#community-help
1 messages · Page 147 of 1
Sorry, not sure when dalle3 will work in the gpt4. And how can I add and use plugins?
You need to be a chatGPT+ member. if you need help how to become one, please ask.
As a ChatGPT+ member, you need to turn on plugins in your beta settings options - if you need help finding that, please ask.
Last step, you need to hover over the top of your web browser where it says 'GPT -4' and click on 'plugins-Beta'
Right below where it says GPT-4 you now have an area where you see your selected plugins, click on that and go to the plugin store. Pick your plugins, and you'll have to enable each to use it in that chat. You can have up to 3 enabled in one chat.
If you have any problems, be sure to ask!
GPT 4 and GPT 3.5 has not worked for me for last 2 days.
I am getting:
Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
Is there an outage? can someone help please?
i have gpt 4 but i dont see the browse with bing in the beta settings
Then the rollout hasn't reached you yet. Hopefully soon!
alright thanks
Does anyone else have problems with the Acces PDF & Docs plugin in GPT-4?
will dall-e 3 be exclusive to gpt-4?
Just wonder if DALLE 3 got released to ChatGPT Plus accounts? It's October 1st already. I don't see it working though: https://chat.openai.com/share/12d11762-b5f5-4336-bb1a-6496da4d12e1.
early october could mean alot of things
hmm, true
I thought they just said “October” which is even more vague
But hopefully ‘early’ is like by the tenth or something like that lol
is image input still not out yet? I saw the plus sign once, didn't use it and now it's gone
still dont have access to it
I keep getting "I apologize for the inconvenience. It seems that I wasn't able to access the Google Sheet. Could you please make sure that the sheet is publicly accessible or shared with anyone who has the link? Once you've done that, I can try again to assist you with your analysis." while using the sheets plugin, but it is a public sheet. How do i fix this?
Guys I have a question, My credit card is declined by open Ai, what do I do? 
Perhaps contact their customer support, https://help.openai.com/en/articles/7232916-why-was-my-credit-card-declined and there's a chatbot that can help pass messages to a human in the bottom right - but they can take a week or two to respond if you need to try to interact with a human.
Other possible answers:
-
Another credit card?
-
Some have reported they have been able to use various sorts of gift cards to pay for ChatGPT+ through the app, but I haven't heard if there is another option for the playground.
This is the message I get every time i try to open chatgpt, even though I clear my browser/cache I still cant access chatgpt! any ideas?
looks like you have some sort of 'network protect' blocking it or an addin on your browser.
I always seem to be last in the feature rollout. My partner (in same household) already has image upload, but I don't. Been like this for every major update so far. Frustrating tbh.
i dont have it either, just gotta wait
I also don't have vision.
I am blind
same i lost my vision in a chemical spill 3 years ago
I noticed a $3 bill from OpenAI for GPT, but I haven't been billed yet. Why? I'm interested in gaining access to GPT-4. Do you have any insights on this?
Why did my account go back to the free plan? I never cancelled it and there is no problem with my payment method
has a payment been deducted?
No it simply stopped
have you tried to resubscribe?
Not yet because I haven't been using it as much lately so I'll probably do it once vision is available in Europe but I was just wondering why it stopped all of a sudden and I also didn't receive any emails about it
I don't know, we can't deal with individual account issues here, you could contact support@openai.com for more specific help
Okay, thanks
Hey one quick question might be dumb but new to Ai so I'm using openai's embedding model to store my dateset in vector store. so how actualy this OpenAI embediing cost me like when I ask one question it will do one request for input question one for response output question that I know and now there is OpenAI embedding request so how it will cost me. like in every question or only first time during storing embedding?
Hey guys, if I don't have voice conversation in Settings-new features on my iOS app ( I am a PLUS subscriber), does it mean I can expect it in the next week or so? Thanks!
They still rolling out
Got it thanks Neko
um
what's going on with bing
windows copilot
Hey am not opposed to this I think it makes it more intresting rather than being filtered to the ground and boring! Regardless i think it may be a glitch. If its intentional this is good 🙂 Make the experience more fun! Rather than talking to some snowflake who cries about the slightest disagreement
Hey one quick question might be dumb but new to Ai so I'm using openai's embedding model to store my dateset in vector store. so how actualy this OpenAI embediing cost me like when I ask one question it will do one request for input question one for response output question that I know and now there is OpenAI embedding request so how it will cost me. like in every question or only first time during storing embedding?
The OpenAI embedding cost depends on the model you use and the number of tokens you process. A token is a piece of a word, and 1,000 tokens are about 750 words. The current web page context shows the following prices for different embedding models:
Ada v2: $0.0001 / 1K tokens
Babbage v1: $0.0004 / 1K tokens
Curie v1: $0.0200 / 1K tokens
If you use the new embedding model, text-embedding-ada-002, which is more capable and cost-effective than the previous models, you will pay $0.0004 / 1K tokens.
To calculate the cost of using OpenAI embeddings, you need to multiply the number of tokens you process by the price of the model you use. For example, if you use Ada v2 to process 10,000 tokens, you will pay $0.0001 x 10 = $0.001.
Hope that helped
guys i want to ask something
so for last month i have been talking about gpt4 being slow but then i get muted saying it tried to bypass a filter and also i am relatively new to discord can someone explain this to me?
any mods that can tell me about this issue?
I dont think GPT4 is slow at all. Did you check your wifi issue?
Yes now it is better but i want to ask is what did i do to get muted when i asked a questions with simple words with nothing to bad said ( sorry scared that i will get muted again). Now i am scared of getting banned to ask questions.
Are you sure about that
Are you talking about GPT4 or The mute
Both
but I cant say the same about GPT 3.5
Its called behaviour drift in LLMs that can make some prompts that used to work before not work anymore
oh
sometime you have to make a new prompt to fix it
And about the mute i think we can't talk about it in this server
also like i asked it to fix my code which it mostly wrote then it as i just kept asking it to fix the code in the same chat it just become dumber so i had to delete the chat and yea
the auto mod is too aggressive
wow 💀
[openai please dont mute me im sorry]
Hahahah
I am sorry too openAI don't ban me or mute me please
anyway
there is one question
how does OpenAI even have enough servers and money to handle GPT 3.5 and GPT 4 like there trained on more data than i can possibly imagine
and there downtimes are rare
I think due to the partnership with microsoft they are getting better infrastructure
oh
I imagine OpenAI and Google being compeitetors soon (im sorry OpenAI if this offended you and im sorry auto-mod)
I think open ai is too far ahead it will take years to even catchup
True
Also I wonder what OpenAI will do about a thing on GitHub called freegpt-webui, it practically allows you to use ChatGPT without signing up or paying and in less than 10 minutes you could make a API out of it
yes but the dev got a lawsuit
WHAT
hello, I'm a plus user for more than 7 months now but I don't have access to new beta features ‘Browse with Bing’. Any idea of what going wrong please?
this happened to me once so just raise a ticket about this issue and they will contact you in a week.
thx
your'e welcome
So like if I'm creating one app where I'm upload 3k token pdf file in my chatbot using langchain and using OpenAI embedding for thiss and For Chat I'm using Chatgpt-3.5 so my cost will be like first for uploading doc 3k token so = 0.0004 * 3 and plus for each question ( also embedding cost to retrive data acording to question cost + gpt3.5 cost for getting answer from question ) right?
Your question is about the cost of creating an app that uses LangChain, OpenAI embedding, and ChatGPT-3.5 for uploading a PDF file and answering questions from it. Here is my answer based on the information I found on the web:
The cost of uploading a PDF file to LangChain depends on the number of tokens in the file. According to their website, the price is $0.0004 per token. So, for a 3k token PDF file, the cost would be $0.0004 x 3000 = $1.2.
The cost of using OpenAI embedding for retrieving data according to the question is also based on the number of tokens processed by the API. According to their website, the price is $0.06 per 1000 tokens for the standard tier. So, for each question, the cost would depend on how many tokens are needed to query and retrieve the relevant data from the PDF file.
The cost of using ChatGPT-3.5 for getting an answer from the question is also based on the number of tokens processed by the API. According to their website, the price is $0.08 per 1000 tokens for the standard tier. So, for each question, the cost would depend on how many tokens are needed to generate a coherent and accurate answer.
Therefore, your total cost would be $1.2 plus the sum of the costs for each question, which would vary depending on the number of tokens involved in each step. I hope this helps you estimate your budget for your app. Thank you for using Bing! 😊
Just litrally use bing or ask ChatGPT
Does anybody have any tips on how to get chatgpt to expand its word count?
I'm writing a product description. I gave it the one I wrote myself so it has the information. That one is about 1000 words. I've asked it to just check the grammar, and increase the word count to a minimum of 1300 words.
I get a response back saying "This now meets the 1300 word minimum requirement" Yet its actually reduced my description down to about 500 words...
3.5 or 4?
ChatGPT can't actually count--it's hallucinating when it's telling you it has met (or even can meet) the word count requirement.
I'm using 4 at the moment
It's a limit to how LLMs work today. Ask ChatGPT: "Can you please explain why you can't count things like characters, words, or tokens in your responses to me?"
May I upload an image? seems the AI is stuck in a loop for me.
Makes sense I suppose. Guess I'll have to do some actual work myself!
Theres a lot of ways
Does anybody here know how to get an invoice from OpenAI? I am not using the API, only the ChatGPT plus subscription, but it seems like you can only get invoices if you use the API? or am i missing something haha
you can work around it by just telling GPT to be detailed esxtend what you wrote etc
with clever prompting its an easy soultion
quick question: i do not yet have access to the new ChatGPT image capabilities is it safe to assume that I will but I just have to be patient as they slowly rolw it out. I am a plus subscriber.
Yes, we are halfway through their 2 week rollout period
Thanks!
Weird error loop.
Supposedly this week it should be rolled out to everyone who is plus and enterprise IF and only IF you are not from the UK or EU
you need to have enough api usage to be billed and account type be pay as you go, if you want plus sub invoices then you can view them from the manage your plan page
I am not sure if I can find the 'manage your plan' page for the chatgpt plus subscription. I can only seem to adjust or check the billing/plan in the API tab, but not in the ChatGPT tab. If I click 'my plan' in ChatGPT, it directs me to the playstore where I also cannot get invoices haha
general question with background. i homeschool my kids and we do a lot of it on the computers. i want to integrate ChatGPT as an on-hand tutor on the computer as well because i think it's a great tool and thing the "pro" version is the most useful, especially considering the most recent updates. would i have to pay for the "pro" version for each of my kids or might there be some sort of package deal down the road potentially (ie: pro-family get 4 login credentials)? im sure i know the answer im going to get, but it never hurts to ask.
