#gpt-models
1 messages · Page 218 of 1
Bro is French he doesn’t know English
not that big of a deal 😭
I know, but how else would he learn?
Probably doesn’t wanna learn English
Gpt4 should shame him everytime he makes a mistake in his prompts
This should be a plugin
Should have used cgpt to translate instead of google translate tho
You can use gpt4 in french
I never make a mistake in my prompts.
I am looking for the new beta feature to continue generating a message beyond the maximum token limit.
how can I do that?
Will gpt4 shame me in french automatically? No? Thought so
You can't change thr max token limit
If you start the conversation in French and ask it to then it would. But then again that's not exactly automatic.
that's not what I read here
https://help.openai.com/en/articles/6825453-chatgpt-release-notes
After a long enough conversation it willl forget about it
dont know much about the french I took...
Plug in and browser available this week ?
Huh, I didn't know that. I've never tried talking to ChatGPT in a language other than English
If you prompt it in a certain language, it will answer in that same language
For any language cgpt forgets instructions
But it will keep talking to you in French, right?
Yes
Except if you had a long conversation then ask in English, it may not answer in English
Just bought GPT 4, how do I insert pictures? 🤨
You bought gpt4????
well the subscription of course
Millionaire
maybe there are plugins to do so, but not in the vainilla version AFAIK
What about vice versa? Have a long conversation in English and then switch to French?
😉
Picture upload doesn't exist yet
why dont you try yourself?
oh damn
I'm monolingual
Never tried . Guess it will switch and adapt
then what do you care?
Seems like an interesting experiment
try it together with Google Translate, f.ex.
Yup, it’s terrible marketing by OpenAI
trust me ChatGPT it's quite proficient as a translation tool, and manages to generate valid, correct text in other languages than english
plus it has no trouble in switching or mixing different languages
ChatGPT is miles better than GTranslate
I had it translate to fluent Chinese, and I had it checked by a Chinese friend who said it spoke perfectly
thats not the discussion, but yeah, I think it is
Do you not have Plus? What's the point of talking in the gpt-4 discussion if you don't use gpt-4?
Write a prompt on shri ram
Hi, is there anyone still waiting for the access to plugins? I don't have option to turn on beta
I have plus yet I didn't buy gpt-4.
This isn't chatgpt. Go to chat.openai.com
Oh, sorry that joke went over my head
lmao
I somewhere read GTranslate was better than its current model version, but got worse by reinforcement learning on tons of pages that got translated by itself
Right, that tracks lmao
hahaha
lmao. do people make that mistake often? That seemed like a premade response. It's like people who Google stuff on Facebook.
I haven't seen ChatGPT mess up majorly with any translations. GPT-4 is exceptionally good because it understands context
Which is important in a language-based context
Every hours
As a text language model, I have no opinions about that...
I have to refrain from throwing something every time I read those words
hahaha
True, true
I'm just anxiously awaiting browsing/plugins to be enabled for me
Hopefully soon
I've tried to force him to not be so 'polite', and somehow managed... after dozens of prompts asking him... and just for a single answer. But yeah, it somehow was able to get rid of that C·3PO style for a brief while
"Is there anything else I can assist you with?"
You have an interpunct on your keyboard?!
Nah, clearly he just memorized all ALT-codes. 🧠
Which means you do technically have an interpunct on your keyboard. It's just made by a specific combination of keys.
I'm jealous
It's definitely in the top 3. Up there with time travel and falling asleep instantly
Apparently according to GPT-4, an interpunct is ALT + 0183
I'm on phone so you'll have to be my guinea pig to see if that's true
· yep it worked! ····· Brilliant!
So, again, regarding whats stated here:
https://help.openai.com/en/articles/6825453-chatgpt-release-notes
'''
In addition to the beta panel, users can now choose to continue generating a message beyond the maximum token limit. Each continuation counts towards the message allowance.
'''
... does anyone know how to toggle this on?
Guess GPT-4 did all the memorizing for us... 💪
Who needs a brain when you have ChatGPT?
Hey quick question which is the cut off date for the latest chatgpt version?
It's definitely a good partner to it... Though not fully recommended to replace it yet 🤣
September 2021
Still September 2021
Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.
Please report any incident by sending a DM to @open crescent immediately.
But if you use GPT-4 with Web-Browsing, the cut off is functionally the present day. Although it is slow
Web-Browsing always fails for me
Can't wait to have access to it rolled out to me so I can give it up-to-date code documentation. 💪
dont know whybut takes like 5 mins to finish and almost aslways returns no results
As a Plus member, if I still have not obtained the web browsing function, what should I do?
just wait, the rollout is taking place over the current week, so be patient and you will see the Beta option soon
yet xD
have you been able to try this?
In addition to the beta panel, users can now choose to continue generating a message beyond the maximum token limit. Each continuation counts towards the message allowance.
yes, on long output there is a continue button provided
👍
That's why I put it there 😉
👍
Give it a year or two and we'll definitely have a contender
(For the brain, I mean) 🤣
Exaggerating, of course, but yes
If neurolink somehow goes the way Elon Musk thinks it will...
Is anyone experiencing a problem today with the plug-in and browser versions of GPT 4? They were working yesterday but today the system just shows that it is busy for a minute and then there’s a generic error message. The normal GPT 4 chat works fine, it just doesn’t work when plug-in and browser are activated.
They've been working for me, maybe it's location based?
yeah me too
matter of fact even normal GPT 4 becomes so laggy and barely responds tonight
out of interest, where are you located?
