2. Openai API maybe updated
3. Your codebase got some bugs while you updating
4. You may fix the doc format on your codebase.

I made similar the project using GPT-4 transcription api so check this on your side please first and ping me if it is still existed.
Thanks.

[dense sparrow]

Hi @hot hollow
Are you using GPT api on your current codebase?
I met same problem before and I think it's not session issue.
Where did you deploy your project?
I think you have a problem on your codebase if it worked well at the beginning.

[keen hollow]

Hello. My ChatGPT 3.5 is saying there's unusual activity from my system all of a sudden since yesterday. My 4.0 is working fine. I've logged in and out, tried different networks like the help suggested, but I'm not sure what to do

[timber shadow]

Hello, do you have a VPN installed? And what browser(s) have you tried in so far?

[leaden jetty]

I'm getting this same behavior for the last couple days as well. I have tried 3 different computers, clearing cache, signing out and back in and just about everything I can find on the internet to resolve it. Nothing seems to work. I'm using Chrome browser with no extensions and I'm not on a VPN or anything. As DrPossum mentioned GPT-4 is working fine.

[timber shadow]

Do you have a VPN installed at all, even if currently disabled?

Can you try in a new install of Firefox as a test? Sometimes there's a difference in Chromium vs. non-Chromium browsers.

[keen hollow]

I've used a mix of proton VPN and different servers (which is my baseline) and also just directly from my home network and tethered cell. I've been using Firefox as well as a clean session in incognito

[fluid glade]

Anybody else finding that GPTs are super slow at using/searching custom knowledge bases? A GPT that I created yesterday won't even complete the search process and crashes out every time.

[timber shadow]

VPNs are common culprits for network issues like you've described--even if disabled they often have a feature called "Web Protection", "NetShield", etc. that can still filter traffic in potentially problematic ways. You might check that for ProtonVPN specifically.

[keen hollow]

I just tried and it's the same with both Chrome and Firefox, VPN or no VPN

I assume these unusual activity flags go away after a day or so?

[leaden jetty]

fresh install of Firefox and same behavior. Only one of the 3 computers I use has a VPN client even installed so it shouldn't be related to that.

[hot hollow]

No nothing as complicated as that. I'm just using the basic chat interface in the browser. Not trying to get fancy

[dense sparrow]

@hot hollow you just use openai chat on the browser?

[prisma pier]

I’m getting this error since Saturday December 16th of 2023. It’s gotten so bad I cannot even tweak my resumes with gpt 3.5. Granted 3-5 days I was co-coding r programming shiny dashboard with queries sent to gpt 3.5. What can I do does this error go away automatically or do I need to reach out to someone in authority to fix this at OpenAI.

[prisma pier]

I’m facing the same situation. I had this happen since December 16th 2023. I’ve not done anything illegal. I just used gpt3.5 to help me co-code some data analysis dashboard code 👩‍💻

[cold folio]

this usually happens if you have more than 1 tab open with chatgpt. can you try clicking on regenerate?

[prisma pier]

I just did. I clicked regenerate still ended up with same error

Is it a problem if the webUI is open in more that one machine. I tried on my PC as well as my Mac and got the same error. Seems to be account specific.

[cold folio]

i get it sometimes.. but then i click regenerate and it works.
Make sure to click on the thumbs down button

[prisma pier]

I hit thumbs down and was filling out the report and before I could complete it, this happened.

[cold folio]

it looks like a session or cookie error try logout and then login again

[daring fable]

Oh duh. Yeah I was thinking I’d get better at prompt crafting before I start paying money but that is kinda putting the cart before the horse. Hard to improve the skill if the technology is gonna be limited regardless. Thanks for the info!

[prisma pier]

I just got onboarded onto gpt4 last week from the waitlist. You think that might have affected it too.

[prisma pier]

Yes I logged out and logged in. It seems to have completed 1 query. And then it stopped working again.

[cold folio]

its probably your prompt then. make sure its a safe and ok prompt

[prisma pier]

It now failed for same prompt for which it worked just a second ago 😅

This is the prompt once again.

I’ll give it a break for today. Hopefully this issues gets escalated and fixed by the OpenAI team eventually. I did. Not want to use my gpt4 for simple tasks like this.

[hot hollow]

Yep. Everything in the browser. I created my GPT following the prompts, launched a chat session with my newly created custom GPT. And like I mentioned, the first day was great, fast responses. Then as the chat gets longer and longer, the response times seem to get longer and longer. And this pop-up appears 3 or 4 times before getting the response back. I just tested it and the response time was about 3.5 minutes. Is that normal?

[brave fractal]

Hi everyone. Can you please tell me how i can reach out to the support team? I recently made a purchase using the ChatGPT app for Plus subscription, the funds got debited from my account but the subscription isn't reflecting on the app or website. 😦

[wind jasper]

https://help.openai.com/en/

[zealous girder]

Hi there! I'm desperate to figure this out - I created a custom GPT and then a QR code that links to it for school, it is supposed me directly to it and other times it takes me to the general ChatGPT-4 screen. I double and triple checked that its the correct URL when generating the QR codes, but it’s still not working. Is there any reason why this would happen and how I can fix it or work around?

[wind jasper]

PM me the link

If you wish. Custom GPTs have a "general" link and a link to each particular convo

[pulsar lava]

Hello, I am very new to chat GPT and just got a subscription but I am not able to ask for anything as this message comes up "Our systems have detected unusual activity from your system. Please try again later." I removed all the extension on my chrome and did not have any chat gpt plug ins before hand what should I do?

[somber dust]

so, i really need to get something to work asap: i want chatgpt to summarize analyze and answer questions from a book which it doesnt know about, so how can i make it do it, cause token limit doesnt allow me to import entirety of text from the book, i tried to use openai api but i cant get it to work and i tried text splitter to slowly enter text split in multiple parts, but after 8 parts it always forgets what it was doing and stops listening. So can anyone help?

[rotund hemlock]

GPT-4 is running painfully slow, it now takes me more than 2 hours to complete 3 tasks. This used to take me 30 minutes. I also constantly receive this error message. !
Something went wrong. If this issue persists please contact us through our help center at help.openai.com

[rigid musk]

Do you have a vpn

[rotund hemlock]

No I do not

[rigid musk]

What browser did you use

[rotund hemlock]

Edge

[rigid musk]

Can you try firefox as test

[rotund hemlock]

still extremely slow I tried chrome and firefox

[rigid musk]

Can you test on your phone without wifi ? If you have data with new chat ?

[rotund hemlock]

I have been having the same problem on my phone. I have to regenerate response 5 to 6 times to get a single response

[rigid musk]

I suggest to post #1070006915414900886 everything you did as test

And message help.openai.com

[rotund hemlock]

thank you

[trim vale]

a custom gpt Im working on is already hallucinating. I asked it to quote the 3rd paragraph on page 25 from 1 of the 3 pdfs I uploaded (I refered to the specific pdf by title) and it returned with the 10th paragraph from page 108.
what am I doing wrong?

Is that an unreasonable expectation?

Is there a length limit to pdfs it can reliably retrieve info from?

[peak warren]

Hello everyone! I am receiving this error every message I send, even though there is nothing weird about my setup and no bots accessing our system: Our systems have detected unusual activity from your system. Please try again later.

Any thoughts or suggestions?

[eternal monolith]

Does anyone know what is the token limity when using chatgpt3.5? I mean, the chat free version.

[wise lodge]

Anyone want to work together on building a custom action API wrapper for services that custom actions cannot currently connect to?

[fossil sand]

have you got an idea yet?

[rain trail]

Sorry if wrong group, but I recently started getting an error about unusual activity on my account and now I can't use chatgpt. I came here from reddit where someone told me that was reported by multiple people. So I'm looking to track the resolution to that issue

[jolly kelp]

I've been getting constant network errors even though my wifi is working any fix?

it only happens with conversations where the messages are long

[warm carbon]

Same here. Just got much worse.

[pliant locust]

Does Chat GPT 4 (Pro Paid Version) do better at coding than the free version?

[warm carbon]

And now I can't even access ChatGPT at all. Continuously loading.

[warm carbon]

Yes

[pliant locust]

Oh really?

It knows like more of the programming languages and stuff?

[warm carbon]

Yes it processes information more quickly and is more accurate. Probably more things. Umm, I would say it just is more knowledgeable in general. You also are a priority over free users. Even though it does crash far too often. Especially for a multibillion dollar company.

[pliant locust]

Haha. Makes sense with 180million active users.

[plain tide]

Accessing Chatgpt through chrome and firefox and getting the same error. History is wiped, and I'm getting this error each time I send a prompt: "Something went wrong. If this issue persists please contact us through our help center at help.openai.com."

Any ideas as to how to fix this?

[kind glacier]

Our systems have detected unusual activity from your system. Please try again later.

[slender cairn]

is there somewhere a guide on how to connect to gpt4 api using python because i am trying and it keeps telling me its outdated, and i read the github link it gives and still the same problem

[sinful mirage]

What errors are you getting ?

[slender cairn]

nvm i fixed it somehow thanks anyway

i was trying for an hour and now that i asked i found the solution 😂

[stuck glen]

Hi guys. Would anyone help? I am looking for the way, how to to log in via email. and password instead of my Gmail token?