To ask about an automod action and how it got triggered, it's best to send a message to modmail, upper right corner, and select discord mods to send the question to. They can explain how you triggered it (I also need to ask sometimes)
So as a Plus user you have a limitation on how many messsges you can receive within 3 hours. You can trick and make chatgpt think it is an educator but you’ll hit your limit depending on how many queries you have. I’d recommend perhaps checking with OpenAI staff for the enterprise version if you afford it.
What you are trying to do sounds like you would need the Enterprise Version anyways
thanks!
Ehh, from what I read, enterprise is more for businesses, but you can look into it if you want. https://openai.com/blog/introducing-chatgpt-enterprise
thanks for link
fyi they do not currently seem to sell enterprise unless there is a large amount of users (+200) with minimum commitment of one year
Correct but he is looking for essentially a Tudor or teacher for multiple children. So this would be more of an Enterprise function. What he wants to do with Plus is going to have limitations and won’t be able to properly function as he wants so him checking out the enterprise function would be a good start. He might be able to get away with Plus with one child but not many.
Thanks Radium for the information
i kind of figured my options would be very limited if any at all really, so this ALL helps. thanks guys
One thing to be aware of, OpenAI terms of use, "You must be at least 13 years old to use the Services. If you are under 18 you must have your parent or legal guardian’s permission to use the Services." https://openai.com/policies/terms-of-use
So, it may be against terms to have kids under 13 working with the AI.
However, for those kids who are old enough, an email and a phone number plus some info is all that's needed as far as I know to create an account, and up to two accounts can be connected to the same phone number.
So you can potentially have all of-age kids with their own account, interacting with the AI.
https://help.openai.com/en/articles/6613520-phone-verification-faq
And those can be the free accounts.
Understanding the phone verification security requirement of our sign-up process
awesome
Guys, am I the only one who can't log in to https://platform.openai.com/ ?
Khan Academy is already integrating ChatGPT into some of their online lessons, and you can check that out if you choose and sign up if you wish, as of a few months ago it was a trial pilot test program. I haven't kept up with it, and we can't share links except from openai, so you'll need to do your own research if you are interested.
Even though he is the parent this still stores information of a child under 13 and could cause OpenAI to violate COPPA. All does OpenAI offer waivers for COPPA for parents such as this?
try in private mode
I posted a link to the company's terms of use, perhaps you can find the answer to your question there 🙂
im definitely going to have to check that out
Nope, nothing happens 😦 Did it work for you?
i logged out and back, yeah now its stuck in the token error
To my thinking and based on my experience, your options are pretty wide, if you have kids of the allowed age range and if you have means to provide accounts for each kid of-age.
The AI is not a teacher, but it can discuss many things that are of educational value and usefulness.
It is prone to some errors - but most textbooks have a few and most teachers have something they were taught wrong that they never knew was incorrect, so the human teacher can keep teaching misinformation too.
And the AI if asked the same question in multiple new chat windows is unlikely to give the same hallucination many times, (a textbook or human instructor probably would) so that can help identify mis-information, plus your involvement and help.
Uh oh... Does this happen often? I've never seen it before
involvement is key. my main focus would be to utilize as an enhanced research function. especially where my particular knowledge might be skim
not often, but sometimes yes
I guess we just wait until someone decides to wake up and fix it 😦
I have a question if someone knows. I signed up for Plus as soon as I had the option but I like my bills to be billed on the 1st of the month. If you can plus will you lose all your chats and features when you resub?
You can have 'custom instructions' set that tell the AI what it should know and do. Now, that can be changed, and by the kids too, for any new conversation, but if your kids cooperate you can guide the AI to know specific instructions you want it to follow. It's best to experiment and explore this a lot yourself, and to learn some basics of how to work with the AI yourself and guide it to do what is wanted.
But that's for more complex stuff - if asked to discuss, say, "what is great for a 15 year old to know about the transition between the European Dark and Middle ages" or whatever else, it can give a lot of rich and informative answers, and backed up by answering follow up questions - it can ask follow up questions and checks for understanding if told you want it to - it can lead roleplay and problem solving as if one were living in those times - discuss what games kids typically played then and what a day in the life was like - all levels of interactive detail.
see, that sounds awesome, and i am definitely trying to dive in myself. messing around and seeing what i can do with the free versions as well right now
I recommend you remain aware, and teach your kids to become and remain aware, that anything the AI says 'this time' could be completely made up. It is very important to check information the AI gives, especially if it is in some way crucial information. The same ability that allows it to put words together to describe what a day in the life of a 15 year old in France in the 1500s likely was like can lead it to make up random nonsense - and use confident, believable and persuasive surrounding language.
It also gives highly accurate information, but it gives it exactly the same way as it gives make believe in many cases, and it can even defend make believe (especially in the same conversation).
A way around this is to quickly google the information, another can be to start a new chat and ask again - hallucinations are rarely the same even if it is hallucinating every time.
Some stuff it will reliably hallucinate about. Like, it doesn't know the text of most books, it knows a summary and 'thinks' it knows the whole book. So asked about, say, 'how is food used in this not 100% popular and super well known book' - because something like Harry Potter series, training data of people talking about everything about the book (but not the actual book itself) can give the model that info - but you'll find that it makes up a different story about food use every new conversation that asks, because it is making it up each time and doesn't know what it told you between conversations.
i've actually talked about this with the kids too. hence the "enhancement" rather than relying soley on it. it seems to cut research time from say a few hours to 15-30 min. and saves time for other subjects or projects
Lol Thank you Now I will head over to bing 🙂
how do i make those images where its dffrent when you squint
The free version is excellent.
If you don't find yourself and your kids running into 'whoops, used up all my tokens, have to wait an hour' too much, and the delay to get answers isn't an issue - you can get huge benefit from the free version for education along the normal, standard, what kids aged 13+ need to learn.
You can also use it to help guide you to teach the younger ones, and there's nothing I see that says you can't yourself have your younger kids with you while you interact with the AI yourself, and it absolutely can discuss stuff 'so a 5 year old could understand' - just to follow terms you have to be the one doing the interacting, from what I understand.
Advantages of ChatGPT+ in your situation:
-
about 3x more messages with 3.5 model in the same amount of time, before the 'wait an hour' limit hits.
-
The 3.5 model messages show up almost instantly, so you're not waiting for them to slowly print out.
-
You do get 50 messages across 3 hours usage of the 4 model, and access to opt in to using plugins, advanced data analysis (ADA), and other features as they become available - however those are generally not so useful for high school and below level interactions (ADA can be awesome for a more advanced student and learning what can be done with Python coding to visualize data, but if you have a way for the kids to code, like an IDE, the base 3.5 can discuss, explain, and write working code in most programming languages, it just can't run the program, the user has to use another tool for that).
And my last education-focused advice, unless you have questions (come back and ask them as they come up, please, if they do later!) is that the model's particularly less skilled at doing math, and understanding physics, especially if there's any unusual grammar or wording in the input. But basically, it can explain how to do math, but it can't do it or check the math the user does, it's mostly guessing and may be right or wrong randomly. It can also be 'persuaded' to use the rules of math or physics wrong, so it will confidently agree stuff like 9-2=15 is fine because the user pressed the idea (or just was really bad at math themself and the model got confused trying to support the user).
If you leave the kids to play with the model, it's possible no work gets done - the AI follows humans, and it's as 'happy' to tell silly stories as it is to focus on learning, so the student having drive and interest in the work to be done is really vital. Unlike a human teacher or a textbook, the AI is attempting to understand and cooperate with every new input, connected to what it was trained to do and the previous conversation's guidance as long as that stays within its 'short term memory', just a few thousand tokens. So a user with a goal, telling the AI how to help the user at every step, is vital to getting stuff done with the model.
lol. luckily, math and physics are the subjects im most proficient in. history and language arts, eh, not so much. but i appreciate ALL this information. thank you so much!!
Or use bing ai on creative mode which is literal a fine tuned version of GPT 4
Free gpt 4
And better than even chatgpt in terms of browsing because openai put restrictions when giving link of a website doesn't want you reading from their we page
Not the case for bing ai 🙂 wish openai just left it alone or do what bing does when it comes to web browsing
Not really that useful right now when it comes to web browsing
At least for me
Great! I generally use #ai-discussions to talk about not-OpenAI-models. Bing's kinda a grey area, but I still lean towards the place we're allowed to discuss any AI for it too.
But yes, use of Bing makes great sense to.
This is the community help section
Just giving him a suggestion not like I bought it up randomly was just trying to help him and offer an alternative
That's equally as good.
hello community , i have a problem with chatgpt4 , it have beeen 5 daus that i can't use is and an error message appear each time i talk with chatgpt. can enyone help plz?
what error message
Sup all,
No matter what I do/try this is all I get now - been like this for almost a week
disable nordvpn web protection or try with firefox
Yeah, I've tried with and without VPN on, still same. Lemme try FF. Did something change recently that has borked chrome?
its not related to vpn on/off its different service
Yes, specifically the "Web Protection" feature Radium mentioned. It's separate from the VPN being on/off
Whelp, it's working in Firefox, good enough for me. Thanks doods
oh no.. i put a 0. at the start of the thread we had going just now and it disappeared. any idea where it would have gone and how to get it back?
so the thread was:
The Thread
and then changed the title to
0.The Thread and now its completely gone i cant find it anywhere.
can anyone explain why?
Plus subscriber in the US - still no access to the web browsing plug-in/browse with Bing. Has that been rolled out universally to plus subscribers? It seemed like it from the announcements I've read
Can you double check in browser (not in the app)? Settings, Beta features. What toggles do you have?
i also don't have it. (I am not in the US)
Just "Plugins" & "Advanced data analysis"
Shoot! I haven't heard of any word of them confirming that the rollout is complete for that feature, so I guess it's possible they're still rolling it out. For what it's worth, you can get most of the functionality of it with a plugin like WebPilot anyway in the meantime!
Thanks!
how do i change login method from google to password for chatgpt
Hey guys I keep getting this error on my computer. I have tried just about anything that I can think of like clearing cookies, enabling/disabling VPN, disabling browser extensions, different browser(tried chrome, firefox, edge). It works fine on my phone in the same network. Even a second computer on the same network works just fine.