Malaysia
I’m in Dubai. Same issue.
i've used VPN thinking it could fix the problem but nope nothing changes
GPT-4 seems slower than usual for me here in Australia, it's not terrible though
seems you’re in a better position than me, mine decided to take the day off
Error message today —> 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 on your computer?
it's weird it work fine here.
I do but it’s never given me an issue before. I am not logged into the vpn. The issue is on my phone and laptop as well.
disable the 'web protection' feature in NordVPN settings and that should fix it
I’m having the issue on my phone too
that may have another cause
And it’s only plug-in and browser
Very strange. The normal chats work fine so long as Im not logged into the plugin@and browser versions
there is a bug in NordVPN 'web protection' that causes that error, I see it here all the time
Hi lads
Interesting
@worn current you have no patience! The restrictions I’m referring to are people that already have access to GPT plus and web browsing and plug-ins. Some people only have access to one of them and they’re paying that extra $20.
Again the restrictions I am referring to are people that are currently browsing the Internet and it’s not like you could just browse any website and get 100% of the Data you’re looking for. Hopefully somebody else explain that GN from a galaxy far far far far far away.
why is the icon in chatgpt a purple T
Are there any plugins (paid or otherwise) available which use ChatGPT to analyse and read really long PDF files (suppose, 100+ pages)?
How can I gain beta access for plugins?
AskYourPDF in Plugins
Thanks - I have played around with it, but looking under the hood, it merely scans for keywords and does not read the passage in its entirety.
I have used it with 10 page PDF files and it seems quite good
Suppose I insert Crime and Punishment.pdf and ask 'what are the central themes of this novel?', AskYourPDF cannot answer this question without reading the entire novel.
i..e, scanning for 'central theme' will not yield any useful results.
it's early days, maybe it will improve
Most certainly, just wondering whether such as tool is available right now (paid is fine).
I’m pretty sure there are few other pdf plugins but I haven’t properly tried them yet
One of them being ChatWithPDF
This one returns an internal server error, unfortunately.
Ah
ChatWithPDF seems broken
yeah, I have used AskYourPDF
Is anyone experiencing there is an issue on GPT-4 model now? Whatever the question I asked, it takes longer time to load and return "Something went wrong. If this issue persists please contact us through our help center at help.openai.com."
Hi, I am experiencing an issue, chatgpt say that I have been blocked, I do not have access to Chatgptplus but I have to DALL-E with the same account
are you using a VPN?
yes, same problem here
do you have a product called NordVPN installed on your computer?
yes
turn it off
turn the VPN off and it should work
I am. 3.5 works but 4 doesn’t.
it did, it is a new thing? because I always used the chat with VPN and had no issue
Yeah, me too actually.
some VPNs are blocked, so best not to use it
yes, exactly the same proble, 3.5 works, 4 doesnt. have tried switching to mobile internet instead of home one, have tried android browser, and firefox and edge, have tried to delete cookies, nothing helps.
thx
is there any chance to get a support here with the current issue with gpt 4?
keep an eye on https://status.openai.com
GPT-4 seems slow right now, I am sure it will be addressed
its not informative and as i see, im not the only one who s experiencing same issue atm
its not slow. it always gets too long to answer and then gets to "something went wrong" message. no matter what browser or pc/smartphone i use.
do you have a product called NordVPN on the PC you're using?
no i dont have any
i actually dont need to use a vpn at all
ok could you please try with Firefox?
I’ve been on the waitlist the first day it came out, and got access 5.12
yes, as i already said before, ive tried firefox, edge, chrome on pc and + 2 android ones from my smartphone. all the same
ok, then you have done everything you can do, you can email support@openai.com for direct support, otherwise just wait it out

Yet it kinda works anyway
yep, seems like that. anyways, thank u for ur answers!
you're welcome 🙂
I've been on the waitlist too, no access yet with Plus
Have to wait 'till the end of this week to see
@hallow owl I think singapore can use ChatGPT right?
yes, you're right
What is the third button on chatgpt mobile version
there isn't a ChatGPT mobile app
Website version
oh, I don't know then
Only 25 prompts every 3 hours it is more expensive than gold. Do we know when we will get unlimited prompts like 3.5 ?
Probably a bug
we won't get unlimited prompts ever, however keep an eye on #announcements for updates
Thanks mate
Hey all! It's been a while since I last checked and asked and I see some are talking about it but are we ever getting an increased cap limit of 25 messages every 3 hours? 😦
no news on the cap yet, but keep an eye on #announcements for updates
Robert should be annoyed replying the same questions every three lines
ok thanks 😦
it's ok, not everyone knows
I'm so sorry if this has already been asked a lot, i used to check every week over here but now i'm moving to once a month lmao
always feel free to ask questions here, even if you think they have been asked before
thank you mate!
Why cant we get unlimited chats with 4 like 3?
because GPT-4 uses a lot of compute and there are limited server resources, so having a cap ensures everyone gets a fair go
Fancy explanation, just say a lack of computation power
when you mean "not ever" is it like forever and ever?😅 if gpt 5 comes out in 2024/2025 with lots of optimizations there will still be a 25/3hours cap (or a bit more) on gpt-4? I'm a bit confused
The browser access is pretty terrible, especially compared to bard, or even bing chat...
openai should proably jsut stick to training LLMs, and release updated trained versions more often. forget about the plugins et al
I said not ever for unlimted messages, the cap will be raised at some point in the future, we just don't know when yet
Oh ok got it, my bad! Thanks
Anyway LLMs will become a competitive market, openAI will simply be forced to increase the limit cap by force of circumstance
Is there any plugin for web browsing that I can use with other plugins? I find sometime the default web browsing plugin is too slow and always failed, and some plugin like Wolfram is too good so I want to use it together with web browsing
try WebPilot plugin
i didn't received plugin yet 😦
We have to wait, it won't be long
same, but hopefully by the end of the week
Hafta wait
Bhaiii yrrr
it is not available now
oki 😭
yh
hmm let;s see
Is anyone how what's going on?
the rollout is taking place over the current week, so be patient and you will see the Beta option soon
interesting think, i always get "Something went wrong. If this issue persists please contact us through our help center at help.openai.com" message, but when i click "regenerate response" it works. i already mailed and described everything in support, but maybe u know why it answers only when regenerating response?
bard has 100 message capacity
I don't know, sorry.