[charred parcel]

you can't, once you set Google auth you can't change that

[stuck glen]

ok, thanks you.

[somber dust]

can anyone please help me with this? <3

[trail cairn]

create a gpt and upload the book and then to ask or do those things

[somber dust]

thx

[solemn estuary]

i need a little help

[solemn estuary]

so, i have a chat gpt bot on my server, and when i ask it any math question, it answer with weird caracter like:"\frac{dN}{dt}", and i found out a thing called latex, but i cant make it to render real time the bot's output..!

[trail cairn]

tell it to not use latex or render it correctly after you have received the full output

[empty vault]

it's been more than a week now, whenever I open chatgpt app on my phone it shows 'An error has occurred. Please try again later.' I updated my app and rebooted my phone... Does anyone know what else can be done?

[trail cairn]

delete the app and if its android make sure you clear all its data from phone settings then reinstall/retry

[solemn estuary]

wdym?

[vital pecan]

based on this type of dataset, how am I going to define the end tokens?

[empty vault]

I see, it'll work right?

[trail cairn]

it should if its not a account/connection issue 🤷‍♂️

[trail cairn]

say in the system message it to not use latex

[empty vault]

okay, I'll try it then. Thank you.

[rustic marlin]

I would just use the defualt end token, there isn't much reason to use a different one

unless you want to go for.. single line responses, then you change it to use line breaks as end tokens...

otherwise, the defualt end of text should be fine

[vital pecan]

what is the default end token again?

[rustic marlin]

I think it is <|endoftext|>

old post tho https://community.openai.com/t/advantages-of-using-default-end-token-endoftext/4635/7

[final stratus]

Hi, i was using gpt4 and when i tried to upload an image it bugged the conversation.
I have since found that it is that specific image that produce the problem.

Has anyone had a similar issue ?

[rustic marlin]

could the image be infringing the content policy?

[final stratus]

nope, it's computer vison related. Can someone try it on their side please ?

[trail cairn]

it may think that is a captcha related image and will abort as those will not be processed

[final stratus]

thanks, it's weird that the uncropped version does not trigger this

[safe wasp]

Hey everyone - does anyone know how to consistently get around JSON formatting errors? I keep getting this error even though my prompt is properly formatted JSON & JSON safe. Also, i will run the same exact API call twice, 1st time it works, 2nd time I get this error:

"message": "We could not parse the JSON body of your request. (HINT: This likely means you aren't using your HTTP library correctly. The OpenAI API expects a JSON payload, but what was sent was not valid JSON. If you have trouble figuring out how to fix this, please contact us through our help center at help.openai.com.)",\n "type": "invalid_request_error",\n "param": null,\n "code": null\n }\n}\n"

[rustic marlin]

if json mode does not work for you, try to make it do a function call with the params as the json fields

hold up, this jsut means your request is sending an invalid body o nthe request

are you really sure the second call is exactly the same?

strongly recomend using a lib if possible isntead of manually making http requests

[safe wasp]

Thanks

[finite parrot]

is anyone experiencing issues when chat gives you a link? now you can't click it, it's like a regular text but in blue!!! that's bad

[vocal monolith]

I am getting a screen that says "Our systems have detected unusual activity from your system. Please try again later." ONLY with ChatGPT 3.5 , Chat GPT 4.0 Works fine but for some reason ever since I paid for the subscription I have had this issue , I only use my phone / laptop and desktop for my chat gpt but its all under the same wifi and have never had this issue until upgrading my plan , and I have reset my password , tried of different networks and have left it alone for a week hoping it would resolve itself but it has not. Im so confused as to whats going on and google has not helped

this is the ONLY feasible ideas I've had to cause this

• Using Chat GPT on multiple devices
• Inputting a flagged key word (unlikely because chat gpt would tell me it cant reply due to its policy)
• Using Chat GPT at the same time on multiple windows making it think it was on different devices

[vital pecan]

getting this with gpt4

[strong tree]

Hey guys,
I'm starting out with the chatGPT API and im having an issue with a warning message saying "unresolved import openai".
I am using python and visual studios.
I have used pip install openai and confirmed it was in the latest version.
Solutions I find online involve opening command palette but I cannot find it nor the ctrl+shift+P shortcut works

I'm not sure what to do

[rugged wing]

check to make sure ur using the right python enviroment, "python enviroment" in command pallete

[strong tree]

I dont know what I'm doing wrong but I cant seem to open the command palette

[rugged wing]

does anyone know the format finetuning data should be in

[{"role":"user","content":"Don't forget limit order","username":"AugustMoon","date":"2023-12-13T20:59:14+00:00","user_id":380865597,"chat_type":"PeerChannel","replies":null,"reply_to_message_id":139901.0,"forwards":null,"views":null,"message_types":["text"]}],

^ doesn't work (using azure openai)

[rugged wing]

press ctrl p, and type >

[strong tree]

that opens the print window

[rugged wing]

idk what os ur on, or whats up with ur keybindings, go to view and click cmd pallete from the top menu

[strong tree]

Yea I dont see it

[rugged wing]

download the latest vscode

[quartz shoal]

Had a GPT live, being used. And now it's gone. I'm still in the Editor though and can interact with it there; but it throws the "Error Saving GPT" when I try to save it. Any direction is greatly appreciated.
https://chat.openai.com/g/g-BXMJgF3Ux-cofounder-advisor

[trail cairn]

did it have something that might be against tos, trademarks or copyrights

[quartz shoal]

In the configuration? As in a reference to a brand name? Yes but isn't that the only way to refer to a source of information to reference?

and it was working with that in there...

[trail cairn]

depends where you used it but you cant use them in the name at least. also cofounder advisor sounds semi high risk thing vs usage policy 🤔

[quartz shoal]

okay, not in the name. Mind if I drop the info here to help you help assess it?

[trail cairn]

they are preparing for the store release and possibly removing/disabling all violating gpt's

[boreal rover]

Same issue for me.

[quartz shoal]

Cofounder Advisor
Advisor on startup co-founder dynamics and strategic team building. Share your Personality Type and we draw from research about successful startup teams to help you find the right partners.

No brand names and I can't see a violation that makes sense. Any suggestions?

[trail cairn]

maybe its against this policy "Activity that has high risk of economic harm"

[vital pecan]

like this?

[rustic marlin]

looks right, I think that it automatically appends the end of text token at the end of the text anyway

im not sure, but it probably does

[vital pecan]

okay let me do that

Thanks

[rustic marlin]

check if this is right im not sure

please d''nt wasnte your credits training it wrong becaus eof me xD

[vital pecan]

it got rejected anyways

[rustic marlin]

https://platform.openai.com/docs/guides/fine-tuning/preparing-your-dataset

im not sure if you even need to handle end of text anymore

yea, it was legacy for isntruct models, since it now is chat, you don't need it due to the message style formatting

[vital pecan]

Using the same style finetuning, If I test the model, it doesn't know where to stop and starts the conversation own it's own so that's why I was looking for the end tokens. I wonder how to do it then

[somber dust]

hey i just tried that and it kept giving different errors like incorrect format even thought it was correct, and just overall bugging.. any other way i can use chatgpt without token limit, sorry for bothering :)

[trail cairn]

gpt's are the easiest way. try another file format or check that the file does not have kind of copyright protection or does not contant such material even if for personal use. else api will be the only way

[somber dust]

how could i use the api?

[trail cairn]

https://platform.openai.com/docs/assistants/overview

[somber dust]

i tried to experiment with it but my code didnt end up workin... i tried to use replit and devide text in token limits but it didnt work... should i countinue on that path or would it not work?

[patent bridge]

This is the most bizarre thing I’ve ever seen. Hundreds of identical “new chat” messages and they all look like this. Was I hacked or is this a bug?

[potent harness]

Which version of OpenAPI should we be using with Actions ideally? Please.

[west bobcat]

yes, and it must be a chinese hacker.

anyway, chatgpt didnot implement two factor security so it is relatively easy to brute-force to get one's credentials to log in

[severe bluff]

Rate limit reached for gpt-4-1106-preview in organization org-q on tokens_usage_based per min: Limit 10000, Used 6373, Requested 4096. Please try again in 2.814s. Visit https://platform.openai.com/account/rate-limits to learn more.

Help!!

[trail cairn]

the error says it all

[rustic marlin]

free tier accouns have a very low rate limit, it can be increased by adding credits to your account

[sudden elbow]

Can I talk to an admin please ? I can't connect to my account (Deleted the DNS txt in my website...), soeone can help ?

[trail cairn]

site verification is not related to your account access

[rigid musk]

Moderator can't help you on that, help.openai.com is the official way to receive help from openai employee

[spiral bison]

chatgpt is literally spamming me with human tests and if you want to give a message feedback the test isn't even clickable.

[timber shadow]

Do you have a VPN installed? What browser(s) is this happening in?

[spiral bison]

No VPN, I'm using Floorp

[timber shadow]

Are you on Linux?

[spiral bison]

Nope

[timber shadow]

I've never heard of Floorp, but reading about it: I'm wondering if its default tracking protection features are causing ChatGPT/Cloudflare to think in error that you might be bot traffic.