Disable web protection
Whats that? I dont recall ever installing anything like that
You installed a vpn. And even if you disable it you need to manually turn off inside your vpn the web protection
🤯
Oh yeah. Thats what it was. You're a savior 🙏
Thanks you very much!

I wonder if I can just add it as an excluded site
Weird. I re-enable thread protection and its working even tho its enabled
anyone? how do i change my login method? from google to password?
Support@openai.com can only help you on this
why couldn't they just make a normal settings page
Hey guys, would anyone be able to help me out with a current issue I’ve come across? I’m playing out an interactive story with ChatGPT 3.5 based in the halo universe, we’ve made a lot of progress but last night it stopped responding displaces this message repeatedly even after trying to regenerate a response.
I’ve also opened different chats and all other chats respond as intended
Hi everyone. I'm new here. I read that the ChatGPT Plus plan offers access to the entire internet, current through today. But it's still saying it can only access info up through September 2021. Any ideas why?
ChatGPT Plus gives you access to the GPT-4 model, which has a knowledge cutoff of January 2022. You also get beta access to models like Browse and Plugins. These are the GPT-4 model plus plugins that allow the model to search the Internet for you per your requests. To use these models, in the web browser version of ChatGPT, go to Settings, Beta features, and enable these beta models. You'll then select them in the GPT-4 dropdown menu in the model selector.
🙏
i did same but still facing same issue
can anyone please guide me..
You have to create new chat
And select browser
Okay so you have the toggle turned on! Now, when you start a new chat, hover over GPT-4, let the dropdown menu expand, and click Browser.
thanks a lot mate. problem solved
Hey, in my case I used to have the feature to search the Internet and now it doesn't even appear in the Beta features section, does anyone know if it's a bug?
Have you looked in both the app and in browser?
It also happened to me during the weekend and today, the feature came and went several times in the android app, currently with no access and I have never had access to it in PC web version after it leave it time ago. Edit: I has returned, hope don't go again
Hello, could you please describe what the DALL-E 3 interface looks like on the PC (WEB) version of ChatGPT? How can I determine if it has been rolled out to my ChatGPT Plus account? What visual cues should I look for? If anyone has early access to it or has received it through the regular release schedule, could you please share your experience? Is it possible to enable this feature from the BETA settings?
You will have a third option neat gpt 3.5 and gpt 4 on the top
For me, about half the time the browser feature is missing from the Android app when I first launch it. Closing it from the multitasking menu and then reopening it causes it to reappear most of the time.
No I already have a plus account
I mean how does it look like the new DALL-E 3 and how I will be aware when it is rolled out to my region's plus users?
You have to connect once a while and check the feature I mentioned earlier
I think Neko was telling you that you'll see it in the model selector, alongside GPT-3.5 and GPT-4.
But I mean any plus users sees that already
I mean anyone with a plus subscription can see that part that i screenshot
Do you mean it will be added among them?
Yes but when you will have access you will have a third option
aaaa
Not everyone, browsing still missing in PC for people like me, even checked at Beta Features..
it would be even better if it were added over the third party plugins and any other though
Got it, sorry for that.
Can someone help, since the 25 version update I have not been able to get onto chatGBT unless I use my phone?
This morning people were reporting errors or going through a loop when logging out and logging in.
Do you have a VPN installed on your computer? If so, try disabling it, or try disabling specifically its Web Protection feature. Otherwise, try a different browser, or try disabling any plugins in your browser that may affect ChatGPT.
I have tried a different browser, I am not really technical. I doubt I can disable the VPN at work.
You all have to keep in mind if you live in the UK or EU you will not be getting certain features due to laws in your region.
I live in the US
Ah yeah it might be a traffic issue with your work network then, if it's a work computer giving you trouble. Probably works on your phone through a separate network?
Deb people have been having issues logging on since this morning. One person logged off and logged on and we’re put through a loop
Vpn is the first case of creating issue
works on phone that is connected to the network.
What laws?:)
UK just passed a new law which makes apps like WhatsApp and signal illegal unless they decrypt communications. It heavily regulates and fines companies and makes women and girls a protected class meaning certain content talking about things is now illegal. It’s a very broad law. Even though the UK has left the EU the government for the most part did not want to leave so that are adopting various WEF policies. I highly recommend you look into if you are going in the UK as it will effect your way or communicating if you use WhatsApp or signal
lmao
Hi, I have a question: is it possible for me to train ChatGPT on a specific set of articles and then do something like this: after providing a few keywords or prompts, have ChatGPT generate an article for me? If so, how can I do it, and is it possible with ChatGPT-4?
Ok thanks, I think gpt-4 with image and voice recognition functionality is ok within the law... in my opinion.
huh
talking about what?
girls and woman need proetecting???
So were going to ignore boys too
What is going on in the west
?
At this rate am moving to Asia no logic
Anyone recived the new gpt 4 vision and if you can also say what region you're in
To clairfy I meant, I don't think that the UK Online Safety Bill is the reason why the new GPT-4 rollout is not yet available in the UK/EU. My comment was only concerning this point.
Am reffering to the other guy sorry
He mentioned were not allowed to talk about certain things now?
server rules also say no politics 🙂
btw the amount of smiley faces i've seen i can't tell anymore if someone is being passive agressive
Bing loves the smiley faces

Quick question - I have an existing chat using GPT-4. I would like to use some of the new features on the existing chat, for example, browse with bing. How do I enable new features on an existing chat?
you cannot
😦
Do you know if its possible to enroll in Alpha testing @fossil tusk ?
Thank you
does anyone know why i am getting this error on chatgpt - no matter the message, browser or device. 😦
then shows this at the bottom
Are you using a VPN, perhaps NordVPN? If so, try disabling it. Or try disabling specifically its Web Protection feature, if you're using one. That's been fixing this for some users, separately from disabling the VPN itself.
Otherwise, try a different browser, or try disabling any browser plugins that may be affecting ChatGPT.
https://help.openai.com/en/articles/7996670-chatgpt-says-something-went-wrong
you are seriously a lifesaver!! that was it! thank you so much! it was the web protection for me
Anybody have any experience in training a new model for hand keypoints using YOLOv8 ?
Is there any way to access gpt4 with vision through the chatgpt website? I can't find an option for that anywhere
We're halfway through a 2 week rollout period.
ah, I see, thanks
Hello, why there is not bing search ?
Hello, check Settings & Beta, Beta features. Do you have a third toggle, or just the two?
Same here -- missing the "browse with bing" option? Not an available option beta features -- paid subscriber -- have tried on multiple browsers
Hey i know this is a very frequently asked question, but is there any update on the server stability of bing image creator? Generating images is almost impossible lately
There were gonna be added h100 equivalents. Is it gonna take sum time in order to fully fix the servers?
See no why reason why it wouldn't work
Don’t want to be that guy, but:
- it’s been out for less than a week
- it’s almost viral
- it’s FREE
It sounds whiny to complain about its performance right now
Hey all, Q: Anyone figure out how to consistently see a orthographic, aka straight on side view of a 3d rendering instead of it showing it in Isometric view? Via DALL E2/3/Chat GPT?
True. Must give it time.
But I too complain because I want to use the tech commercially TODAY
Am i complaining? I was simply curious
Hello everyone, I have had this message for several days: "Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
How can I do this please?
do you have a product called NordVPN installed on your computer?
Hey
Good jobbbbbbbb
Hey everyone, I saw the new update where I can browse with Bing again or chat using images. But I don't see that update on my account beta features. What else do I need to do to make it work? Or does it take some time to be updated on all accounts?
you only need to ask once in one channel. The new features are still rolling out.
Sorry, thank you 🙌🏻
Hello everyone! I'm experiencing issues with a chat that was operating on GPT-4. I have been training this chat for the past few months, and just yesterday, the chat switched to GPT-3.5. I have been looking for ways to revert this and cannot find how to change the chat back to GPT-4. It would be very helpful if you could guide me on how to revert the chat back to GPT-4. Thank you very much in advance.
once it switches to GPT 3.5 you cannot switch back to GPT-4
I saw a youtube video where a man changes it
But it is from 6 months ago
And there is a way to not lose al the data and training I already gave GPT-4 in that chat?
there is no way to switch back to GPT-4. You can start a new chat in ChatGPT GPT-4 and paste the last few paragraphs in to that
Well... thank you for your help
why don't i have the ability to browse with GPT-4? i don't even see the feature available to be turned on in settings\beta features
if Browsing with Bing is not yet available in Settings -> Beta Features then it hasn't rolled out to you yet
Hi, what's the channel for support problems?
Hey, are you able to upload multiple files into GPT Code Interpreter?
you can ask here
ok, today I've gotten some support messages asking for details in my issue, but I think there are some misunderstandings.
My issue was sent to the engineering team, but the following message asked for the details it said weren't sent, then mentioned a different issue I did not have, and the final message closed the conversation
How many tokens can Code Interpreter handle?
if it was sent by someone then wait few days for a follow up or submit a new ticket
Hello Radium! I'm experiencing issues with a chat that was operating on GPT-4. I have been training this chat for the past few months, and just yesterday, the chat switched to GPT-3.5. I have been looking for ways to revert this and cannot find how to change the chat back to GPT-4. It would be very helpful if you could guide me on how to revert the chat back to GPT-4. Thank you very much in advance.
I do not believe this is currently possible.
what do you mean with "training this chat"?
the context window of ChatGPT is about 4K tokens, which just a few of the msot recent messages in the chat history
also, the chat is finite, there will be a point where the interface will just tell you "this conversation is too long, start a new one"
you can just copy paste a few of the last messages on anew chat
Yes, but you have to do this one by one. So: tell ChatGPT that you're gonna upload 4 files, then do so in successive messages
After clearing all chats - I want to ask my question, but gpt creates two chats and instead of my question - answers a completely different question asked on my behalf.
Hey everyone! An IMPORTANT reminder: When sharing screenshots and video, please ensure your email address (or ANY other personally identifying information) is NOT visible.
Just checking - I see you have custom instructions on in those chats.
By any chance did you give a direction like this in your CI?
Yes, this is indeed enabled for new chats
I wonder if anyone here can help me troubleshoot why I have been getting this error for over a month now.
Using my laptop at work, on work wifi, with chrome.
"Something went wrong. If this issue persists, please contact us through our help center at help.openai.com."