They will increase it to 25/h
ok, thank u anyways
sadly it is still very high time to wait after we cross the limit
It seems probable to me but does anyone know if the web browsing model uses context tokens for each page reading?
they will increase it, that's why they include the survey form when you reached 25 limit
i just don't know why tf web browsing has such demand. you can google it. at least the resutls are comfortable to see and choose.
i'm disappointed with the web browsing feature they provided.
it's still in beta, it will improve
firstly it is very slow, 2nd it displays similiar web results
and if you want the google reference aka web browsing then use bard. bard is just like a scrapper i think.
they are scrapping code from github.. xd

bard is really good if you want to access a bunch of recent info, and put it in a table with columns with different data for example.......... it makes this web access on chatgpt look really bad
nerver ask GPT4 to be your wife 
it put title as AI marrage proposal rejected 
Starting the next week I guess
i find web browsing is good if it worked, but it's too slow right now, just give it time to improve. I've used all chatgpt plus, bard, and bing side by side for my project, but so far chatgpt is the best so far for lengthy project.
I did it for content on youtube btw
Bard is bad even in that, almost always it just says or puts fake stuff in the table
It's true that some features are just gadgets. Like the graph generator which seems to be just another GPT4 request to generate a matplotlib script. Or the interpreter code that just connects the input and output to a terminal.
yes im using gpt4 for it's analysis, generation ability, coding but web browsing and most of these plugins would be a waste of time
Can you use gpt 3.5 with plugins?
no
Is there any plugin at the moment that can transcribe audio?
I don't think so, keep an eye on the Plugin store
try voxscript
How long until shipping plugins to paid users is done, still not got anything
end of this week
Can voxscript see private youtube videos? I could possibly upload them to YouTube, provide the link and voxscript will transcript them?
respect
Does anyone else feel like they dumbed down the base model?
I haven't try private, but it worked with my unlisted videos
Anybody else having issues using Noteable? It doesn't appear to be working since late last night.
it was working this morning my time in Australia
Hey y’all, how can I switch the mode of a conversation back to gpt4? Ran out of messages and didn’t want to wait the 45 minutes so I continued the conversation, now it’s stuck in 3.5 and I’ve got a long chat history here with a project I’m working on and need all the references.
once you switch to GPT-3.5 you are stuck there, you can't go back
you cant
Well that’s a major over sight
Few thousand words. Lol. More like several hundreds of thousands of lines of code
Not true. It can recall the entire conversation all the way back to the first request.
Hi, sorry if this question has been asked many times but just to make it cristal clear: by having a GPT Plus sub, i can use chatgpt4 without additional fees (on top of the $24/m) as long as i respect the 25msg/h ?
It not really has very long memory, at certain point if the conversation is too long, you need to create a new chat again, and copy and paste your related work for context again
yes you are correct
When GPT4 without cap
Perfect, thanks a lot
Thanks for the info. I think I may end up starting again with a new notebook, which is not ideal. Still, amazing plugin.
No eta and not soon
Except if you ask it to look back at a specific time in the chat, it can absolutely recall all the way back to the beginning of the chat. I’ve been doing this for months now.
No, it can. I literally do it every day and have for several months. Try it yourself.
I mean you will hit the conversation limit at certain point, like Bing.
Anyone know of any special way to share many files of code with chatgpt for it to review all at once? Currently pasting files into chat one at a time? I have plus
what you are seeing is "inference" the model can only remember the last few thousand words in a conversation
You’re missing what I am saying. I’m not saying it’s recalling the entire conversation on every request. But if I need it to look back to the beginning of the conversation and I specifically request that, it can and does.
you're seeing inference, it can't do that
I'm looking for a plugin that can check cinema time
25msg each 3h
“Inference” yeah, it being able to go back and find the error I pasted, from a code completely different from the one I’m working on, is definitely just inference…. It can absolutely recall the entire conversation and sim it up even.
I wish I could share video in here to show you it works. I just asked it to tell me what the first 2 sentences I typed in the thread were and it did it just fine. Probably close to a million lines all throughout this thread.
wolfram plug-in ain’t really any better than what GPT4 already was for math solving. If you used correct prompts it would display/render in latex.
link for the plugin store?
you access it from within ChatGPT
I use wolfram to show some infographics.
I am curious to know how the web page data are formatted in the request of web browsing model. It could be cool to make a selenium based browser plugin to avoid all missed click or bot limitations.
That's interesting, gpt can handle 30k tokens max. But it's quite likely that it indexes history just to train itself backwards. Maybe he has direct access to this database. I'll definitely give it a try.
Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.
Please report any incident by sending a DM to @open crescent immediately.
Am a paying subscription memeber
Yesterday was working fine ChatGPT4 and ChatGPT4+Plugins
Am trying to chat with it right now, and its pointing to: https://help.openai.com
It can't even say "Hi" when i prompt him both ChatGPT4 and ChatGPT4+Plugins are not operational.