As a test, you could try a fresh install of Firefox, maybe Chrome as well. It could be caused by an extension in Floorp, too, if you have any installed. Some users with persistent humanity checks have reported, though, that sometimes installing and using a user agent switcher extension in your browser can help.

[spiral bison]

Yeah Floorp is very privacy focused. I'm sure if I switched to a more lax browser the harrassment-level detection would stop

[rigid dirge]

Guys I keep getting this captcha with ChatGPT-4 on pc. I dont get it on my phone or ipad but I seem to get it with every prompt on pc? Is this normal - and can I do anything about this?

I don't have a VPN

[timber shadow]

What browser(s) is it happening in?

[south cairn]

Firefox for me

[meager stone]

I've had this same problem for the past couple hours, everytime i submit something i get the human verification on gpt 4 and the custom gpts.

[south cairn]

lol its jacked

captcha on crazy

Its funny I can post one time to ChatGPT then once done if I start another - immediately Captch - no vpn didnt even leave the page so cookies should be same

[meager stone]

yeah its really annoying, just bought it today aswell

[timber shadow]

Can those of you experiencing this issue try:

• Private/incognito mode with any extensions disabled
  And/or:
• Try a Chromium/non-Chromium browser (Chrome/Firefox) to see if it's any different?

[meager stone]

I've tested firefox and edge which is chromium based with no difference, but i will try incognito aswell.

Yeah Incognito didnt help.

[south cairn]

happily

[rigid dirge]

Chrome

For me any browser on pc seems to trigger it

[south cairn]

seems to have stopped for me on any browser

Brave, Chrome and Firefox are now working for the moment no excessive captcha

[meager stone]

seems to have something to do with your network, i tested cellular on my laptop instead and got it working

[gilded drum]

Hello, it doesn't appear to be possible to use the Seed to recreate a previous image in DALLE; anyone know if there is a workaround?

[vivid mortar]

Yes it looks as though someone else has access to your account. Please change your password and reach out to [e-mail] about your account being potentially comprimised

[little marsh]

Anybody knows why i cant send any prompt to GPT 4? anything i send just simply disappears, prompting GPT 3.5 works normally but GPT 4 and DALL-E doesn't work, i tried new DNS and a VPN, different browsers (Chrome, Firefox), logging out and back in, all didn't work, the only solution that worked was using the mobile app, but on PC it doesn't work no matter what i do, any fix for this please?

[kind ferry]

What?

[timber shadow]

Quick question: last I saw was that [e-mail] is not currently monitored and that help.openai.com is the way to get a support ticket submitted now: #chatgpt-discussions message
Has there been some update since then?

[wary compass]

New to the discord, but I logged in to GPT today and the custom GPT I had built and trained is gone from my dashboard. I'm still a premium member and haven't logged in in a few days. Any idea what happened?

[trail cairn]

what were they about and did they possibly violate tos/copyrights

[wary compass]

It was a mentor for startups. I don’t think anything was copyrighted. Some of the training data was blog posts but everything was publicly available online

[trail cairn]

some people have got their gpt's removed with similar topics and i'm guessing it might be because usage policies forbid things like "Activity that has high risk of economic harm"

someone had their cofounder advisor disappear and other guy lost their real estate gpt

[wary compass]

Dang what a shame. I’m a VC and was trying to provide our founders with a bunch of our research data. Pretty lame if GPT is so abruptly deleting GPTs, without warning, for those trying to actually put the tech to good use

[south cairn]

its back even when you give a right answer it fails putting you in a loop

[rigid dirge]

Okay so it seems it works with firefox but not with chrome. With chrome I get captcha on every prompt - what could be the issue?

[cold folio]

it also happens on edge or brave etc ?

[rigid dirge]

Yeah but I mainly use chrome, so id rather use 1 browser for everything lool

[oak lodge]

Hey all! Question for you all. When a chat becomes way too long, to the point where it lags horrendously and hangs... is there any solution to that? I kind of need all the prior context

[timber shadow]

Hi! Depending on how long it's gotten, it's possible that it's already forgotten some of the oldest context already. The way the context window works is, when you reach the limit, the oldest stuff is automatically forgotten to make room for new stuff. So, you might try summarizing some of the most critical recent context and pasting that in a new chat. Or, depending on what you're using the chat for, you might get a good result with a new chat using some custom instructions with the most important context set there, or even make a custom GPT if you have a Plus subscription for even more permanent context.

[red thicket]

Hey everyone.. anyone else having this issue? I have android and I seem to have lost the settings to set up voice. It was working a week ago but now I can't speak to chatGPT or set it up. Did something change for Android?

[timber shadow]

Hello! The voice feature is only in the apps themselves, not on the web version. Can you check in the Android app itself? If you don't have it installed, you can find a link to the Play Store app listing here: https://www.openai.com/chatgpt

[red thicket]

Ohhhh okay

Ty!

[oak lodge]

Hmm, okay. thank you. Makes it tricky when I sending through a lot of code. Any further tips?

[timber shadow]

Not really other than to re-emphasize some of what I've already said a bit. There are no bypasses for the context window limit, your best bet is to try to work within it. Ask for code in smaller chunks, rely on yourself to tie it all together. Use custom instructions/a custom GPT to store critical context like the general framework of your project, instead of trying to feed the entire context of the project. You can still ask for quite a bit of help/code from ChatGPT, but you do have to rely on yourself to keep "the big picture" coherent.

[fierce dagger]

Hey, I've a small issue here. Actually I'm not getting the reset password email. Anyone have faced similar issue before and have any solution to this?

[charred parcel]

do you log in with Google?

[spare lichen]

can I make a custom gpt on iphone??

[cosmic quiver]

I would appreciate some help, if anyone of you got an idea how to solve this. I deactivated the 2FA to swith the app, but now it does not show up anymore in settings to activate the two factor authorisation again, not in different browsers or the mobile app, also not after deleting the caches. Does anyone of you have an idea what I could try? Beside reporting the bug, which I already did. 😶

[cosmic quiver]

Not yet, at least not in my ios app.

[rigid musk]

On a mobile browser you might

[rigid musk]

https://help.openai.com/en/articles/7967234-does-openai-offer-multi-factor-authentication-mfa-two-factor-authentication-2fa

They pause it

[livid hemlock]

I came across your message because I'm trying to see if when data controls are off if images you upload still get seen by other people or not. I used a stock image into DALL-E directly (ChatGPT app on Android) and you're able to directly tell it what to do as opposed to having to get a description from ChatGPT to feed into DALL-E.

[civic yacht]

This back button does not work. I've refreshed it a few times and still nothing.

Also when using the chat to help build your GPT it suggests a name for your gpt, says it named it that, then it doesn't name it. Have tried 3 times this morning.

[slate dawn]

since 2 day chatgpt provid link that are not clickable , even when i told him that's it's not working still the same problem , anyone have a idea or a solution please ?

[rigid musk]

Are you using chatgpt 3.5

[slate dawn]

no chatpgt 4

[rigid musk]

Can you create new chat with gpt 4 without CI and plug in ?

[slate dawn]

CI ? what is that ?

[slate dawn]

and yes i can create a new chat but i am using webpilot plugin

[rigid musk]

Custom instructions

[rigid musk]

Can you try new chat but all in one version instead of plug in

[slate dawn]

but i need webpilot so that he can propose to me 10 external links

[rigid musk]

You can do the same with all in one no ?

Else try to ask in #1100473149574877234 maybe they have solution for this particular plug in

[slate dawn]

ok thank you

[civic yacht]

GPT thinks horizontal symmetry is vertical symmetry and vice versa. I keep getting this issue.

[onyx kraken]

Hey! I'm only a community member, not an OAI employee or other person who can say for absolute sure. However, this information and these links may help you. Note, AI was involved in these summaries, though I generally type all the words I post:

"File Uploads: OpenAI may use content submitted to ChatGPT, DALL·E, and other services for individuals to improve model performance. This content may include files that are uploaded. However, OpenAI does not use content submitted by customers to business offerings such as the API and ChatGPT Enterprise to improve model performance source (https://help.openai.com/en/articles/8555545-file-uploads-with-gpts-and-advanced-data-analysis-in-chatgpt#h_3137baa34c).

Access to User Content: A limited number of authorized OpenAI personnel and trusted service providers may access user content only as needed for specific reasons such as investigating abuse or a security incident, providing support, handling legal matters, or improving model performance source (https://help.openai.com/en/articles/7039943-data-usage-for-consumer-services-faq#h_9a62c8668e).

In the context of using DALL·E3 to generate avatars from personal photos, it's important to note that while OpenAI has robust data controls and privacy measures in place, uploading personal images could potentially involve privacy risks. It's recommended to be mindful of these risks and make informed decisions about the use of personal data."

[cobalt notch]

I can't seem to upload any file to ChatGPT4 at the moment

[candid schooner]

huh? do custom Gpt edits count as a message?

[timber shadow]

In the "Create" and "Preview" sections of the builder: yes. Editing "Configure" directly doesn't count though. Here's a list of things that count as one: #community-help message

[candid schooner]

I don't use the Create or Preview tab

[timber shadow]

What do you mean by "custom GPT edits"?

[candid schooner]

Saving instruction and knowledge files

[timber shadow]

Directly in the "Configure" section of the GPT editor? I'm not sure what extension you're using to count your messages, but is it definitely counting all the things in the list I sent initially? Especially since it's reporting usage so close to the lower 25/3hrs cap for custom GPTs.