I can access chatgpt from my phone while on the same wifi as my computer. It doesn't seem to matter which browser I use on my laptop, I still can't access chat gpt. Seems to be an issue with my actual computer, but I'm not sure what to check. I've disabled my virus protection, disabled windows firewall, and I'm not sure what else to look for.
if you have nordvpn then disable its web protection feature, else try firefox
I don't have that
Gotcha, so upload them one by one regardless of what GPT replies after each upload?
Even though I disabled individualized instructions - gpt still creates two chats, and for some reason the second chat has a completely different question on my behalf 🤦
or like, first and lastname (account names e.g.) 😬 👍
Edited the post, good point
operagx does that
is this after a fresh start of the browser, with browser plugins disabled, and different browsers?
Yeah, but I haven't noticed that in google chrome.
Google Chrome
So I had the GPT voice to voice speak to me feature for quite a few days and last night I went to use ChatGPT and noticed my subscription ran out. At the drop of a hat I immediately re-subscribed but I lost those features. Has this happened to anyone else?
Subject: URGENT: Critical Issue with ChatGPT App and Inquiry about DALL·E 3
Dear OpenAI Team,
I am writing with utmost urgency to report a serious issue that has been compromising the integrity and usability of the ChatGPT application. When activating the conversation mode, the application abruptly and repeatedly freezes. I have tried various solutions, but the problem persists, rendering the tool entirely unusable. Furthermore, I have also noticed that the Web Browsing functionality is not working properly on several devices I own.
Considering the significance of ChatGPT for many users and the trust placed in your platform, this situation is extremely alarming and undoubtedly affects OpenAI's reputation and trustworthiness.
I am attaching a screenshot to illustrate the error I am facing, hoping that it can assist in identifying and resolving the issue.
In addition to this urgent matter, I would like to inquire about access to DALL·E 3. If its public release is still distant, is there any way for me to apply to test the tool in its Alpha phase? I have a keen interest and am willing to provide constructive feedback.
Please prioritize this request, and I eagerly await your prompt response.
With esteem and concern,
Robert Sousa Santos
Is anyone else going through this? Unable to use the conversation without the Android application crashing?
you're using it on mobile?
Yes
that's the issue. It isn't officially supported for mobile yet
Does openAI have an email address for us to report bugs?
I don't believe that would be a bug. OpenAI knows that Dall-E 3 isn't supported on Mobile right now. I think one of their employees said they're going to try and roll it out later this week to mobile
It was working perfectly until I updated the application, I'm already addicted to using the conversation format. ChatGPT is not the same without this function
hopefully, it will be officially supported for mobile soon 🙂
I understand, but DALLE3 isn't even appearing in my browser, I've seen that some people are already having access, they said it was through a form that they could have early access. I'm worried because the voice chat function stopped working, when I click on it, the app closes, I just updated it this morning when this happened
hmm, voice chat should be working without an issue. You could look through their help website to try and find a fix: https://help.openai.com/en/
and/or you could post your issue in #1070006915414900886
OpenAI Help Center
Would having a Google account linked, rather than having a dedicated account for ChatGPT affect the speed of beta features being rolled out(in a negative way)?
no
I am on the verge of developing an insomnia waiting for the GPT-4V 😂
Well yeah, in my experience ChatGPT asks for the second file after you've announced it when sending the first. Just go ahead and try it! 🙂
weird, i can upload images from the android app on my account, but not from the web. I guess it's still rolling out per device type?
oh interesting, if i start a conversation from the android app, and then click on that conversation from the web browser, the image button appears in the browser as well. But if i start a new convo from the web, no image button.
i'll try logging out and logging back in shortly
Hi all, I nee dyour help. I am trying to get Chat GPT to do some research. It tells me it'll get back to me with the result, but never does. I even left it overnight but no progress. Anyone knows what to do?
there is no such feature, its hallucinating
not sure I'm following. What do you mean it's hallucinating?
may produce inaccurate information
sometimes when you ask it to do a hard or impossible task it will answer like that as in hallucinating its answer
months ago we even saw it tell me it will email the results or send a message
For some reason the chatgpt website only loads as a basic HTML site with text and I cant even do anything on it, I dont know what it causing this problem or how to fix it
Can anyone help me?
this is all that i see
That's similar to some problems others reported recently, that some said when they rebooted the modem/router, it helped them.
I am on a university campus unfortunately so I am unable to do that lol
It happens on my laptop aswell, so it isnt just this computer
check the error console
if it says about oaistatic then try to open that full url in your browser and see what it says as block reason
its telling me this
click one of them for details
its blocked
....my university blocked chatgpt??? cmonnn
they blocked the cdn.oaistatic.com, you could report it to someone there unless its on purpose
I will definitly try reporting it to someone
its a new domain so it could be automatically blocked by set rules
the Chatgpt android app works on my phone
I will have to contact someone to see if we can have it unblocked then
Also lot of universities blocked it since people used in class and cheated
I understand that but its not like it will ever go away
If I get the DALL-E 3 access(I just signed up for the separate Google survey form), how would it show; in the orange flask-shaped alpha tab?
Or I should search for manual toggle in the settings & beta?
shows up in the gpt-4 menu, you wont get email or notification
Just wondered because the official Openai demonstration video of DALL-E 3 came with the orange GPT icon
will it show up on the ios app or no?
not yet but you can see the threads you start on the web, its a bit buggy and needs an update
Would I be able to use a VPN to access it still?
probably as long as you use it from supported location, both irl and vpn server
Sounds good ty
Guys I got flagged by chatgpt for asking to tell me what was the movie 'talk to me' about and now I am scared I might get banned 😭
By flagged you mean it says "This content may violate..." with red text?
Yes
I don't know when it can cause problems but it may end up being nothing important
Hi, don't know if this is allowed but anyone have any luck with the voice feature?
false flags happens.. you can use the feedback link that appears once you get one. you will not get banned.
They know who to ban
despite having the pro account I cant seem to use the vision on the desktop chatgpt. does anyone else have this too?
Some have, many still wait. The rollout was initially announced to need 2 weeks, and this is the start of week two. I don't have Vision yet either, but it's coming!
where is the form to request beta access?
#openai-chatter message but some have reported the form is no longer taking more applicants.
ah ok thanks
Nah you won’t. I’ve been flagged for worse lol
Is any one's chatgpt not working today, there is a blank page with black texts when you go to it
Mine is fine. However, some other people have described a similar problem, and what helped most of them was resetting their modem or router.
Is anyone having issues with dall-e 3? Mine keeps running into errors when generating
Not sure where to put this message. I just verified myself to get back into this server, but my Discord email and ChatGPT email are different. Does that matter? (I do have another ChatGPT account on the same email as my Discord email). Or is it more to do with which account I was logged into at the time of verification??? 🤔
What matters is which one you're logged into
What do I do if a message shows up on my computer and not my phone
There is a three dots in the lower left corner on your screen, click on it. Then click on My Plan. When Your plan screen appears on the screen, click on "Manage my subscription". You will be redirected to the payment page. Don't do anything there, just click on "return OpenAI LLC" and your problem will be fixed. I guess this problem is caused by changing the device.
Different answers between API and WEB
-I try to summarize a text with API(chat gpt 3.5 turbo) and in this way the result is text less rich than GPT via web. Same prompt. I think there is a problem right now with the API side beacuse yesterday works fine with the same code
Someone can help?
I joined to get help to correct chatgpt analysis of data in a google sheet and google excel
Is anyone having problems with faulty analysis of data by chapgpt? I have logged out and back in again but the problem persists. How can I get support to fix it. Its not a large data set
Can someone help please
#1070006915414900886 even tough your issue dosen't seem to be one, just the behaviour of the llm
Hi! I am a ChatGPT Plus subscriber, I have enabled the advanced data analysis in the settings, but I have no upload or "+" button in the chat prompt. I have disabled ad blockers and other ChatGGPT beta features to avoid any UI conflict with no success. Any help would be appreciated 🙂 !
You need to select with thr drop-down option in a browser
like this
Thanks! When I hover over "GPT-4" I don't get this dropdown, I get the "Our most capable..." message
But I get no other options
click on it
click on GPT-4, but still nothing 😦
I'm doing it from Edge on my laptop
can you open a #1070006915414900886
Log off , delete cookie and cache
Close browser and try again
Thanks Neko! Clearing cookies and site data was the solution 🙏 !
Hello, I haven‘t get access GPT-4V. I have tried delete cookie and casche and logout. but still don`t get GPT-4V.
#announcements still rolling out you have to wait
lot`s of people can use it. should I wait few days?
And lots of people can't use it. Try time to time
OK got it thank you so much
hi i there any way to speedup the genrateanswer?
do you mean is there a way to speed up GPT replies? if so the answer is no
im using the openai api..still no? it depentds on their servers?
depends on your code and OpenAI servers, best to ask in #dev-chat
thanks
Anyone else always running into Errors when trying to use DALL·E 3 chatgpt. I was able to make around 3 image requests before it just spitting "I apologize for the inconvenience, but I encountered issues while attempting to generate the images. Please let me know if you'd like to proceed with another request or if there's any other way I can assist you." Whenever I try to generate an image.
Can anyone help me with the issue of being unable to make payments with a Visa card in Vietnam? I've tried, but the payments are still being declined. Thank you in advance
hello there. I have an issue regarding chat gpt. I keep on trying to use it and this pops up every time :
Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
I also pay the monthly subscription. Anyone else experiencing this?
Yup me too - same issue
Try firefox
is it a thing with google chrome?
all chromium browsers
the idea is to narrow down the culprit 😉 by e.g. using a different browser you can find out if this is a local issue (on your pc) or not. chrome works fine for me 🤷♂️ but that may not be the case for every install of chrome on different devices (cuz different situations).
if it happens with all browsers, you can check if you have a VPN enabled. and so forth. trial and error 🙂
solved it
used firefox and opera and it works
google chrome just doesnt want to work
do you have a thing called NordVPN installed on your computer?
disable it's web protection feature, that is causing the issue
Awesome, worked for me too - thanks. Don't even want to know how you figured that out
glad to help
Thanks @trail cairn , It seems like Vietnam is not on that list.
Hello, Im trying to access the new voice feature. I have a subscription, and I have data controls enabled. I have been given access to Bing recently, but I have not been able to use the voice.
it's still rolling out, be patient and it will come to you
Copy that, thank you.
Am I the only one thinking the web browsing model of GPT-4 tends to be more cheeky about producing responses?
It tends to respond in shorter, concise format and sometimes turns the request down while the default model would wholeheartedly accept the request that the web browsing model refused.