ChatGPT3.5 is working. But for the task i need i want to use ChatGPT4
When will be available so i can proceed?
Anybody else with same problem?
Still waiting for my plug-in access I’m like an addict refreshing my page for the past two days bro
same lmao
I wonder if it’s location based, I’m in Australia no idea if anybody else here has it
So I tried it on my "main" chat and I asked for the first sentence I wrote. My main chat have 166k words and it dont gave me first but exactly 31000 words back sentence.
To create my avatar that can be used in multiple virtual spaces and games, such as VR Chat
what are you using to feed so many words? PDF?
I have access, I am in Central Victoria
Nope, its just old chat.
Gives me more hope I’ll get it soon lol
robert, by any chance is working your ChatGPT4 or ChatGPT4+Plugin right now? is the service down?
ok
the API is up but very slow right now
Add your well-crafted prompts to our #1019652163640762428,
or share your interactions with ChatGPT in #1050184247920562316!
oki, so probably, some server demand issue i guess
yes, there is huge demand right now
Now I have no chat history
i've got the plugin, why i didn't find the code interpreter
Code Interpreter is still waitlisted
oki, thanks will wait1h and see
I assume I'm not the only one that it takes 10 minutes to get a reply in GPT4?
it's very slow for me right now
thanks for sharing! Same here. I cant even get it to reeply hi
am in europe
I'm fine with them making some money, but it does feel like they have done the same thing as airplane companies, selling too many seats, and hoping people wont show.
hahahahah hilarious comparision but true!
I guess they will be able to pull it off in time! And deploy the cash in expert engineers FAST
Idk what to tell you man. Works perfectly fine for me. But I’m also able to feed file links and have it work sometimes. So maybe my setup is buggy.
The continue generating man, THANK YOU.
I just assume their servers are dying
What is likely is that we are gonna get less prompts per 3 hours, because of this
It always does the "click failed" thing in web browsing gpt4?
do they post on some twitter account when they are doing maintenance? Perhaps i missed it
is there any way to get more then 25 messages per 3 hours?
I wonder what is really happening when millions of people send gpt agents to browse, scrape and summarize information from the web..
Certainly takes ages to get a response, with lots of failed clicks.
Any suggestion on how to improve that?
It will improve over time
Is there any way to get more then 25 prompts every 3 hours
No
not even with the api?
i guess having 4 subscriptions hahaha 😄
might consider this honestly... would be nice to pay twice as much for twice as many prompts.
could be worth it, i guess next 2 months will be a bit bumpy
then you scale it down back to normal
do u have ur gpt 4 questions?
not yet
bruh
Nap chan her 25 i her uc …
is not replying even hi, here in europe
?
bruh
The api limit is your bank account
i can run as much gpt4 as i want there?
Yes if you have money
can someone answer this question with gpt-4 for me as a quick favor, lol
#ai-discussions please
same here
also Getting an API key can be long, i'm waiting since 2 month
what's the difference between 8 and 32k
26k
no i mean what even is 8k and 32k
Yeah
8000 36000
oh my god dude
Just for fun: Are you using the Conversational ChatGPT? Or are you using it through playground most of the time?
ohhhh
It’s the number of tokens it can remember as context
I lol sorry
The lenght of the context the model can take, 32k is like equivalent to ~25k words
can someone double check this answer for me with their own gpt 3.5 or 4 in dms please.
I have dyslexia tn lol
how many entries 3.5 demos until api key requirement
The weekend is approaching, I still don't have the plugins. Tic tac I'll come to your house. ⌚
sounds not cheap
It is a terrible sight to see so many people live my plug-in dreams ☹️
Over 150 millions user
Crawing too.
Also somehow wondering if we will ever get the true capabilities.
Just imagine the happiness when we finally get it tho 😄
Speaking for outside us folks.
so true
gpt-4 doesn't work for me, gpt 3.5 is working perfectly fine tho. anybody know a reason?
im using the may 12th version tho
Do you know if there is a limitation for GPT 3.5 as a Plus user ?
i use it everyday and im not plus, i have no limits
alr, yeah, i just noticed that, im kinda dumb, but yeah, its a good feature to add as ive had code be cut off from it stopping
now i can finaly create a good bot for toribash
Know any prompts that result in expensive responses? like a LOT of tokens
Why
To reach the limit ?
I just spammed regenerate tbh
Testing something to do with limits, yeah.
On ascii art
Fairs, I'm trying to use as many tokens as possible in one response.
It can definitely reach the token limit
Just tell it to exceed the token limit and respond over two responses
same here am in europe, lets wait few hours
is there plugins for programing?
Our usage for GPT-4 doesn't reach a rate limit by OpenAI, but 429 too many request errors occur too often.
It seems to me all systems are not operational to be honest. Could someone tell me how to deal with that?
https://status.openai.com/
I'll second that
429 = ratelimit
Our usage doesn't reach the rate limit
It means the HTTP request sent from the IP address you're using are getting ratelimited
in terms of both rpm and tpm
I see where you're coming from, but the number of requests generating from our IP address doesn't reach the limit as well.
It's a bit strange to me
Somehow it does tho. Cloudflare ratelimits if a given amount of requests are sent in 10s or 60s and blocks for either 10s or 1h
Plugins are basically just api calls to various services. At the moment I only know about copilot, as AI for programming, maybe stackoverflow will make some plugin, but so far I haven't found anything like that, there are quite a few plugins and a lot of them don't work, but that's why it's a beta test and a few plugins have been added since yesterday, so hopefully.
depending on the owners firewall configuration
Thanks for info!