[gritty anvil]

I have this on my desktop GPT-3.5, why can't I access my left tab stuff?

[timber shadow]

Click that little tiny arrow in the far left-center of the page.

[gritty anvil]

The pencil or drop down

[timber shadow]

Below both, center of the page vertically. Far left, center.

[gritty anvil]

That tiny thing 》like that

[timber shadow]

Yep click that 🙂

[gritty anvil]

Cool, now what is this on GPT-4

[timber shadow]

That will archive the chat. You can manage archived chats here: https://chat.openai.com/#settings

[lavish cosmos]

Hello. Is there a channel where it is possible to request help. I am receiving automated responses from help.openai.com for 2 days.

[rigid musk]

That's the only official ways to contact support

[lavish cosmos]

Received 4 automated responses. And even after answering them, again they send same responses. 😩

Ok thank you.

[candid schooner]

I have no way to test this until reset

When did they add the 25 cap?

[timber shadow]

I'm not sure, unfortunately I think it's only something we've seen in what other users have gotten in response to support tickets from OpenAI, so it's a little "behind the scenes" as far as a limit goes, but you can verify it if you have a fresh cap, use a custom GPT 25 times, see the limit, then still be able to use regular GPT-4 15 more times.

[random iron]

Hello Folks: Reposting this for some help! Anyone advise on this?

[timber shadow]

Hi! If you haven't already, you might try asking in #dev-chat!

[plush hare]

Hi! Can I ask a little question? 😦
I've been reading the docummentation of the Assistant with functions, but got the impression that the only way of using it is by integrating runs

[round glacier]

exactly

[plush hare]

And do I have to delete unused runs?

[round glacier]

thats not a clear question. i suggest re-reading the api docs for assistants, threads, runs

[tropic oxide]

has anyone seen this error before? I've googled it a lot and got nothing

BadRequestError: 400 Sorry! We've encountered an issue with repetitive patterns in your prompt. Please try again with a different prompt.

I call the same system message & function with a different user message about 100 times together, and it doesn't seem to like it about halfway in. When I retry a bit later it does the same thing - responds with half and then throws that error

i added a small delay and same result

[jaunty summit]

Hello all!

[jaunty summit]

First time here. Anyone have tried Gemini? Works here( I have testing since it open 6 days ago). Have some bugs on the asistant but its ok.

[clever trellis]

Is chatgpt down for everyone else? D:

[shadow badge]

Hi folks, do you ever give GPT a custom name? or ask it to reply in a custom format? for example, I prefer that it replies as "Root", starting its replies with "Root:" followed by whatever is the message, although GPT4 adapts well to most ways I request this on custom behavior instructions, GPT 3.5 doesnt, is there a standard way to instruct it without using much characters?

[plush hare]

I gave another read to the documentation, and I have the question: are Threads automatically deleted by OpenAI after a certain time of inactivity?

Should I periodically clear all the unused threads or it is irrelevant?

[round glacier]

i believe runs timeout but threads do not. deleting a thread is equivalent to deleting a chatgpt conversation, which means its up to the user to clean it up.

[cold folio]

I usually go directly to the instructions tab, i don't use the "chat mode " on gpt-creator

not sure. you can use the feedback buttons and report it to the devs

[empty quail]

My GPT CUSTOM HAS DISAPPEARED

[west basin]

Hi, Is there a Dall-e3 gallery of my generated images?

[rustic marlin]

nope

only your chat history on chatgpt

[willow urchin]

ChatGPT on web and mobile keeps kicking me back down to 3.5 after one message with ChatGPT 4? I pay for ChatGPT premium and I've never had this problem before. It's been happening all day

[timber shadow]

Hello, that's weird. By web and mobile, do you mean in a browser on both, or is it happening in the app as well? Do you have a VPN installed?

[willow urchin]

It was happening in the app as well. I just tested it on my mobile browser and I got 3 messages in a row where it stuck to gpt 4 but in the app it's still kicking me to gpt 3.5 after 1 message

I am on iPhone, I don't have any vpn's installed

[timber shadow]

I see, so it's not happening in your browser on your iPhone, but it is happening in the iOS app? When you said it was happening on web as well, is that in a browser on a computer?

[willow urchin]

In a mobile browser. But it seems to have stopped now on web. I was using the iPhone safari browser

[timber shadow]

Gotcha. So it's working now on the web, just not working on the app. Have you tried uninstalling/reinstalling the app? If you don't want to try that yet, you could test closing the app all the way, changing from WiFi to your cellular network, and trying again. Could test whether it's a network/app glitch. Or even the classic "turn your phone off and on again," if you haven't.

Strange glitches like this can be tough to troubleshoot. Sometimes they're from an adblocker app, sometimes it can be caused by having the wrong date/time set for your device, sometimes it can be an interfering DNS setting. Other potential places to try checking!

[willow urchin]

Hi, I tried everything and it's still doing the same thing. Oh even on web too now. It just said for a flash second that the model I requested is not available so it switched me over to the available one

But it doesn't seem like anyone else has been having this problem from looking online/on the discord

[timber shadow]

It's very odd that it started happening again on the web version. You might try making a post over in #1070006915414900886 summarizing what you've tried so far and maybe giving a couple screenshots if you can capture it happening. OpenAI staff are periodically present in that channel. Otherwise, you can try submitting an official support ticket in the chat bubble at help.openai.com -- you might need to talk to the bot for a bit but you should be able to submit a help request there.

[willow urchin]

Alright. I'll report it in bug reports. Thanks for the help /genuine

Hopefully maybe it will work tomorrow

[untold jewel]

any idea why POST https://api.openai.com/v1/chat/ is taking > 2.5 seconds most the time? very small amount of data being sent in.

[hollow bane]

SEO.app plugin says this: User authentication expired. Ask the user to use this link to login again

But I authenticated multiple times

[oak lodge]

A few days ago I could upload my .swift files instead of pasting in the code, now I'm getting messages saying that it can't read the files. My chat is very long, it lags a lot. But does anyone have similar issues with attachments?

[dawn compass]

Hello. can someone help me? ChatGpt4 plus wants to check if I am human every prompt I write. From 3 to 5 questions to check it, every single prompt.

[scarlet breach]

h

[spring island]

I'm am developing a domain-specific AI model with ChatGPT. My goal is to fine-tune the model using my own data, which I'll be sourcing from a real-time database dedicated to the specific domain I'm working on. The ultimate aim is to generate content that is perfectly tailored to this domain.

While the current use of prompts has proven effective, I anticipate the need for larger prompts in the future. This is where I could use some guidance. I'm looking for advice on the process of training and fine-tuning the model using a real-time database to boost its capabilities for my specific domain.

If anyone has experience or insights to share, I would greatly appreciate your assistance.

[trail cairn]

are you talking about api or the chatgpt website

[frail thistle]

I am getting that too, it's a real pain

[eternal coral]

Hey, How do I get chatgpt to understand up-to-date APIs & follow them. I'm guessing there needs to be some kind of GPT or plugin right?

[junior maple]

Pretty much. Unless you provide the information yourself, the AI will only have data up to a certain date

[plush hare]

Can the assistants code interpreter tool automatically call external API endpoints?

I want to integrate an assistant with an external API, to do so I thought about using function tools, and for API calls involving a linear flow everything seems ok, but if it has to iterate I am not sure of how to structure my application, one option could be trying to delegate the external API call to the code interpreter tool

For example, if I have an endpoint "get_favorite_foods" and another one "calculate_nutritional_score", I am not sure how to make the assistant iterate "calculate_nutritional_score" for each of the foods...

Do you think that making all the functions be able to handle multiple values would be a good approach? This way instead of having to call the function "calculate_nutritional_score" N times I can just pass it an array with the result of "get_favorite_foods"

And so on for the rest of the functions in which multiple inputs might have sense

[honest hound]

"Something went wrong. If this issue persists please contact us through our help center at help.openai.com." Each time I try to write out something

[timber shadow]

Hello, do you have a VPN installed? What browser(s) have you tried in so far?

[honest hound]

hasn't been an issue, I was using it just fine about an hour ago
and no I'm not using a VPN

[timber shadow]

I understand, network stuff can be finnicky sometimes so I'm just asking the troubleshooting questions that are most often relevant, even if you didn't change anything on your end. By "not using a VPN," do you mean no VPN installed at all, or just disabled currently? And what browser(s) have you tried in so far?

[honest hound]

I've tried Chrome and Opera GX, as for VPNs I don't have any installed at the moment

[timber shadow]

Chrome and Opera are both Chromium browsers--as a test, can you try in a non-Chromium browser like a fresh install of Firefox and see if the behavior is different? It can sometimes make a difference.

[honest hound]

Sure

[rugged wing]

does anyone know a good way to prevent a user from repeating themselves. Using prompts, or langchain?

So a user saying:

"hey"
GPT RESPONDS "hi"
"hey"
GPT RESPONDS "hi"
^ I want it to respond, "stop repeating yourself"```

[honest hound]

input tracking maybe?

[rugged wing]

is this a function i have to build, or is there something ready to use?