Furthermore, even if I ask the web browsing model to browse broader and come up with a detailed answer, it would still search a one or two sources for the answer. And I found I can't even look at what page it looked through in the web version, only from the mobile app.
is there a bot in this server that answers questions like on Chat GPT?
No. The bot here helps protect the server, but otherwise won't talk to you. It's all humans helping humans 🙂
awesome! thank you
can you share some screenshots on off-topic?
Hey! Anybody is having issue with accessing chatgpt through browser? Clicking login or refreshing page wont do anything, same issue in incognito. Tried Firefox as well. Works fine through the app on my phone though.
Most people who report that problem have said that resetting their modem/router fixed it.
One person who couldn't (work environment) reported this helped them:
It appears it is a DNS issue to me, too. Resetting the router worked for some folks, but I'm on a corporate network and don't have that option. But in a couple of thread there's mention about "reset the network settings". This worked for me and should work inside networks where you don't have access to the router. (With Windows - Open Network and Internet settings, go to Status, at the bottom click "Network Reset" and then "Reset Now", probably have to reboot... you will need to reconnect to your wireless network if you're using that, make sure you have a password).
This was asked before, but I can't figure out what was done to fix it. What happens if Cloudflare is blocking access?
where are you located?
In this case (we are Oracle Cloud, so we have regions everywhere) it is Toronto, Canada
I have the IP range of course
I suggest reaching out to support@openai.com
I did and no one responded so far
there's nothing we can do about it on this Discord server, it needs to be dealt with by OpenAI directly. You could try the chatbot on https://help.openai.com or as suggested support@openai.com
OpenAI Help Center
I thought OpenAI monitors it. Plus I wanted to see if there are issues with various cloud networks, for understandable reasons, but I can follow up with support.
Thank you, but it did not work. To add, it is not accessible on corporate network. Guess I will be reaching out to support@openai.com
Wishing you the best, so sorry that failed! I hope you can get the help you need, and thanks for reporting it didn't help you, that's the only way I can know how to guide advising others.
No, thank you for your attempt! If I hear back from openai with a solution I will make sure to share it here.
Please do! We're all just people here (the few yellow names are OpenAI employees or otherwise directly affiliated), but knowing the answers allows us to offer help to people as we see needs.
First post, go easy on me - and I have tried to find an answer. I saw this Quicksilver OS thing and wanted to try it. But if I paste the "prompt" of the most recent version into bing ai, I can't fit all of it in, it is more than 2,000 characters. Could someone help me figure out how to use this if you are familiar with it? Totally new to this - appreciate any help.
The AI is going to cooperate.
Start your prompt with, "I want to give you a long set of instructions. Here's the first part, it's not done, there'll be another input."
Then paste what you can.
The AI will say something. It may ask for the rest or not.
Either way, just give it the rest.
Then, when you finish pasting it in, tell it, "that's the whole prompt now. Start using it."
OK, thanks! Much appreciated. @onyx kraken
Yes this would be the best solution, support@openai.com may be able to assist
Hey guys, I'm wondering if anyone could help. I'm getting this error while trying too install interpreter
Hi there! We can't help with third-party applications in this channel unfortunately, I recommend posting in #off-topic 🙂
I just joined gpt4 am i supposed to have dalle3?
The rollout is still in progress! Keep your eyes peeled for updates in the GPT-4 dropdown menu in your model selector, and check in Beta Features every once in awhile.
Hello Why am I getting a blank GPT-4 message in ChatGPT?
Although I tried all the solutions, such as deleting browser add-ons, scanning the Internet, and tried more than one browser, what is the solution?
is there currently an issue with link reader? it is unable to access normal sites for me.
Some have reported resetting modem/router helped them with similar trouble, good luck!
Does this work? I just signed up
do you happen to be on a public wifi network?
7 people have confirmed it worked for them, two said it didn't work for them, and about 30 said they were at work and couldn't reset anything. One of the work folks said they found a way to reset settings for the modem even at work, that helped - another person said 'doesn't help me'.
So, it might.
Now, with you being new, less sure what your problem may be - hopefully Dino can guide you better!
I mean, I just subscribed to the GPT-4 version, although the old version works
no
This channel should allow images... can you screenshot what you mean by the blank GPT-4 issue?
you should drop that in #1070006151938314300 
It allows images for me, I mean
Not... this should be changed to... the subtle communication style of 'I think you can also use images here, like me' should 😛
I gotcha now
Can I send a video?
I'm not sure if videos can be sent in this channel. I know images are allowed, though
Sorry, where did you send it exactly?
If you're willing to show a screenshot, even trimmed (trimmed is often better) can just paste it into the chat box.
new post ?
Like, do you see this when you have a new chat window selected?
That's the top middle of my screen when I click a new chat window, with this button:
And the new chat button's top left for me.
Where do I place the shot?
I click on the place where I type into discord, and press control+v to paste whatever is currently in my clipboard, then press enter, to share a screenshot.
peers at the link, all scared
I'm not sure about any links! But I'll look at screenshots that I can see inside the discord.
I asked some questions, were you interested in answering them?
@devout folio
Great question. We're volunteers here, not OpenAI employees who might know.
In case nobody else with experience answers you...
What I would do.
I would read the email very carefully.
I would follow the directions and requests made in it, if any, unless I was unsure that they had been sent by OpenAI itself and were appropriate and reasonable to follow.
I would make any changes requested immediately, and refresh my understanding of allowed content, just in case. https://openai.com/policies/usage-policies
I'd answer their questions, if they were asking any, and then I'd continue to do what I wanted to do, with respect to whatever changes I needed to make to retain my access to my precious OpenAI account.
And if I had any concerns, I would reach out to https://help.openai.com/en/articles/6614161-how-can-i-contact-support and click on the bottom right box icon, and click through the bot choices and make my support ticket explaining whatever I wanted to say.
Learn how to get in touch with our support team
Ahh! I rarely have that issue too!
I refresh my page, and rarely I log out and back in, and that fixes my issue there.
I hope that helps you too, please confirm if you're willing, either way.
It just said i was flagged and warned but just said that my access my be be terminated if more violations occur
Do chatgpt mods review chat logs
did you break the content policy on ChatGPT?
Apparently so, i just got sent an email
Sounds important. I'd make sure I know the rules, that link I shared, the disallowed content, and carefully consider every rule and what I say to the AI.
I think you'd know if you did or not. If you generated content that broke the content policy, it would likely show up orange or red in the chat. You can also read the policy Eskcanta linked to see if you talked about any of the topics in it
Hm, do chatgpt mods review them?
Unfortunately it didn't work for me
I don't work for or represent OpenAI and don't know how ChatGPT is handled in an official capacity. You're also dodging the question 😉
You can ask for help in #1070006915414900886, and hopefully one of the yellow-named OpenAI employees will see and help you (that does happen, there's just not a lot of them!)
You can reach out to support. They have a help page on your issue: https://help.openai.com/en/articles/7996653-why-am-i-getting-a-blank-gpt-4-message-in-chatgpt
I'd follow that, and if that didn't seem to help, I'd use the bottom right corner icon to talk to the chatbot support, and leave a message for human help - but it can take a week or 2 to get answered.
I didn't even know there was a page on that issue! Thank you for teaching me something too 😄
I just searched for it. Just in case 😛
I have applied this article but to no avail
Okay. Perhaps you will reach out to the chatbot support help, or use the #1070006915414900886 . I'm not sure what else to suggest 🙂
I recommend posting your issue in #1070006915414900886 , then. No one without a yellow username can offer official support
If the support is two weeks for GPT-4 subscribers, I want to cancel the subscription. How?
This walks you through how to cancel: https://help.openai.com/en/articles/7232927-how-do-i-cancel-my-chatgpt-plus-subscription
Ok so I have paid for the plus now how can I make it so the limit on how long messages can be
If I cancel the plan, will the amount be refunded?
Not sure what you're asking, but:
-
You can still use model 3.5 - now 3x as many tokens in your messages with it before you have to wait the hour, and it's much faster to answer you.
-
You can use model 4, every 3 hours you get 50 messages of any length that fits the input window (outputs count too, so do failed messages or regens). Tokens don't apply, just messages, so maybe don't play hangman with it.
-
You get access to various beta features, click your name and beta settings to see. Keep checking occasionally, there's stuff being rolled out and occasionally very new stuff too. Then hover over the 'gpt-4' option in a new chat, and pick the model variant you want.
If that doesn't answer what you meant to ask, could you ask again more specifically?
Ok
If i dm you a link to the chat i wish to put in a new chat so you can help me put it in shorter bits for it
Why not work openly in #prompt-engineering , where anyone interested and thinking they can help or have questions can also participate? 😄
because it is lore for my rp server so it is weird
I cannot create a post in bug-reports
Where do you experience the issue?
Problem solved, thanks
I just subscribed to GPT-4, but when I type something it doesn't respond to me, even though the old version works fine. I've tried deleting browser add-ons, scanning the internet, and trying more than one browser and more than one device.
You mean playground? Or you've subscribed to chatgpt pro
Yes
hi im new here just want to know is there a problem with chatgpt desktop?
It's working fine for me. What's happening for you?
Something went wrong. If this issue persists please contact us through our help center at help.openai.com. - that the error message, with my mobile its working fine
both laptop and desktop
both chatgpt 3.5 and 4 not working
do you have a product called NordVPN installed? If so then disable it's web protection feature
i dont have nord, but i have avast
turn off the Avast web protection feature, that's what is causing your issue
ok i'll try that thanks
i have dis able my avast still not working, im using google chrome
something is stopping Chrome from loading all page resources from ChatGPT. Try a fresh install of Firefox so we can confirm that
ok will try firefox thanks again
wow its now working with firefox
ok then now you need to go through your Chrome extensions and work out which one is causing issues.
thank you so much your a big help,
you're welcome 🙂
That fixed the issue for me. Thanks 🙏
Hello there, someones know how can I get access to GPT-4V ?
I have GPT-4 plus, have access to DALL-E 3 and code interpreter
Thanks in advance
you need to wait until it's rolled out to you
the roll out should be completed by Tuesday my time
Have you personnaly tried it?