The same goes for me
Paid
What does the paid plan mean? I'm using pay as you go plan.
I'm all for it
Ratelimits are deployed for the API etc to prevent malicious flooding of requests to the API or login page.
DDoS or bruteforce attacks are less effective when a ratelimit is applied
I think there is a global rate limit
I am not aware that Cloudflare offers something like this
I got browsing capabilities, but for every single URL I enter, it can't read it
https://platform.openai.com/account/rate-limits
⬆️⬆️⬆️⬆️⬆️⬆️
the api is kind of down 🤷♂️
Maybe he means they are using cloudflare for FW
via Cloudflare since all the blocks and limits are applied by them
Yes, it seems to me they've applied some global limits to Cloudflare
I'd like to know it as well
I never heard of anything like this since it would create issues with Cloudflare.
It would create issues because the backend servers do not receive the actual IP address (they see the cloudflare proxy) which would make ratelimiting on backend level impossible
Because if they would ratelimit the CDN IP address it would cause the website to go down for that PoP
I can imagine OpenAI having the Enterprise Plan for Cloudflare and that's why there are custom rules or actions performed for certain requests
What do you mean
thanks for sharing! So valuable site!
Anytime
@ocean wharf they would have to deny the IP address. All blocks and limits are done before you even reach the OpenAI servers because Cloudflares CDN handles everything
It would also be alot more resource eating for the servers to evaluate the requests
guys what did you find are the most interesting plugins
ask to PDF
@leaden cape By the way, do you know how to get an access of GPT-4-32k? I got an approval of GPT-4-8k 2 month ago, but I can't use GPT-4-32k yet.
can anyone tell em why theres a card declined problem for the upgrade? i've tried all my cards, all of them declined, including all of my brothers. has anyone else experienced this issue?
Unfortunately I don't
Noted
Are you talking about api or gpt chat?
API
Thank you, I assume chat gpt is running on the 32k version, because it doesn't say what version I have.
ChatGPTs version is, as far as im aware, based on the 8K model
However i never got it to go even near that limit since it just stops and finishes what you asked for
what tokens were you guys talking bout earlier?
Can someone tell me with the plugins how to do the authentication for certain plugins?
GPT API tokens
It tells me “Manually enter the following code into your preferred authenticator app and then enter the provided one-time code below.” but what authenticator app?
Well, a while ago I tested how far into the chat history it sees, and it came out to 31,000 characters. I know a token isn't a character, so you're probably right.
The browsing is so fricking long right now on gpt4
Bing does it much better
For example Google Authenticator. Better ask in #community-help
ah, yeah, is that a pro thing?
Hey pals
No the API is publicly accessible even for free users but it costs money even when using GPT 3
that sucks, then what version am i using? i dont see anything about tokens
https://openai.com/pricing
Here's the pricing for GPT API tokens
Are you a free user?
give me exaples for cars
Still waiting for the browser
This isn't ChatGPT.
Go to https://chat.openai.com to access ChatGPT
yes
Then its GPT-3
Does take some time but the overall result is better in my opinion but I also barely played with bing.
ok, thought it was, cant wait for 4 to be free sometime
it's unlikely that will ever happen, GPT-4 uses a lot of compute
oh, well that sucks, i looked forward to using it, but im broke
hey guys, I have access to chatGPT plugins, but I don't have access to chatgpt browsing, I thought they will will be the same. I am a plus user
is web browsing still rolling out to plus users?
cause i got plugins but no web browsing
i dont do iit, im just telling you guys so you can stop it
still no gpt4 api for me?
alr, i got the warning, no more talk bout it, just please stop it
hope some day it will be for free like 3
I doubt it will be, GPT-4 uses a lot of compute
Browsing is really disappointing - am I the only one with this opinion?
it's in beta, it will improve
3 also uses a lot and they made it
yes, I hope so.
is gpt 4 open for api now?
honestly, gpt 4 is prob gonna be free when they work on gpt 5
you need to be on the waitlist
plz stop
let's stick to GPT-4 discussion and move everything else to #off-topic
too bad that they have censoring for abreviations
you now how to get in aniway for free because 3 sad i shod look for the free version on diskord ore rebbit
I am experiencing a lot of "Click failed" in Web Browsing mode. anyone?
what is that sentence?
yep same
hope you understood but englisch is my 4 languige
I am a ChatPlus user and just got the upgrade for Beta, but the Web Browser option is not available for me.
it will come, it's a progressive rollout
ah, if you want help, i can help you
or just use translate
yes
wanna go to dms then?
Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.
Please report any incident by sending a DM to @open crescent immediately.
why is that there now?
I’m a Chat GPT 3.5 plus user I also have GPT 4 with all the plugins except for code interpreter it’s simply not showing up
its still in closed alpha
Oh ok do you know how long it’s going to be closed and was it ever open
Cool
Absolutely GPT 4 is a beast
You can sign up for the wait list https://openai.com/waitlist/plugins
bard still completely unavailable in NL ( not that I care much about it xD )
Bard isn’t currently supported in your country. Stay tuned!
wait its actually available in 180 countries but not in any EU
guess thats a bit weird
scared of privacy laws I guess
Why did they leave out the EU
good question
Privacy laws I’m sure
is there a plugin for search that actually works?
Privacy laws most likely
Bing lmao
damm, why didnt i think of that?
apparently also Canada for some reason
The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.