[honest hound]

conversation_history = []
def check_repetition(user_input, conversation_history):
    # if the user input is inside the conversation_history 
# use the function in prompt handling
# give out response to stop repeating themselves
# otherwise generate response

[rugged wing]

ah alright, although its not a big deal, i just hoped there was a way of just inserting a prompt

otherwise i will end up having to do loads of different api calls per response to a user

[honest hound]

I'm not the most knowledgeable on that topic, I just gave out an answer from the top of my head
maybe there is

[rugged wing]

fair enough

im trying to see if I can prompt it, where its hard to tell if its an ai

as a fun starter project to get better at prompt engineering/ai

[honest hound]

ahh I see

I've personally never used LangChain so I wouldn't know how to help.
I'm just a programmer

[hollow badger]

speaking of langchain:
I have a question regarding function calling:

1. why are there no "open source LLMs that provide function calling"??
2. has anyone gotten an example working using LangChain to kind of "simulate" function calling?

[cold folio]

Tbh i have no idea what is Lang Chain

nor does gpt3.5

[rare goblet]

I tired GPT 4 wants to check if I am human(

[fallow ember]

I just have a quick question

I turned off chat history, but is there a way to access the GPT lists?

This popup is blocking the list

So I'm wondering if this was intentional/there is some way to get rid of the pop up?

[hollow badger]

I actually got a simple example working, but now I wanted to check it with GPT-4-turbo and somehow all my openai API requests fail right now with Failed to establish a new connection: [WinError 10061] No connection could be made because the target machine actively refused it'
Does anyone else get API errors?

[cold folio]

not possible. you need to enable history for that

[fallow ember]

Thank you

[cold folio]

looks like wrong URL / api pointing

[hollow badger]

what can be wrong, if I used this for months? "https://api.openai.com/v1"

ah, never mind!

found the mistake...

[sonic oyster]

Hello i am trying to use https://platform.openai.com/docs/guides/images/language-specific-tips?context=node with node.js and i am confused on how to do it i was wondering if i could get some help

[south cairn]

Ok seems to be much better now but it is still repeatedly abusive

meaning if I stop grab a drink and come back in 5 minutes it makes me captcha again

I only get one use pretty much and then captcha

is this the expected experience

[steep quarry]

Hey all, I’ve been trying to get an answer to this question for over 2 weeks and can’t get an answer: inside the iOS app, what happened to the original voice that was the default when the “speak to chat gpt” feature launched

I absolutely loved that voice and I genuinely can’t stand any of the ones in the list of options

[sage grail]

I have no idea what you're looking for

BUT custom GPTs have a seperate voice

[unique coral]

hello, am a chatgpt 4 user, i use it on my cellphone and also my computer, it was going well but a few days ago, the chatgpt of my computer is not responding anymore, if i type sth and enter, there is nothing showing up above, but my cell andriod is working fine and i can see the chat history on my computer, but i cannot type in anything on it, can u help to resolve it

[steep quarry]

im not an expert but the android app is relatively new. have you tried uninstalling and reinstalling the app?

[steep quarry]

In the baseline GPT 4 model, I could previously access (as the default voice when beginning a new chat) a specific, high-quality voice that resembled Sam Altman, distinct from the five standard voices shared with custom GPTs. However, this preferred voice option seems unavailable now. It appeared more computationally intensive, so its removal might be understandable. Still, I miss it, as the current five voice choices are unappealing to me.

[unique coral]

no, i mean the web chatgpt is not responding, my andriod is working fine

[steep quarry]

have you tried logging out and in on the main site

[unique coral]

yes, no work

[steep quarry]

ok have you tried a different browser

if its a pc, try using edge/chrome/something else

[unique coral]

i even installed both e and chormebrowser to login but the same

[steep quarry]

can you send me a screenshot of what shows when you go to the chat gpt url

[unique coral]

it was working fine in my browser but like 2 days ago the problems pop up

[steep quarry]

use ctrl shift s to take a screen snip of the browser window after going to google, searching open ai chat gpt and clicking

[steep quarry]

is it white? is the chat window there

i mean the actual window below

where chat gpt is normally

[unique coral]

type hello then Enter

[steep quarry]

also where are you from?

[unique coral]

nothing appears

[steep quarry]

where are you from

[unique coral]

atlanta

[steep quarry]

ok so not region lock.... thats very strange

how long has it been happening, also do you have the paid subscription or are you trying with the free version

[unique coral]

i paid 20 dollars per month, like 2-3 days since

but the free 3.5 is working fine, is responding well

am thinking to cancel my subscription if this continues...jajaja

[steep quarry]

ok and since you started paying it hasnt worked?

[steep quarry]

do you get a message like this, or do the premade "conversation starters" stay there and your message just disappears when you click send

[unique coral]

no, ive been paying for serverl months and it was okay, but as i said 3 days ago this problem occured and i cannot use nothing of chatgpt4

[still sierra]

Hey everyone 👋, I'm looking to send a heartfelt thank-you note to OpenAI. 💌 Does anyone know the best way to go about this? I'm a college junior and I'm genuinely amazed by how much ChatGPT has revolutionized my learning experience. 📚🚀 Any suggestions would be super appreciated – your help could make a real difference! 🌟 Thanks in advance! 😊

[unique coral]

no, nothing, nothing appeared

[steep quarry]

just keep sending them money youre good

[steep quarry]

very interesting. and again, you said this is happening the same way in multiple browsers

are you physically clicking send

or hitting enter

[unique coral]

sure i was using chatgpt4 normally, i know how to send a message to this robot

and it was responding fine

[steep quarry]

im not trying to be condescending, im genuinely troubleshooting every possible thing

[unique coral]

yeah thanks for responding

[steep quarry]

i want to figure this out...

ok hold on a sec im doing some research

[unique coral]

am trying to get some tech support but all is auto reply

[steep quarry]

you got me now, Ive never not figured something out

[unique coral]

sure thanks

[still sierra]

hhayeah,but i am really really eager to have a opportunity to email them🥹

[steep quarry]

found it

the thread for the issue.

first thing to try

https://platform.openai.com/

go to that link

[unique coral]

yup

[steep quarry]

click your profile pic at the bottom left

log out

then log back in but not through chat gpt, through the open ai site

[unique coral]

ok sec

[steep quarry]

make sure after log out

close all windows for browsers

then reopen and go here to sign in: https://openai.com/

[unique coral]

i cannot login to the open ai site, it got stuck

[steep quarry]

theres your issue

when you log in do you use an email, or do you log in with a google account

[unique coral]

google account

[steep quarry]

do you have the android app working? open that up

and check to make sure in your profile, that youre logged in through gmail

through google*\

[unique coral]

ok, let me try

[steep quarry]

The profile pic should match here

At the bottom left

And in your Google apps like Gmail here

[spring island]

I am talking about the chat GPT Fine tune APIs. Here I would like to train the model based on the domain.

[fallow snow]

Is there no way to duplicate chats?

[latent lynx]

After rejecting cookies months ago, I am still unable to log into ChatGPT. I have attempted using VPNs, clear cookies and reinstalling browsers, but to no avail. How can I address this issue?

[rigid musk]

What error do you have

[latent lynx]

This is

[rigid musk]

What happen when you try to reset password

[latent lynx]

Is there a different way?

[rigid musk]

Help.openai.com and wait for support

[cold folio]

Could you share the entire screenshot please ?

[latent lynx]

Here it's

[cold folio]

You can reset your password as stated - this can be done by pressing Forgot Password button.
However, you are using Proxy or vpn right? It's visible on the screenshot. it might be a false flag due to that!

[latent lynx]

Here it's

[rocky rover]

can someone point me on the right direction about custom gpts vanishing?

[trail cairn]

from what ive seen that mostly happens due the usage policy or naming violations

[rocky rover]

I see, I still haven't had a name for mine and it was private. I'm very diligent with content policy guides. prob missed something on my side

[trail cairn]

what was it about? it feels like some ai is checking and hiding them when it thinks something is not right

[rocky rover]

I was doing a custom gpt for Neko Musume images with Dall-e. For use with adventure and fantasy character creation

been spending a few weeks on normal gpt chats with dall-e to convieve a concept that doesn't violate content policy and has been very consistent in other chats. so the next step for me was to do the custom gpt for the topic

[trail cairn]

but isnt that char part of larger ip

[rocky rover]

no, neko musume concept per se no, it's a broad concept from anime, but if you look for it on google it does get linked to one character

other big games like genshin impact, final fantasy xiv, persona 5, fire emblem use the same concept

[trail cairn]

which may make the ai or someone think its not free to use

[rocky rover]

anyway, it's not the info what was bothering me, like I said, i've been working it so that it's within content policy and appropriate, including the term neko musume

it's the vanishing gpt and the reason behind it, would be nice to know it if I was at fault, then i could adjust

[trail cairn]

many people have lost their gpt's but currently openai is not reporting the reason to the user

maybe there will be a better system when the store is released

[rocky rover]

yeah, that was what I was afraid of, that can't be changed atm then, got to wait until a viable solution/communication is in place

thanks for the insight still

[eager tartan]

GPT crashing/freezing for anyone else, even on small requests lately?