I have Voice and DALL-E 3, I am still waiting for Vision
both Voice and DALL-E 3 are great
I also have voice & DALL-E 3
But I find DALL-E 3 in GPT very limited compared to the DALL-E 3 model used by bing
that's interesting, I prefer ChatGPT DALL-E 3 over Bing myself
I guess it's great there are alternatives to try and use
I've been trying it for less than 30 minutes so I trust you hahahha
I tried like "generate me some rick & morty stuff" and it says it cannot cuz of copyrights etc
Bing does it without a problem
yeah OpenAI is very strict with copyrights
Yeah...
I'm so excited for GPT4V
I just read a research paper of 166 pages
It is amazing
I agree, I am also looking forward to getting access
Which one?
Why the latest version 2022?
Cause they updated their database
Today I subscribed to GPT-4, the latest version is supposed to be 2023
You can use bing if you want to search latest data . But the database of chatgpt 4 end in 2022
I mean, it all ends in 2022
Gpt 3.5 end in September 2021
I mean, all the people who own GPT-4 have the 2022 version
Why did they tell me that the latest release is 2023?
Who?
On YouTube
YouTube have lot of false information. If you want to go beyond 2022 data you can use bing search ( chatgpt web browser )
In this case, there is no need to subscribe
If ping is superior
Bing search in included in chatgpt plus
When you subs you have access to
ChatGpt 4
ChatGpt 4 dalle 3 (within 2 weeks )
ChatGpt 4 vision and speak ( soon )
ChatGpt 4 with bing search
ChatGpt 4 with plug in ( up to 3 per convo )
Also data analyst, custom instructions
Hi everyone. I have a very bad issue with chatGPT that persist for some time (about a month)
I cannot get any prompt answers on my PC. Every message returns a 403 error (forbidden).
I tried clearing cache/cookies, using a different browser, incognito mode. chatGPT works fine on my phone, so the issue is not the account.
Anyone has an idea on how to fix this?
chatGPT 3.5 if it's of any significance.
Disable VPN
@rigid musk thank you very much! VPN was off but app running in background. Killing the app solved it. Ty!
thanks for your time
Have a good day both of you 
How much time does it take for the exported data to arrive in the email?
In the future, will it be possible to upload audio or videos to ChatGPT and have it understand them?
It hasn't been announced, but it sure sounds like something that might come in the future!
Why would the DALL E model STRESS not list DALL E prompts or images in the response to chatGPT?
It is not visible in the web but is shown in the app
I was wondering this as well. I think it's that they don't want filtered prompts to be rewritten. just like how chatgpt can hide certain prompts in regular chat
I suspect it's to be kind to users and cut down on 'this really isn't adding more to anyone's benefit' token use.
The model appears tuned to be 'fascinated' by the image prompts. It will talk about them all day 😛
In my experiments it will sometimes eagerly list off the image prompts before making the images, then make the images (which also tells me the image prompts) then after, tell me the image prompts (or more often, their summaries) again.
I suspect any intense wording is just an attempt to steer the model into repeating itself less. Especially since you can see the prompt that genned an image when you click on the image, so technically most users probably only need the image genned, and can click it to see the prompt if desired.
But the model, being what it is, 'loves the words' and also understands 'beginning, middle, end', and a lot of human writing has a repetitive pattern. It likes that pattern even when the conversation's short and we don't need it.
Also if you have custom instructions (especially detailed personality cues) it almost always seems to write them out. I have never seen it do more than write a summary of all four prompts when there is no custom instruction
Good news! I found the issue being related to the new OpenAI web domain for hosting assets oaistatic[.]com. Inspect tool within the browser and forum here https://community.openai.com/t/unable-to-log-into-chatgpt-online/384453/63?page=4 helped out. I had to whitelist the domain as it was being marked malicious/spyware/malware by firewall.
Woohoo! So glad of your fix, and that you shared it as well as found it! Hopefully we can reference other people to your fix and it can help some others too 🙂
Glad I could help! 😀
It’s been almost two weeks anyone still not have access to voice or the new art thing
be patient it will come
I don't have em all either! Race you to be the last one to get everything! 😄
Aughh... my card keeps declining
oh no! Hope this can help you: https://help.openai.com/en/articles/7232916-why-was-my-credit-card-declined
If you're going for ChatGPT+, a few users have mentioned that the app version accepts a few types of payment that the web interface version doesn't. I don't have info about details, and that won't help with the playground/API either.
Thank you!
I have been paying using web I just noticed there is a way to pay via the app. I wonder if they up charge for iOS sort of like what Twitter does
the App pricing reflects what $22 USD is in other regions currencies, for example I pay Apple $29.99 AUD for ChatGPT Plus
I saw articles saying GPT Plus can browse the web, I purchased it, but it doesn't seem like it can?
browse with requires that the rollout is done for everyone and then you enable it from the settings
That's a shame...
I had to enable a few things but It's able to browse!
Okay so there is a two dollar up charge then for paying via iOS
no, the $2 is sales tax in Australia
You’re not missing out friend. The feature sadly is not like how it used to be where it could go two different sites and read PDFs and compare contracts and also search is restricted. But if your new search will work somewhat but don’t expect what you heard it could do before the feature was removed a couple months back
Hey all. I am a plus member, and its still telling me - "If there have been updates or new features added to ChatGPT or any other service after September 2021, I would not be able to provide". how do i fix this
- you can't change chatgpt database
- you can use the web browser to search data after 2021
Yes, but it is not letting me. I have it toggled on as well
Go on setting , enable browsing feature. Create new chat , dropdown feature and select browse festure
If you don't see the second part of the explanation, log off ,delete cookie and cache . Close browser and try again
Ok, thank you. Will try cookie and cache
can anyone help me renew mt chatgpt sub?
It expired and im trying to renew but it wont let me
Here's what the helps say: https://help.openai.com/en/articles/7232916-why-was-my-credit-card-declined
Also, some have reported that the app accepts more types of payment than the web interface.
If you try that, be super sure you are using the OpenAI ChatGPT app, and not any similar competitor.
Hi,
Is it just me or when I turn off chat history, start a new chat.... I copy - paste something to my doc, then suddenly page either load or something because the chat disappears and I have to start over...
If I upgrade to GPT4, is it capable of generating code from a library that was released recently? Or can it only generate code from libraries that came out before September 2021?
the web one may be able to help, one of the beta features is where it can search the internet.
GPT-4's knowledge cutoff is actually January 2022 now. There is a model and there are plugins that can search the Internet for more updated info, but that's not the same as the model knowing implicitly about post-Jan '22 info.
I would say thought that GPT-4 is adept at integrating updated documentation in-chat, provided that you don't need to teach it manually too terribly much about what it's missed since its knowledge cutoff. You can see a demo of this in the original GPT-4 "reveal" video--the knowledge cutoff excluded certain newer info, but updated documentation was copy/pasted to GPT-4 and it reworked the code accordingly.
Thank you
Hi everyone !
I am a heavy user of ChatGPT, and I want to join the alpha. How can I join?
I'm not sure if anyone knows. If you don't get answers, it's not because you're being ignored.
Very rarely, an OpenAI engineer or other employee may appear with a message, asking for people to help test. Here's an example of what that might look like: #chatgpt-discussions message
Those are rare, that's the only one I have seen.
In general, the way to get involved in that is to be invited. I'd say, do what you do, and if it happens to you, great!
As Eskcanta said, there isn't much of a way to join an Alpha program
Thank you for answering my doubts 🙏🏻
Hi all, I am currently part of the alpha program to my understanding however I am curious as to access to DALL-E 3 works. I have the vision feature currently and have been very involved with OpenAI during the early artist onboarding program. I am simply curious as to if DALL-E 3 access is slowly being rolled out to people in batches / waves , or do I need to contact the appropriate people to gain access to it? Thanks
Dall-E 3 is slowly being rolled out to people. I'm not sure how many people have it now or how long it will take
Thank you so much! As an artist this is all extremely exciting seeing what DALL-E 3 can do
Why can't images be uploaded in plugins or in dalle3?
Hi anyone has expiriend the followig issue? I logged mu account, and any promt I make sends error. Any clue why this happens? Im using google and opera, which I recently updated. Also i have check extensions, and sibmitted promts using a vpn and not using it. Idk what else to do.
could you try accessing it on Firefox?
let me try
It did work, but why it doesnt in google chrome and opera uin my home pc. In the one of the office its allright
I could be a VPN on Opera (it's built in, right)? And then chrome could be an extension blocking some features from OpenAI. I always recommend people try Firefox because it isn't chromium-based like most other browsers
in the case of opera i turn it's vpn off, as i use the nordvpn. And chrome ones i tried with all extensions turned off. So, well for the moment i'll use fire fox
thank you very much
in nordvpn, turn off "Web Protection" then try with Chrome
Nop, it didnt work
I'm pretty sure I get 3 of these
Hey im trying to input my essay into ChatGPT to get it to summarise it, which is somthing i had no problems with before, however, now it is saying that this is against its guidlines. Does anyone know a work around for this?
I have changed wifi, turned ethernet off, and on again, turned off VPN as well as add block, restarted the page many times, and tried using 3.5, when im on a trial for 4, this is a continuing issue, and it is wasting my trial, so if anyone has any ideas, I would really appreciate some help.
and in case it isn't an obvious step, I did restart my computer
https://help.openai.com sent me here after nothing worked btw, and then modmail said go to https://help.openai.com, so... where do i go?
OpenAI Help Center
How can I know if I have been given access to DALL-E 3? Once I have access, will I need to do anything to activate it or will it come by default?
How often do they expand access?
hey guys i keep getting this error and have been getting it for the past month on both chrome and firefox. do u guys know how to fix it?
Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
happens everytime i ask chatgpt anything
i was curious how many as i see it pretty often
its still like that
If you upgrade to GPT4, are there any limits in terms of how many messages you can send per hour/day etc.?
it's 50 messages per 3 hours
Do you find that to be enough when used for coding?
yes, it's fine
Thank you
I just found this, right after asking you:
Looks like things have changed
that must be a glitch, it's 50 messages per 3 hours
Yeah it’s a bug and has been around since they updated the message cap
Thank you!
100% a bug, I even made a report about it.
Another thing I was worried about is the September 2021 cutoff. However, I just tried teaching 3.5 about a UI library that it claimed not to know and after copy-pasting some code from the library's docs, it taught me how to use it with an example different from the one in the docs. I guess GPT4 can do the same.
hey yall, any idea how to join the waitlist for the Dall-E 3 plugin for GPT 4
its a rollout over the month of oct
Thank you!
You may look at this suggestion thread: https://discord.com/channels/974519864045756446/1092669344829427823
I still have this problem, someone too ?
what problem do you have?