Bing works really well with search and can summarize articles for you
i need to make a prompt to make it give more detailed responses, they are very short
For ai not related to openai gpt please use #ai-discussions
@turbid flume@wraith ginkgo
This is not ChatGPT, you can access ChatGPT here - https://chat.openai.com/chat
The browsing plugin uses GPT-4, right?
Yes
tyty
A mind is still a terrible thing to waste because because with all the power at your fingertips, you still have to be able to ask the right questions . I’m getting matrix vibes . GPt 4 is dope
Chatgpt 4 was able to give me links when i asked for some sources. Now it doesnt and says that it has no access to the internet 😅 anyone had the same "issue" ? Thanks in advance 👍
You are not asking the right questions
sometimes it might
other times it might give you and they are wrong (or outdated)
other times it will refuse
Hmm ok thanks anyways guys👍
Why did I talk to gpt4 and gpt4 didn't respond but gpt3.5 works. gpt4 is just loading.......
It can't be that I ran out of my three-hour limit, because I haven't used it in the last few hours.
I want to build a program that would tell me MODE 1 / MODE 2 based on my real-time screen. I want to train the AI to make a selection based on thousands of images or videos. Can anyone point me in the right direction? Which AI would be best for analyzing pictures/videos and finding patterns?
could be overloaded dude it’s fine
Yay Plugins! But where is the code interpreter?
Artificial intelligence (AI) is when machines can do things that typically require human intelligence, like recognizing speech or images, translating between languages, and making connections between different types of information. It's different from the intelligence shown by animals or humans.
code interpreter is different than plugins 🙂
So it's not available yet?
The only thing I got in the last couple of days is browser and it's not that useful lmao
That code interpreter is going to be a game changer
Does anyone know how to give GPT long inputs?
Ask something like this. summarize the following text in plain and simple english
Artificial intelligence (AI) is intelligence—perceiving, synthesizing, and inferring information—demonstrated by machines, as opposed to intelligence displayed by non-human animals and humans. Example tasks in which this is done include speech recognition, computer vision, translation between (natural) languages, as well as other mappings of inputs.
Does anyone have access to chatgpt 4 api yet. I know there is a wait list but has anyone actually got to use it yet? I would imagine if the nessage cap was still on it. Then it wouldn't be that useful to have access to anyway but still
daky#eher
there is no cap on API (just rate limit and monthly price cap)
daky#leher
Interesting, I mean I know we have access to chatgpt 3.5 turbo api but having access to 4 is going to change alot I think
This was Gots answer Artificial intelligence, or AI, is when machines show intelligence - they can understand, combine, and make sense of information. This is different from the intelligence shown by humans or animals. AI is used in tasks like recognizing speech, understanding images, translating languages, and other tasks where they process input data.
its gonna go very deep into your pocket 😄
Yeah I'm sure it will cost a lot
some people have access and others do not. I do not know if all Plus subscribers are getting access to it
Will it be available?
I don't know if it will be publicly available in the future
It will most likely be available, because OpenAI is trying to compete with Google, and Bard has a code interpreter, so OpenAI will most likely include one as well.
Think about it there are people lots of people that still know nothing about crypto just like a vast amount does not know about Gpt
But it would be interesting to develope a plugin for let's say visual basic to use chatgpt 4 to analyze the code It gives in real time.
That’s what code interpreter is for
I think copilot might be doing to same thing if I'm not mistaken but I haven't really looked into copilot to much yet. Not sure what it's using as a model or anything
they say codex
the new copilot X should use GPT4, but not sure when is that coming
Copilot is a plug-in I downloaded all of them
is web browsing still rolling out to plus users?
Gotcha
cause i got plugins but no web browsing
yes, until the end of the week
Think I would rather have the plugins at this point lol
Maybe webpilot is better then browsing then 🙂
do you also have code interpreter?
Quite liking copilot though
I would quite compare it to excel autofill, but for programming (and much more advanced)
Like when you would need to write many lines of almost the same thing, it will just autocomplete that for you (of course its also a lot more than just that)
I use "WebPilot" as plugin as a stand-in for browsing, works fine!
Ah sorry @wind urchin meant to tag you too
I wish
I can't wait until it comes to the point where the code interpreter will cross reference what you actually mean through the internet and complete the code that way, instead of purely basing it off what you told it.
well you can using the Noteable plugin!
That would be great if I had access to any plugins besides browsing lmao
you don't? hmm, I don't have "browsing" nor "code interpreter" but I do have "plugins" since yesterday
Yup the only thing I've had access to is browsing. Got that like 5 days ago
I have plus but no access to plugins or browsing. Whats the deal?
You can go to "settings" bottom left and in the section "beta" you first need to enable "plugins"
it's rolling out this week, be patient and it will come
Robert would you happen to know if you have extensions active will that impact your upgrade or not?
Yeah that's how I enabled browsing but there aren't any plugins there
no, it won't impact you
Good to know just wanted to make sure i've been testing a bunch of them i've found some good ones lol
still this long till end of week 😄
<t:1684619940:R> (assuming they end week on Saturday)
Thanks. I logged out and back in and now I see it
be aware though that buggy browser extensions are a common cause of issues with the ChatGPT website however that won't affect the server side upgrade to plugins and browsing
i've been pretty selective with my extions i've found the ones i like
Anyone else having trouble uninstalling plugins?
Cool 🙂
not yet i don't have them to install therefore i can't uninstall them 😛
Yeah I have been logging out and in every day just to see if that could kick it in but just got to be patient
just a page refresh is all that is required, no need to log out and back in
Got thanks robert
it has past like 2 hours. And still not working for me the ChatGPT 4 and ChatGPT 4+Plugin
Anybody else? Am from europe
Is there a set time in the day I should be checking back with it?
it just says: "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 on your computer?