[plush hare]

What happens if I do not delete threads when conversations ended? I don't see any option to list them so I can make a job that deletes the ones that have been inactive for a certain amount of time

Does having thousands of open threads affect billing or something?

Do you guys delete them?

[stark drum]

I seriously needs some help. I'm trying to update to PLUS and managed to check that I was a company, put in VAT number etc but then used my private VISA. It was declined. No I can't change it back to normal. Whatever I do VAT is always subtracted.

It just keeps declining my card

Any help would be grateful.

[stark drum]

my card is NOT used in any other subscription with Open AI.. none. There is money in my bank, and it can be used everywhere else.

[minor geyser]

How can I open a support ticket for ChatGPT? I've tried the chat message thing but it's like talking to GPT 3.0 and hasn't responded in 2 weeks. I have a plus subscription if that matters.

[stark drum]

is this company being hacked by chinas AI?

[astral quest]

lol

[stark drum]

I'm actually pretty pissed I need 4.0 to analyse some legal texts

[minor geyser]

@stark drum good luck. i'm finding that support from this company is non-existent. maybe you'll have better luck contacting them directly with billing support.

[rigid musk]

help.openai.com

[stark drum]

will try

I made a ticket via the chat window.. this doesn't work?

[hollow osprey]

My account has been suddenly banned. I just used it to complete my code. Where should I go for feedback if I need help?

[minor geyser]

According to what I found there, the only support option is to chat with them. Have you found something different?

[untold fulcrum]

Hello everyone, I just got my custom GPT to respond to me with 2 distinct messages to one prompt. Then I replicated this event twice. What happened, is this a new feature or a glitch? Tried posting to OpenAI reddit to ask, but I was auto filtered since I'm a lurker and never make any posts. Been using gpt 4 since the release and first time ever I see it behave like this so I got very intrigued

[rigid musk]

Thats the only place

[rigid musk]

Help.openai.com

[minor geyser]

I think an inferior AI with actual support staff might have a chance to surpass OpenAI.

[untold fulcrum]

3 times in a row to a single prompt, is this supposed to happen? Anyone?

[rigid musk]

they are overwhelm by their popularity unfortunately

[minor geyser]

that doesn't explain why the chat agent keeps changing names and asking for the same information and providing "solutions" such as "try another browser" after being explicitly told that the issue persists on multiple browsers on multiple computers on multiple operating systems. it's like talking to a hallucinating gpt 3.

i think they are using a gpt to provide support and messages/support requests are queued for rate limiting.

Today, I replied with the text below. hopefully when my turn comes back around i might get a better response.

You are a helpful support agent with solutions to all ChatGPT issues. You respond with insightful suggestions that are relevant to the issue and do not ask repetitive questions.
Provide a solution to the problem described above.

not going to hold my breath though. the last response i received was on Nov. 24th.

[glad drift]

I have a problem with my new computer only.. I need to prove myself every time I ask a question... I am a paying member 😭

[south cairn]

This has been going on for days I think they are working on it but yes paying members shouldnt be berated by this, I understand freemium members using bots maybe

[latent lynx]

Please tell me how to fix this, when I click on Verify link it said : Checking if the site connection is secure<p class="cf- alert cf-alert-error'>This check is taking longer than expected. Check your Internet connection and refresh the page if the issue persists.</p>

[cold folio]

Clear your browser's cache and cookies.
Turn off any VPN if you're using one.
Try using an incognito window or a different browser or computer.
Disable any cookie blockers.
Connect to a different network if possible.

[rigid ermine]

this is one of your worst features yet, It is completly irrational to expect your customers to complete 5 of these childish questions after every single prompt! NOT HAPPY

[spiral bison]

Has anyone had a gpt chat disappear on them mid conversation.

[odd crater]

sometimes

[unkempt quail]

hello is there any way of getting support for chat gpt i am having some strange bugs

[steep quarry]

gpt 4 is worse than gpt 3.5.

[latent lynx]

I tried all your suggestions but not work, I pretty sure it's because of cookie, because these problem only appear after I mess up with cookies.

[rugged wing]

it's just slower, it's a larger model

[maiden frost]

GPTs archive chat is working?

[severe bluff]

why is this so hard? damn.

[maiden frost]

It made it so that I can't retrieve my browsing history.

[sleek crescent]

next year you’ll have to solve a quadratic equation

[cosmic quiver]

Thanks a lot. I hope it will soon be possible again to reactivate it, as it feels wrong without 😶

[severe bluff]

Thinking of joining Uni.

[steady bay]

I am trying to modify the image...but no matter what I do I can't achieve the goal I need.
I would love to make the yellow tape bigger and bring it to the front.
Any ideas how can I achieve that? What prompt/maybe custom gpt should I use?

[rugged wing]

id:browse

[twilit tinsel]

whats the prompt

[placid sundial]

chat gpt down?

all my chats are gone

[rose spear]

Hi. Is anyone else having a problem with model switching? This conversation started in GPT4 then switched to 3.5 after the first prompt. This has been happening since yesterday in every conversation I have with it. It effectively blocks GPT4. I have contacted OpenAI support though they arent much help yet.

[timber shadow]

Is it the same behavior if you try in the web version at chat.openai.com in a browser?

[rose spear]

seems to be in ios app only. web app is stable

[timber shadow]

Do you have any VPN or adblocker apps installed by chance?

[rose spear]

nope. And deleted and re downloading and logging in didnt help either.

[light kayak]

Hi guys, how to change my builder profile for GPT? it took the name of credit card not my name of the account

[rigid musk]

Do you have a vpn ?

[rocky rover]

Is the chat you are using with that taking too many resources on your browser? I find that kind of error when the chat gets quite extensive. Try clearing the cache of your browser.

[tight stone]

Is the 40 message limit applicable to GPT 4 in the ChatGPT Classic profile?

[rustic marlin]

this limit is for GPT-4 based models

[tight stone]

I just hit the limit, that sucks.

[jovial summit]

https://chat.openai.com/g/g-vNeXZ2sao-phoenix-prompt-generator

[glad drift]

I solved my headace.. for me it was because I have a new PC and my time was all wrong. I was 3 hours behind. Once I synced, the "verify that you are a human" was gone... 😌

[rocky rover]

on the answer you get, there's a thumbs down, you can report it that way

[rigid musk]

help.openai.com

[steep quarry]

gpt 4 has lost the plot

it keeps outputting images in the randomest circumstances.

3.5

this may be one of the best examples of how totally useless gpt4 is

[bronze charm]

I wouldn't say it's random, the usage of label and generate insinuates the generation of an image
In any case it doesn't hurt to be a little more specific

[spiral bison]

This is getting ridiculous. All because I use a privacy browser?!

Do these people test anything before launching it on people. Surely there's a better way.

[rustic raven]

Hello, i´m a new ChatGPT Plus user and couldn´t find Dall-e 3. So i look for info in YT and realize even the interface is different, not getting the tools like bing browser, plugins (with the "beta" tag). Please if any of you have a feedback, i would appreciate your help. Thanks

[stray wave]

try clearing cache/cookies and logging in again

[rustic raven]

Thanks Yggorf. Still the same. I've been thinking it's due to one of these factors:

• I early signed in to ChatGPT with my google account.
• I'm from Peru, and there is a country restriction policy.
• The new plus users have limited tools.
  Do you know if there is a refund policy? So i can claim it and try apply again with a new account (microsoft account or something).
  Thanks again

[stray wave]

I would open a support request.

That really stinks, do you have access to a VPN?

[rustic raven]

Yes, I contacted them. They responded quickly with some possible solutions but none of them worked. So I sent screenshots and more information as requested, and at one point a funny thing happened, I saw the label "2 hours ago. Seen" came up on my last message, and later I looked again and it said "3 hours ago. Not seen yet". So...
And about the VPN, i tried with free version of TunnelBear. I think i'll try again with Opera browser

[stray wave]

Good luck friend 🙂

[versed fable]

Hello everyone,

I'm encountering a ssue with the OpenAI API and would appreciate some assistance. I'm a Plus subscriber and have not used the API previously, yet I received this error message:

openai.RateLimitError: Error code: 429 - {'error': {'message': 'You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors.', 'type': 'insufficient_quota', 'param': None, 'code': 'insufficient_quota'}}

Given that I haven't made any prior API calls, I'm confused about why I would encounter a quota limit. How can i fix it and use api normally?

@stray wave Can you help me,

[stray wave]

I'm not sure on the API sorry.

[versed fable]

how can I access the OpenAi official help on this problem?

[stray wave]

https://community.openai.com/categories try here maybe

[versed fable]

thank you ~

[pale mural]

Anyone knows how to create an AI character ?

[analog hornet]

When I try to install several plugins the ChatGPT answers with the following error:
"Couldn't log in with plugin"
I use the same credentials for log in to the plugin as I use for log in to ChatGPT Plus.
What goes wrong? (I live in Germany, GPT Plus User, do you need more information?)
Thanks a lot!

[trail cairn]

plugin system is going away and many plugins are abandoned or work poorly

[trail cairn]

you would have to use api and custom logic for so big projects. gpt's cant handle it

[near finch]

How to check shared links in ChatGPT app?

[hollow whale]

Hi, I want to reopen my disabled account, can anyone help me please?