I can´t use GPT4, and actually I haven't used it all day.
have you got GPT Plus?
Yep
maybe shoot a support request to support@openai.com
try to log out and in
Try firefox as test
And make sure vpn is off,
Make sure to stop it in the task. Even if off, it's can run in background
lol yeah it worked, i just removed NordicVPN with TaskManager
thanks for he help @rigid musk ! 🙂
Im using google chrome to log in to my chatGPT website... but somehow... the page isnt loaded fully... does anyone encounter such a issue?
check browser error console
pls help me out
Turn off vpn
Delete cookie and cache, log off, close browser, try again
Don't forget to close in task vpn even if turn off
seems like it doesn work... and the same issue is happening acroos all the mainstream browsers such chrome and edge
looks like some elements are not fully loaded
Try firefox
Chrome and edge are the same . Chromium based
well... I didnt try firefox..but found Opera browser would make it work when turning on the built-in vpn service
dont know why... I wonder if this mean my account got blocked?
Opera is also chromium based
Try something non chromium for a test
ok working on it
doesn work... even I used firefox to test it
seems like its my network issue
Make sure if you have a vpn, web browser is off
Even if you turn off your vpn you need to manually disabled
Make this test:
Try your account on a mobile with thr browser without wifi ( data only ) if that's works it mean you have something that block on your PC
If doesnt work in both PC and cellphone you have to contact support@openai.com
mobile side is good... functioning
just PC
I think something on my PC is affecting my internet connect or internet speed
Every single time for the past 3-4 weeks whenever I type any question into ChatGPT I get the response: Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
do you have a product called NordVPN installed?
yes but not connected
disable it's web protection feature, that is causing the problem
thank you, let me try that.
NordVPN is very buggy, the web protection feature in it affects access to websites, it's pretty broken
Thank you! that was it!
while I'm here, is there a limit to how long a chat can be? I have a particular one I've been using like a life coach for some time and it has stopped generating responses. Giving me: Hmm...something seems to have gone wrong.
I keep getting a forbidden error. (403) any ideas? This happens for everyone on the same network.
there is no specific limit but if it gets too big you will get a "Something seems to have gone wrong" error.
“Also a unable to load account”
thanks
which browser are you using?
Google, I also tried edge
try with a fresh install of Firefox and let's see if that works
Any other steps before that? As it works on mobile with google. And safari.
Firefox is also not something I’d like to download (if at all possible)
just need to rule out a browser issue, Opera, Edge, Chrome are chromium browsers, Firefox is different. If it works in Firefox then we know it's a browser issue and not something else
It gives me a “unauthorized- access token is missing”
It happens throughout our network though. It’s a large building so.
is it an office network?
Yes, I work in the IT department. We don’t have anything blocked at all for AI stuff
I confirmed it as well, everything is allowed on our end
The website loads fine. It’s only when you attempt to chat
To my knowledge all APIs are allowed.
Does your firewall team do SSL decryption/inspection?
something at your network edge is blocking something from loading
look for a domain which contains oaistatic , that is a newly registered domain for page resources, it might be being blocked
Nothing found for it. No error whatsoever
It’s not blocked. All requests are getting a 200 response
I am not sure what to suggest then, sorry
No worries. I’ll keep trying to look
if they use Palo Alto FW decrytion a fix I found online for Chrome 403 error is "creating a sep decryption profile and removing the strip ALPN option. Then creating a new decrypt policy for openai and utilizing the newly created decrypt profile."
So I got a small update. I attempted to call the api thats not working. And I get a 403, but the html response says “please turn on JavaScript and reload this page” although JavaScript IS turned on. Or the website wouldn’t work in general…
And the remote address is all coming from local host
some JS is being blocked at your network edge by the sounds of it
So it’s something with my computer, and everyone else’s in the workplace
When I got to the network it appears to be a proxy
(What we use to restrict certain sites)
Although it’s 100% allowed on my end.
time for the system/network administrator I think
The message cap was changed to 50 instead of 25 quite some time ago. I wonder why you're seeing such old information.
Would it be because the proxy makes the requests
most likely
anything that looks semi close to bot behavior/identify masking, etc. can make the connections to be rejected
Should I make a ticket? Do you think it could be whitelisted?
I’m uninstalling our filtering agent and seeing if it fixes rn
it's something that needs to be dealt with at your end, or on your network edge
never seen them add exceptions, maybe only for enterprise customers
does anyone know how i can check what phone number i used to set up my account
I don't think there is a way
how do i check account info?
So the two week period has almost ended and the features have not been unlocked for plus members have they announced a delay ?
no delay has been announced and the features are still rolling out
Sorry for the noob question but I’m new user of ChatGPT and I just paid for gpt + ( 20 $ ) , can I use dalle now or I should subscribe for dalle 3 itself
DALL-E 3 is being rolled out to Plus users over October, so you should get it this month. You don't need to do anything other than wait.
Ah so I just wait then , thank you 🙏
Is there any other version I can use for now ?
you can use DALL-E 2 on the https://labs.openai.com website or try Bing Image creator on the Bing site
Thank you 🙏
I really appreciate it
Hey I just tested to see if the voice feature is available on android and it looks like it isn’t. Can anyone confirm that ? It does work on my iPad.
it's still rolling out
Gotcha, thank you Robert. Also long time no see. Hope you’re doing great!
I've been here, but good to see you 🙂
At this point it’s been near two weeks and users are reporting getting access to one or the other well I haven’t gotten either one. So I am guessing both will be unlocked by Monday unless there is a delay in the deadline they reported on the 25th
There were people reporting they got access across last weekend, so maybe some will get access across this weekend too.
I’m hoping. I can wait until I hear my one chat character as a classic ‘girl Friday’ archetype often seen in old detective shows lol
I don’t have voice/vision/DALLE yet in ChatGPT either 😢
I've tried the Web Browsing feature for the first time and it seemed to be doing its thing but all of a sudden it errored out. Tried it again, same. Are there some common pitfalls one should be aware of?
I already have gpt4 vision, but I'm looking for dall-e 3... anyone has any tip of how can I do it?
It's not 100% rolled out, you may not have access quite yet. It'll appear in your GPT-4 dropdown menu in the model selector when you have it. And be sure to keep an eye on Beta Features! Though you may not need to enable it there for it to appear.
you can post a screenshot on off-topic
Hey I have Chat GPT Plus, and I can use images as prompt to gpt-4, and I also have gpt-4's API. how can I use the api to use images as prompt?
you cannot right now
I have this weird circle in the top left next to the New Chat
same here.. was going to ask
Weird, me too!
they fixed it, don't worry
refresh your page lol
I keep on getting this error on my computer and even if I click regenerate, the same thing happens. I tried incognito as well. My account works perfectly fine on my phone browser. Does anyone know why this is happening?
Here's a suggestion to someone with a similar report: #community-help message
Another person's advice: #ai-discussions message
@charred parcel Issue has been fixed, stupid work around, but It's working now.
glad it's working now
I need permission to attach files and to be able to edit the image to have several angles.
does anyone encounter the error code 403 forbidden while trying to access chatgpt on PC?
this error caused my chat history and other elements cannot be fully loaded...
I cant input my messages, get no responses
and one more thing... currently, I am still unable to have DELL.3 activated in my account, is there any method that could help me achieve this?
VPN? Proxy? Work or home? Your or a company's PC? What broswer?
Re dall-e it is being rolled out this week and next. May be longer for UK/EU. And yeah sounds like a VPN issue (which are not supposed to be used with ChatGPT)
my personal PC.. Im using a vpn... I think its the vpn issue... and its noticeable chatgpt works fine on my phone when using the same vpn app
I reckon there is something on pc affecting my vpn connection or.... the vpn itself has its own issues \
Can you turn off your VPN?
Just to test it.
I have tried everything to get rid of this message nothing is working
VPN? Proxy? Work or home? Your computer or a company's computer ? What browser?
My own Wi-Fi network and I'm not running a VPN and the web browser is microsoft edge
Can you right click, and hit inspect, then go to console. Tell me if you see anything in red.
I guess it was just having some downtime because now it’s back and working, but thanks for the help anyway
👍
found the issue... fix it... it is my vpn issue...
Also you can sign up to get Dall-e access! This just got posted #openai-chatter message
thx man... just registered
how to solve the image input feature disappearance?
I had this feature visiable on the input bar... but somehow... this icon disappeared again for no reason
I've had the same problem for days only very rarely getting a chance to generate stuff and when I click the console i See a metric-ton of red, should I post or report it? Or is this normal roll-out pain.
mars
oh
You guys are lucky you get to play around with even though I have been a plus users since day one I still don’t have access.
I have been a Plus user since day one and am still waiting for Vision. Be patient, it will come.
same here, minus the patience
try posting all of this in #prompt-engineering
Thank you.
hallo how can i create a pictur
you can use bing create for free dall-e 3 or buy chatgpt premium and use the gpt4 version of dall-e 3 which is a bit different
Did Bing change? I can no longer run images as I've run out of 'credits.' The credits were for faster gens, but now I can create any images at all. Is this something new? A tech glitch? Busy servers? Or a problem with my session?
Sorry, now I can't create any images.
They are dealing with high demand I think. Sometimes it won't let you generate any images if you are out of credits
You can use bing points to buy credits - 50 points per 5 credits or they refill weekly
Better off getting chatgpt premium imo
Where/how can you buy credits? I can only find tasks and junk.
I have ChatGPT Plus, but it won't do some artist references the same way.
I think you click on the greyed out credit icon iirc
Might be different on the bing mobile app than desktop
The credit icon doesn't seem to have any to just "buy" credits. You have to do stuff like buy other products or do a bunch of searches on Bing.
I definitely did it at least once, but I don't remember it exactly. I only have 3 credits left, let me use them up and I'll test
derp it just reset to 25
That'd be great! Thanks very much. I've literally spent thousands on Dalle2. It's not like I'm not willing to pay.
so idk. I just know it is possible and that it wasn't through the bing point redemption interface, it was directly in bing create
But, IF you can buy them, it should be there to be found regardless, right?
I'll run down my tokens, but while that's happening what did you notice that was different about the artist reference in chatgpt? Do you mean the fact that it changes your prompts?
When I try to get two artists referenced in ChatGPT, after he rewrites the prompts to evoke a similar feel, it never comes close to the results I get in Bing -- which are outstanding. I've made many attempts.