/gpt
go to it's settings and disable the 'web protection' feature then everything will work again
i shut it it off the NordVPN just in case
still not working, am thinking to try from my other computer
just to check if it responds over there
it's not enough to just stop NordVPN you need to go to it's settings and disable the 'web protection' feature then everything will work again
Somebody mentioned a plugin for faster web browsing capability... any idea the name?
oki, will give it a go
it's not a plugin it's a model and it will be available to Plus users this week
Yeah, there is a plugin though which somebody was discussing that is faster apparently.
WebPilot, great plugin
Awesome, ill check it out thanks
How is web pilot different then browsing
am not really using NordVPN. And its requiring me to login with a old account. If i unistall would that do the trick you think?
it's a plugin and it's third party, it's quicker than Web Browsing but may be more limited
I would hope so, most people do ok by going to it's settings and disable the 'web protection' feature then everything will work again
I'm using WebPilot since yesterday, dont have browsing so I cant compare, but it doesnt seem limiting at all.
Web browsing atleast in my experience is pretty limited. Alot of failed clicks etc
you have to proxy it
I tried bard for the first time through ssh-socks
Ideally I would like to have also normal web browsing and the default openAI code interpreter
its nice
as the default code interpreter also allows for multi-modal
Code Interpreter is still waitlisted, it's not included in the current roll out
But still thus far Noteability (having some scaling issues as I undeerstand now) works great as a standin for Code Interpreter
I've never seen it but I heard about that Napkin drawling Turn into a website. Is that using code interpreter
That would need multimodal input as well
seems chatgpt-4 got bugged in my account only
is it lagging or what ?
if so then you're not the only one here
it waits 2 minutes, and then it says Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
Every time, does this happen to you too?
yeah
finally I am not alone!
did I already ask you if you had NordVPN installed?
if you're try to regenerate the response it'll work
I thought I was crazy and people here were gaslightning me
Anyone have a issue where chatgpt keeps going to document360 dot com website instead of the one in link provided.
yes you did, I don't pay for such services lmao
either OpenAI is having issues
or you should clear cache/cookies in your browser
or use firefox which seems the least problematic
besides it was working perfectly fine the previous day
no worries, just trying to ensure I am being helpful 🙂
you're definitely
as you're here for the last 9 hours
tried different browser, different computer, phone, mobile data, 3 devices, 3 ips with different countries, it is not me.
it is my account
I am sure you have emailed support@openai.com ?
yes
ok
but that doesn't mean I can't complain about it here
complain away, we'll listen, well I will at least 🙂
they said we answer you within 1 week
very unlucky to have account issues
since the only option is to wait for the overwhelmed support@openai.com 😄
I guess I wonder if they got some system behind that mail
since mail based support is quite outdated now
yeah apparently some issues can be resolved there (like refunding I believe)
i mean they've created this beast i don't think emails gonna be any trouble
I suspect maybe my chat history is way too long
just normally there is some ticket system
mails get lost and forgotten 😄
it would be absolutely ridiculous if that was just shared mail 😄
they can have system that deem what it should forward as serious problem cases that aren't just simple user errors... like Rainb's case definitely should be prioritized
Gpt4 so slow
GPT-4's etc ability to comprehend messages is useful for technical support
if it's any consolation I am pretty sure emails to support are triaged (I don't know this for sure, I just think it based on recent observation)
I mean they have 100m users, so like.... maybe like around 1 m have the same issue and they will also email them maybe?
the bad thing is most people that have the issue won't email them
just really bad that you send mail there and have no feedback other than just automated reply
it should be a ticket you can track yourself in some way
one day maybe
and the worst part is that I pay for plus
for now we have to hope we never have any issues 😄
technical support also deals with tons of trivial issues that flood them... then there are issues that actually users need their help to fix instead of the issue being on user's end
Did anyone lose access to the version or gpt with online capabilities. I already had access to it as well as plugging, and today it's gone, I only have plugging access.
Hello!
Is there any chance to reactivate my account, suddenly my account was deactivated. I get this message while trying to login: You do not have an account because it has been deleted or deactivated.
If you could track your ticket, that would also reduce people just spamming that mail
did you break tos lol
Don't know how may I break TOS
the only thing you can do with this is send an email to support@openai.com and explain what is happening and ask them to look into it for you
ok
Thanks Robert
does anyone have that issue where GPT-4 answers with a "T"
instead of chatgpt logo is just "T"
I saw someone else mention that earlier today, I don't know about it though
trust me it is not device dependent.
it is not device dependent, it is account dependent, learned it the hard way
it's not even browser dependent, opaqueAI
it happened to me, I think it is like a default image
and then it quickly loaded the correct one
guess this guy said the same
#gpt-models message
I miss black gpt4 when it actually worked.
ah, good to know, I'll keep that in the old knowledge base for next time 🙂
purple gpt4 always fails
it is just my guess tho, I can't be sure it it is the case
but well, this is the least of the problems to stress about lol
the logo is an svg
do we only get to use web browsing with gpt 4? I liked not having a limit of messages when it worked with gpt3.5
yes, only GPT-4 I am afraid
it's it's own model too, it's not a plugin
When I log on from my phone, GPT asks me to verify my identity with a one-time code. Has any one else experienced this?