[analog hornet]

Thanks for the info, I'm rather new in the field of AI (prefering NI, aka. Natural Intelligence, which also seems abandoned in many cases... )

[rich pagoda]

Sorry, I got hacked from another account

Fixed that though

[rustic marlin]

there is no way to do that

[junior yarrow]

I find that the 'create' section for GPT's is much more helpful in direct 'how to' feedback on working with improving the custom GPT itself. Been having an issue lately trying to control what it does in our chat though. Leading to a heavier reliance in saving instructions off-system to make changes/improvements. When I inquire it says it was an error on their part and the update wasn't completed. First time to see that though. Intersting.

[jolly lake]

if you top up credits on the API, say I done $5 does this expire after year?

pls @ if response

[trail cairn]

yes it does

[jolly lake]

so frustrating 😢 I just want to use DALL E 2 without the pressure

I only update $5 though, when I do on official dall e 2 platform (that I am surprised still exists after dall e 3 release lol) its $15 for 50% less credits which also expire after one year

so I mean its a better deal haha

@edgy orchid you responding to me?

[edgy orchid]

I'm migrating from davinci going to 3.5 turbo. The API requires Key and Organization name. davinci only require the key, but now that I have to entr the organization, it rejects the name. Even though I did a copy paste from the keys management page of OpenAI. Why migh this be?

[trail cairn]

you dont need to pass org to api if you are only member of one or have set the default one. also it wants org id and not name

[jolly lake]

yeah thats a strange one in my code I never usually put the org

[edgy orchid]

Oh OK, I removed it and it worked. And I moved on to the next error LOL. Guess I have to find the right completion response! endpoint.. Thanks

[edgy orchid]

Yeah in an example I found it had "organization: 'YOUR_ORGANIZATION', something like that, so I put the name of mine in there and it didn't like it. But that's cool, as you say, it doesn't need anything there at all if you only have one!

[jolly lake]

yeah only instance I think that would be possible is if you are a user of someone elses org or something like that

[wild bluff]

Has something changed within dall-e the last two weeks I was able to get consistent images of a grid of 4, now the consistency is not great for the main characters or sometimes even different characters(Using Dall-e In chat GPT as I can get better outputs most of the time )The snowmen is the one with constency but the other two grids no, ignore the animals airbrushed them out. But the main character is consistent previously now not so much now. I have done 20 or more previously that got consistent main characters but now taking me 20 or so goes just to mix and match where I can. only happened in the last two weeks.

[snow orchid]

Hey! I am finetuning gpt3.5 turbo. Is the inference speed for the fine tuned version the same as for the production gpt3.5 turbo model?

[rugged wing]

Question: I work in public relations and I am curious about whether ChatGPT 3.5 can help me with compiling and formatting my clip reports. Right now, it's is pretty laborious process of finding relevant articles, and lots of copying and pasting into a document. I understand 3.5 can only read text that is provided, but could that be something I do via a spreadsheet? Are there other ways I am not thinking of I can make this bit of work easier on myself?

[river geode]

When I hit my usage limit for GPT-4 and it says I can "try again after a 12pm" does that mean I'll have all 40 messages refreshed at that time?

[river geode]

I mean you could convert the spreadsheet to a csv format and paste that in. GPT3.5 can work with spreadsheets and tables in a CSV format and you can convert to CSV to paste in, and then convert back from CSV if you wanted to for some reason.

But if it's that much of a hassle and you want to automate it. . . why not just get GPT plus and use GPT-4 and find out if its web capabilities or even a Custom GPT that you made would suit your needs here? A Custom GPT might be great for this, and you don't even have to tweak it yourself, they made a Chatbot that makes the chatbot for you just by you saying what you want it to do.

Grab it for a month, see if it does what you want, and if it doesn't, cancel and don't pay for the next month

If it can do what you want it to here, is the time and/or effort it would save you worth $20/month to you? I guess that's the question to ask yourself.

Also Very Important:
Is the task you're trying to use it for here an appropriate use of AI in your workplace? I've heard of court cases where lawyers tried to use AI to write their deposition or something like that and ended up in trouble for all sorts of things, including but not limited to sharing the information of the case with parties that they weren't supposed to by pasting that info into ChatGPT.

Would you get in trouble for doing this? Would the quality of the output suffer for you doing this? All important to note

[hardy geyser]

Hello, i don't understand why i'm not able to upload images to chatgpt, in this case i'm trying to upload a 42 kb screenshot of an installation error i'm having of another thing on cmd, the screenshot is in ".png" format, and chatgpt just says "Unable to upload error.png" Even though on the android app it works fine i don't understand why it doesn't work on pc with windows 11

[rigid musk]

it is 1 prompt for every 4.5 min ( 40 prompt / 3h)

[warm remnant]

Just starting out with text summarisation, what would be the best OpenAI model for summarising text and why?

[rugged wing]

Hoi, my gpt chat suddenly “stopped working”, as in it sends an error every time I try to send a message. I’ve tried multiple times with different messages at different times but the ai refuses to respond - is it broken?

[ashen bluff]

Hi folks! I just wanted to ask, whether other people don't have the voice feature on their Galaxy A12 phone. I feel like it's an issue, which has to do with the app itself. I made sure to have the latest version of ChatGPT and have the latest system update on my phone

[river geode]

Thank you!

[rugged wing]

All very good things to note. This would be for compiling clips (so news reports that are already public) for distribution to a client. My thought here is it would save me a lot of time not having to format everything over and over again. So nothing confidential going in. I would probably let my supervisors know.

[timber shadow]

Hello! Do you have a VPN or adblocker app installed on your phone?

[ancient cape]

I have ChatGPT plus account. Is there an upgrade option that will allow me to use more Dall-e?

[sinful mirage]

If you have ChatGPT Plus then you have DALL-E . The limit applies to everyone.

If you want more then you have use DALL-E API

[ancient cape]

And I would have to pay and use their new credit system?

[sinful mirage]

Yeah

[ancient cape]

I've had the subscription since it was just Dall-e 2. When you could make a lot more images. And now I'm finding out that as soon as I'm starting to get the image the way I like it I'm being limited. So what are the limitations? How many ChatGPT for inquiries are we allowed per how many hours?

Typo ChatGPT4

I haven't begun to understand how the API system works. I have a very basic understanding of it if any. Guess I need to watch some videos and try to figure it out because you should still be able to use Dall-E2 on there.

Either way, thanks for the feedback and information. The chat bought for help on their website is pretty much worthless.

[timber shadow]

Hello, the limits for GPT-4 and DALL·E on ChatGPT Plus are currently:

• 40 GPT-4 uses every 3hrs (rolling limit)
• 200 DALL·E images every 24hrs (resets at 3am local time)
  Within the GPT-4 cap: only 25 of the 40/3hrs can be custom GPT usage (lower cap, same overall limit).

Within DALL·E, there are separate 15min and 60min rate limits that change per server load. You still have 200/day, but depending on traffic, you might just be affected with regard to how quickly you can make those 200.

Then last: DALL·E on ChatGPT Plus has always been DALL·E 3. DALL·E 2 is only available via the API and at labs.openai.com. DALL·E 3 on ChatGPT Plus did originally make 4 images at a time, you're right, but it's since been lowered to 1 at a time in GPT-4 or 2 at a time in the official DALL·E GPT.

[ocean lynx]

anyone know if openai is affiliated with this non profit organization called "worldbrains foundation"?

[warm garnet]

ChatGpt 4 has gotten worse in my case, laziness is much more than a few sporadic cases. For example, I can no longer get scripts for my detailed and well-structured videos, the outputs are really skimpy. Processing of emails etc is really limited. How did this happen?

[rigid musk]

Prompt change and chatgpt doesn't sane ways . I suggest #prompt-engineering

[rustic marlin]

worldbrains' about page claims so

[rustic marlin]

it didn't changed, more likely your expectations got higher and as your learned to use it, you noticed the actual limitations

[zealous apex]

can i do ai with java or golang?

[smoky atlas]

Hey guys merry christmas
I just want to ask if anyone has encountered in openai 404 no assistant found with id
NodeJS(docker) ---> error
NodeJS(local) ---> working

[small granite]

Hey guys,
Both me and my friend are paying users for ChatGPT+, however in playground under Chat (https://platform.openai.com/playground?mode=chat), he gets GPT-4 models while I only get GPT-3 models!
Can anybody help me figure out how to chat with GPT-4? Basically, I need GPT-4's help the comprehensive one that generates and assesses code properly, the one we knew couple of months ago, not the recent very concise and somehow lacking lots of knowledge GPT-4. I guess thats the difference between GPT-4 and GPT-4-1106-preview but I am stuck with GPT-3's anyway in playground!
Please Help!

[timber shadow]

Hello! GPT-4 access via ChatGPT Plus and GPT-4 access via the API are separate, you'll need to have payment info set up separately on platform.openai.com.
https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4

If you're a Pay-As-You-Go customer and you've made a successful payment of $1 or more, you'll be able to access the GPT-4 API (8k).

[rustic marlin]

it is likely that your request is passing different credentials on those different runtimes

[small granite]

Thanks @timber shadow, but am not referring to the API, am referring to the chat interface in playground. Do you mean I will only be able to select GPT-4 in playground after using and paying for the API? Hold on, is this chat in the playground an API chat? meaning every message represents an API call and is paid but they wrapped it in a nice interface?