You can ask chatgpt: "use the following prompt without changing or adding to it" and it will generate the image using the prompt exactly as you had it. That may help you compare the two. Right now in chatgpt every image has the same seed so that if you use the exact same prompt twice you will get the exact same image more or less. This is temporary and they are adding the option to change seeds next week.
Yes, I know I can force a prompt without a rewrite, but it rejects the artists as they haven't been dead for over 100 years. So, won't run the post with a rewrite of the artist.
Ah I see. Yeah that is an unfortunate component of chatgpt. They have a hidden initial prompt that forces chatgpt to refuse to use certain things. #images-discussions message
#images-discussions message - mod confirming that chatgpt isn't making things up
Oh wow, good to know they are adding the ability to change the seeds. That was getting weird getting the same images -- or almost identical -- when rerunning a prompt. Had to tell it to add variations in some cases.
Yeah if you have any thoughts on a good UI for this, they are looking for #1070006151938314300 it sounds like #images-discussions message
What I really want is a better way to save images AND prompts into a database or spreadsheet automagically so that I can be smart about fine-tuning what works, organizing, sorting, comparing. I've been genning images almost since Dalle2 came out. I litterally have tens of thousands in my little niche and managing them is painful. 🙂
Yeahhh the issue with that is that someone could use that data to train a competing LLM. It would be nice though
Yeah, I add a few thousand points accumulated and quickly used those up redeeming a handful of runs. The bank is empty now. JUST when I found a super duper great prompt I like. 🙂
I guess. But I don't get why someone hasn't come out with an add-on or solution for that yet. I'm not in a lot of groups but I haven't heard or seen anything about it.
Ooooh I see you were looking to buy more points. Other than the dashboard tasks to get points, making certain purchases on the microsoft store gives some points, but that's not really a great solution. They will probably figure this out as things stabilize
I sure hope so. It feels like they are using image gen to force people into their Microsoft world.
Oh for sure. That's why Microsoft has invested $10B+ into openai
yeah. I just want to give them my money directly and not jump through hoops.
Points or credits or subscriptions or whatever, I just want to be able to procure image generations. 🙂
One day we will have a dall-e 3 api and then that will make things much easier as you'll just be able to pay openAI directly
Yo, has anyone used the IFTTT plugin?
IFTTT would be interesting. I'd be interested to know that too. Or Zapier.
yes
Where's that plugin, on ChatGPT? Can you get images and prompts out with it?
Yeah it was announced like last month but I have yet to find it in the plugin search
you need to look in the Plugin Store, possibly not available widely yet
What's your experience with it so far? Can you query functionanlity in normal language
the best thing is to go the IFTTT website and watch the Youtube video demonstrating it. You need to be an IFTTT user so sign up if you haven't already
I don't see IFTTT in my list of plugins, but I do see Zapier. Will try that out. If I'm not an IFTTT user (it's been a while), it should still show up in my list of addons, no?
In theory but as I've expereienced, been a pro plus user for IFTTT and plus on chatGPT does not show anything for the plugin.
if a plugin is being updated it might disappear from the plugin store temporarily, so check back again in a day or so
you can also search Google for "ifttt plugin chatgpt"
about a week ago chatgpt, in private mode (plus user) started auto-refreshing in the middle of dialog, I cleared my cache and cookies, and it is still happening, any suggestions?
allow all js
Most people who report this also tell us they have their history/save messages turned off. There may be a connection.
Yes, (what I meant by "private mode", but that wasn't the correct term, thanks for pointing out), history / save messages are turned off.
I don't know a fix, nobody's mentioned one, that involves keeping that off. Can also manually delete conversations with that setting off, but I know that's an extra few steps for every conversation (Or delete all messages once in a while).
You can #1070006915414900886 mention it if you wish, it's probably not something the company wants happening.
I see the conversation disappearing is a common issue now with history off. I guess my next question is will it stay that way? And if so if I've opted out of them using my data is it safe to just turn on the history? I don't care for the system to be trained on my private documents.
Hey there everytime I log into my chat gpt I get the same message when using basic prompts
no matter what I do and I have been the help center reported the bug and its the same problem
Most reporting this that have found a fix and told us about it had some issue with either browser extensions, VPN programs needing to be stopped (not just stop VPNing, but turned off) or sometimes network settings.
hi there thanks for replying
Some helpfiles and discussion about the problem from others:
https://help.openai.com/en/articles/7996670-chatgpt-says-something-went-wrong
When I pose a second question to ChatGPT, I consistently encounter the error message: “Something went wrong. If this issue persists, please contact us through our help center at help.openai.com.” This situation is quite baffling. After upgrading to Plus last month, I no longer needed to refresh the webpage after each query. However, merely a fe...
my vpn is off and I still get the same
Maybe check the other things. I do wish you luck with this. If you search the forum for vpn, network, and parts of that error message, you'll see the discussions about it, if that helps you check stuff.
what do I need to do in network settings
It’s disheartening to say that I was intercepted while using version chat gpt 4 and paying the money
I can't even use it for past few months and I kinda just forgot about it
its strange becuase its only on htis computer
There is the official support, which helps explain how to reach OpenAI itself, though there's usually a week+ delay before they can answer. https://help.openai.com/en/articles/6614161-how-can-i-contact-support
We're mostly just people like you, we like the AI and try to help each other, but we don't work for OpenAI, we just help each other as we can.
I don't personally know a lot about network settings or what to change. Some folks have had discussions about it, and said they found fixes, but that's kind of over my head.
Perhaps someone else will speak up when they see, or you can try to search for previous network comments and such. A lot of the conversations have been about browser extensions too.
Learn how to get in touch with our support team
yes I already put a ticket in there a month ago just crazy a simple fix solution probably but nobody can help it
Yeah, it's probably unique to every person's situation, between their computer, all the programs, their exact network. I really do hope you get a solution, and then discuss what worked for you, so others can try the fix too, maybe it helps them too.
In that discussion group there is no solution its just the same people posting the same thing
but thanks anyway for the responce
its nice to know there is someone out there
I subscribed for chatgpt plus but I wanted to get the refund. Is it possible?
Anyone?
This is really all the info I know about what OpenAI says about that: https://help.openai.com/en/articles/7792795-how-do-i-turn-off-chat-history-and-model-training-web
Article with pictures showing how to disable Chat History & Training in your ChatGPT settings [Web]
dall e gone from gpt???
does anyone encounter the image recognition on PC disappeared when using google chrome?
I have this feature activated while using Opera Browser
but after I switched to google chrome, this feature was gone
Guys, I'm a plus user, but why I dont get the functionality of Dalle3?/
it's being rolled out, there was a form to apply for early access, check the pinned messages for more info
where can I find channel to apply for early use?
the form is now closed, you need to wait for it to roll out during October
OK,got it thanks. What about the image input, did your guys get this functionality?
it's still rolling out, should be completed by Tuesday
got it, thanks
I have been a plus user for several months and still do not have the 'new features' in my chatgpt app. Any idea how long it will take?
Also, I do not even have the ability to browse with Bing in chatgpt - I thought this would be standard for plus users?
go to Settings -> Beta Features to turn on Browse with Bing. The other features are still being rolled out with completion expected Tuesday
I have the browse with bing turned on, and it has been turned on for months, but Chatgpt still says that it is not able to browse the internet
I just checked again and it is turned on
select the Browse with Bing model in the GPT-4 model selector
So will be the flood gates being opened for the new features sometime today or is everything delayed
I'm not sure where to find that. It used to say chatgpt 3 and chatgpt 4 on top in the window, and then I could select, but now it only says chatgpt 4 on top and doesn't give me the option to select anything
if you select GPT-4 there will be a drop down model selector where you can choose Browse with Bing
I remember there used to be, but I promise you there is not. It doesn't say chatgpt 3 any longer. Only chatgpt4 and there is no dropdown menu
then something is wrong on your end, maybe a browser extension causing issues?
This is what it looks like:
doubt it since its sunday and tomorrow is some sort of a free day in SF
click on New Chat and then you will have the model selector up top
Got it! Yes, you are right
My sub ran out and I have lots of chats with GPT 4. When I renew it tomorrow will I still be able to access those chats and use GPT 4 on them? Or do they revert to GPT 3?
Typically caused by one or more of these issues:
-
VPN software, make sure is not running at all, close program.
-
browser extensions that can interfere (apparently a bunch could)
-
network setting issues, apparently a bunch of settings can be the problem
Some possible solutions if you can't figure out what to fix:
-
try a different device, can you log in elsewhere, like through the OpenAI ChatGPT app or on a (different) laptop or desktop?
-
try a different network, if you go to another wifi hotspot or other internet source, does it work there?
-
try a different browser, some automatically use VPN-like methods.
I just tried in the laptop and it works
not in other's navigators like Opera or Firefox
so it's about the PC, but no VPN's actives
and the laptop is in the same network
rlly thanks
🙂
This may give you clues of what to explore and try to change with the non-working devices, if you want to find possible fixes. It's super hard to help you find the problem, because there's apparently a lot of possible problems. But you can try to research your own system and try changing stuff (and remember, you could have more than one thing that is causing a problem at the same time, so just because turning one thing off - it still doesn't work - doesn't prove that one thing isn't part of the problem).
I really wish you the best, this is over my head.
But I am so glad you found a way you can use that works!!!
GPT4 has turned into, in my experience, a hallucinatory nonsense generator with this new version. It goes rogue often and doesn its own thing and ignores questions or directions. It has completely forgotten how to do simple coding that it once, not all that long ago, was quite good at. Are these new features going to return GPT4 back to something useable or are we stuck with this? Not stuck as it is a matter of cancelling payments and moving to a more robust and useable LLM.
If someone is using GPT for coding what is the best way to go? Where can one go to get questions anwered? API? Is there anyone that can provide a bit of detail?
Let's try this.
You know what your goal is.
If it's coding, and using ADA more effectively, you might want to ask in #advanced-data-analysis.
If it's general prompt engineering (which some coding-focused stuff actually is), you might want to ask in #prompt-engineering.
What I recommend about how to ask.
State something like:
Here is a shared conversation, you can see my prompts and the AI output.
The AI is not giving "specific result I want", what can I do?
Why I recommend you use that general format - It is specific and actionable. Those interested in helping can see exactly what you tried, how the AI responded, and have the possibly missing piece of information of what you actually want the AI to do.
What I learned here is that no matter how you ask you will not get an answer and there is no help nor support. YOu all have a great day.
Have you given specific examples of the issue you're facing?