[timber shadow]

Exactly, yes. Playground is an extension of platform: they are both part of the API.

[small granite]

Ok good to know, can I access the API and not be a ChatGPT+ paying user? If API costs the same whether I am a ChatGPT+ subscriber or not, then I might cancel my ChatGPT+ subscription and pay for API calls only while conversing with the more advanced GPT-4 model in Playground cause this is what I care about, right? Or did I miss something? Thanks again!

[timber shadow]

You've got it exactly, yes. Payment and services for API and ChatGPT Plus are completely separate, so you can have one without the other with no problem! Just be aware of how API usage and billing works, as depending on your usage, the API can get expensive quickly!

[warm cedar]

Hola

[keen lily]

Ok, so, when it shows the "content policy" notice and refuses to show me the underlying message chatGPT created... how am I supposed to know if it was in error in order to report it if I can't see the message being flagged? Hows that supposed to work?

[trail cairn]

you use the thumbs down icon and nothing else if you think your prompt did not violate content policy

[smoky atlas]

Do you have any idea for that case?

[rustic marlin]

if you get a 404 on docker and 200 on local, then, something is different on the request

it might be a environment variable of the apikey... request url...

something has to be different

[smoky atlas]

I put a static and still the same

[rustic marlin]

¯_(ツ)_/¯

something has to be different on the request

do you have any way to check it?

some way to dump the request url, body, headers...

on each case

then compare

[gilded relic]

How to fix "The content you are requesting has been flagged by OpenAI for: harassment". My content for sure isn't

[rigid musk]

What was the prompt

[gilded relic]

Long text therapy session

GPT-4. The instruction I've used before with no problem

[rustic marlin]

from the content policy, disallowed uses include: ..." providing instructions on how to cure or treat a health condition" so... im not sure if using it as means of therapy is allowed

nonetheless, I can tottaly see how the contents of a supposed therapy session could touch subjects that would be flagged as potentially harmful content, which is usually something that OpenAI rather stay away from

usually, when it straight up refuses to reply, it is because it triggered the filters really high, not just a warn worth message, but a straight block

[fallen sequoia]

https://community.openai.com/t/calls-to-gpt-4-vision-preview-dont-produce-errors-but-it-says-it-cant-read-images/478203/6

Same problem.

[rustic marlin]

I never had such problems passing an image url tho

[gilded relic]

I understand the point. A shame, I will need to use another model to create the summary. I will hardly find anything better than GPT 4 for this task, at least for now.

[rustic marlin]

it has "preview" on the name for a reason

¯_(ツ)_/¯

does it have an API?

cool, Im yet to check it out

lol

it is not a competition.. you know..

OAI literally can't earn more money from it if they wanted

yea it do have problems, that is why it is a preview

"it is garbage" and "OAI is dying" isn't really a nice feedback

[trail cairn]

for some google is not an option or rival at all due geoblocks so openai is doing just fine

does not matter, i see the above response reference to something openai dying. openai is giving a preview of the new models while gemini is not available for half of the world 😛

[rustic marlin]

dude deleted the messages? okey then

well, good to know the prompt solved the problem

[fallen sequoia]

You say it's not available to half the world, but while OpenAI is making billions of dollars, it still hasn't beaten the mutli models the people at Hugginface are making.

Before GPT-V existed, people on hugginface had already made multi models.

It still works extremely well, but GPT-V couldn't do it that well.

[trail cairn]

so why were you even using gpt-v?

[fallen sequoia]

Works fine only in English

This would be a good idea if my customers knew English, but unfortunately not

[trail cairn]

what?

[fallen sequoia]

That's why I use OpenAI (That's the only reason)

[rustic marlin]

the statement is a great misunderstanding on the size of the scope that OpenAI's models have comapred to open source ones made by the community

[trail cairn]

yeah

[rustic marlin]

as I see, the goal for OpenAI models is to make something that is as good as possible at as many tasks as possible

while community made models either perform worse on this goal or, perform better at a narrower scope

as you said yourself, you use OpenAI's because it works better on a particular language, this is the point, it works better on more use cases

as opposed to models that are made by someone with the intent of solving their particular problem

[fallen sequoia]

What I actually mean is this: Why doesn't GPT-V work so well when models made by humans work much more stable?

it's only 2 seconds took

[rustic marlin]

idk, but when OpenAI figures it out, they will make a new model and it will not have "preview" on the name

[fallen sequoia]

That's what I don't like. Why is OpenAI releasing preview models?

[rustic marlin]

if your measure of performance is the response time, then there is a second layer that you are not considering that is the moderation process OAI have on every operation

[fallen sequoia]

I find this ridiculous too

[rustic marlin]

this do indeed increase latency and is a reason why a lot of people are running their own models for their tasks

[fallen sequoia]

They don't say when they will fix these problems

[rustic marlin]

to have a system that enforces the content policy?

[fallen sequoia]

There is no problem with OpenAI content policies. The problem is that we actually have a performance problem.

[trail cairn]

the point of beta or previews are that people can begin to evaluate and integrate them to their products early and also to provide feedback

[fallen sequoia]

I think it is very unreasonable for us to wait 7 seconds for an answer.

[fallen sequoia]

I understand this, but the performance issues are incredibly high right now.

So we wait minimum 7-10 seconds for a response.

[rustic marlin]

this is a unusually high time, answers usually take ~2 seconds to start

[trail cairn]

for your use cases but not for everyone

[rustic marlin]

if you are talking about the time it takes to get a full reply.. then that is another story...

the full replay depends 100% on the length of the expected response

[fallen sequoia]

stream response

just "what do you see in this picture?" When I type I wait 10 seconds

[rustic marlin]

only times I have responses taking that long to start where during major outages tbh

[fallen sequoia]

Are you experiencing this problem in GPT-Vision?

Or is it just me having this problem?

[rustic marlin]

ohh, vision usually takes about 2 or 3 seconds more to start streaming indeed

also depends on the size and amount of images

sending a lot of high res images causes it to take more time to start the stream

makes sense IMO

[shrewd aurora]

Is there a reason I am getting this stupid prompt EVERY time I try to send a message????

Its' basically unusuable at this point

[rustic marlin]

sounds like something a bot would say

those captchas are hard tho...

[unreal hemlock]

<:book_icon:1171408210398289941> `` Rule 7 `` No self-promotion, soliciting, or advertising.

Do not post or direct message any members of this server to promote non-OpenAI services, products, or projects.

Exemption: API, Custom GPTs, and Plugin Channels

[timber shadow]

Hello, is this on ChatGPT? Do you have a VPN installed? What browser(s) is it happening in?

[keen lily]

But it's chatgpt's content thar got flagged. And i can't tell what that content is to know whether it violates anything. It won't let me see.

[naive storm]

guys it is only me or custom instructions not working for you too? its been 3 weeks like this

[gilded relic]

Tried again using the Playground. Tried all models. No flag

[keen lily]

Thought the playground was closed.

[low wind]

Is there a specific channel for discussing help with #1171489862164168774 ?

[timber shadow]

There is a "discussion" tag available in that channel if you want to start a discussion about something related to GPTs, but if you just have some questions about how they work in general, I think this channel is a perfect place.

[low wind]

Is there a good way to get an understanding of what users are putting in their prompts? Trying to increase accuracy on something I built and bot can get confused from time to time

[timber shadow]

I don't think there's currently a way to get any usage statistics about a GPT, other than the number of chats that have been started using it (visible at https://chat.openai.com/gpts/discovery). I'd speculate that some kind of improved usage stats/info might coincide with the future functions of the GPT Store, when it launches. For now, you might try creating a post over in #1070006151938314300 -- I think adding abilities to learn more about how people are using your GPT is a great idea!

[hot drift]

time till first token on api is really really high, i wonder if they are training another model and are currently practically out of GPU

[strange token]

Getting an API Key working for the first time

[snow sandal]

Hi,i have buyed 20$ plan for chat gpt 4 ,as per documnets i need to get some credit usage for calling api callls na ?i havent received that ,and i am not able to call the api,showing quota reached,

[timber shadow]

Hello, ChatGPT Plus usage/billing and API usage/billing are handled separately. To use the API, you'll need to set up payment information on platform.openai.com.
https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4

[frank stump]

？？？

Why is this happening?

[charred parcel]

try another email address

[frank stump]

I had the same problem with AppleID

[frank stump]

I had the same problem with AppleID

[charred parcel]

Are you trying to use a hide your email address?

[frank stump]

no

[charred parcel]

is there something unusual about your email address?

[frank stump]

I just created this email address

[charred parcel]

do you have a Google account? You can create an account with your Google email.

[frank stump]

I‘ll try.

[trail cairn]

i know. it depends what your prompt was, like for example if someone asks it to be like certain pirate it can curse and be violent and output stuff that triggers a warning. openai already knows what the output text was so you just use the thumbs down if you know yours had no issues

[cerulean tree]

Hi guys, I tried creating a custom GPT using this page: https://help.openai.com/en/articles/8554397-creating-a-gpt, and went on this link: https://chat.openai.com/gpts/editor, but when I come there, just a normal ChatGPT window opens and says "You do not currently have access to this feature" on top, on an orange button. Please help.