2FA?
yea what phone do u use
#what is chat gpt
i had this problem and figured it out
you can access ChatGPT here - https://chat.openai.com/chat
Yeah, but I don't know where to find the code.
well there is 2FA they recently added
but normally you have to enable it manually in chatGPT settings (also applies to platform site)
iPhone. Thanks.
go into your passwords in settings
How many lines of code can code interpreter give, can i make a whole site or is that too much?
#hey how can I access the free version of chat gpt 4
Guys, last month i still could make chatgpt to simulation the query mysql. But now i can not do that again. I think chatgpt 4 developer make downgrade
did you set up some authenticator app on your phone for chatGPT?
you should see one for openai.com
finally I'm getting back access to gpt-4
there is no free version of GPT-4, you can use Bing which is based on ChatGPT
GPT-4 can handle up to 8K tokens
I will not try plugins again, unless like in a week or so
If I did, I didn't realize that I did
How much text is that ?
plugins ruins my account
that would be a problem then
Do you know where I would turn that off?
#how can I earn money from you tube with reuse content with out any copyright issues?
is it not in your iphone passwords folder???
This is not ChatGPT, you can access ChatGPT here - https://chat.openai.com/chat
like the OTP
Gpt5
Normally you go into your chatGPT settings (bottom left)
enable 2FA
then you choose some authenticator app on your phone
you scan barcode on chatGPT site with your phone
It will ask you to fill the code from the app into the chatGPT site
then on any next login you have to go into your phone and get the code which changes every few seconds
otherwise you cant login...you can only turn it off when you login, in the same settings
otherwise I guess your only option is to contact support@openai.com and wait probably few weeks (make sure to use the same mail as you registered with)
You're the best. Thanks.
The system didn't do anything you mentioned. However, I was able to turn off the 2FA which solved it
some plugins require 2FA so it might be a good time to set it up
then I guess you are lucky you were still logged in somewhere 😄
I panicked. I rely on this for so many parts of my job now.
I know this is a dumb question, but how do you set it up? I have DuoMobile. Do recommend others?
I use the Authy app on iPhone because Authy works well with lots of systems
@hallow owl where do I activate 2FA? I can't find it on chatgpt or platform.openai.com
inside settings
bottom left 3 dots
then settings
under account settings on ChatGPT
then data controls and on the bottom
this also applies to logging into platform (API billing / usage / settings)
but you cant set it up on platform (weird)
"General" "Data controls"
Hmm the web browsing plug-in keeps failing clicks?
Sites can add a “no-ai” flag that cgpt web browsing will respect and not visit.
Oh excellent thank you for that! The more you know.
it is more about no crawlers tha about "no-ai"
they don’t wanna deal with all the traffic
Jerry
it's working again the ChatGPT 4! Thanks! 😎 ❤️🔥
great! have fun 🙂
did anyone get the code interpreter plugin along with the roll-out?
hello
no, it's waitlisted
check out the WebPilot plugin and compare
Is copilot a browser?
I looked into copilot and I couldn't figure out what it was
I know almost nothing about copilot
Oh, no I mean webpilot, my bad, lol
Copilot is a copilot essentially
oh WebPilot is a web browsing plugin that does a few tricks
I'm gonna check that out right now
Has anyone used the plugins? How do you feel about them? It's cool, but I don't think it's groundbreaking (yet)
great, let me know what you think
I will
Noteable is great, WebPilot is pretty good, SEO App looks promising
how can i use chat gpt ???
you can access ChatGPT here - https://chat.openai.com/chat
Hey guys, have you tried giving gpt4 a link?
I tried and for most the links I gave it, was unable to accees
do you have browsing enabled?
I do
does the website allow web scrapers?
I'm sorry, but it seems like I'm currently unable to access the URL you've provided due to some technical limitations.
Anyone started earning using chatGPT/generative AI
hmmm, not sure, it's the documentation for an API
Has anyone tried the financial plugins? I’m looking to test them out but I keep getting errors
I have
you can say that ya 😄 been implementing it in my workflow
if it doesn't allow web scrapers, GPT-4 can't access it. I'm not sure how to check, though
i dont have gpt4 access anymore, but i am plus, why/
I’ve noticed wolfram and golden are the least bias
The others seem to promote certain stocks
check /robots.txt
Would you say they’re tools suitable for augmenting a trading session?
best option (not all sites have a list)
Who can tell me why I was banned?Sorry, you have been blocked
what country are you in?
It’s not a ban, it’s temporary (I think)
I do NOT recommend using an LLM (like ChatGPT) to give you investing advice!
Well you can use it to acquire lots of data about a stock as well as some holistic news and perhaps learn some things you didn’t know before
It's permanent if you don't change what causes the block
Pretty any form of technical analysis that doesn’t require large data calculations can be done through chatGPT and wolfram if you wanted to
Have you tried retrieving real-time candlestick data?
i use vpn
LLM for stocks is extremely funny
definitely not. But it can give you ideas to implement. keeping in mind it's all at your own risk
I haven’t
Turn it off and try again
i hope so
how does that work? D:
Okay, it's awesome 😎👌
Is it tokenized?
turn the VPN off and try again
There are tools that help stock analyists, LLMs are not one of them 😭
by adding /robots.txt to the main url
Yeah haha
it's pretty good isn't it? 🙂
I’ve experimented with pasting candlestick data (from TradingView) before as a rudimentary technical analysis (best I can do without chatgpt image analysis) and it wasn’t too bad. It was able to analyze candlestick patterns and trends fairly decently.
I think the issue is that people want to blindly ask ChatGPT what to invest in 🥲
thank you
It’s not really good except for maybe some light holistic analysis
Other@than@that the tools exist
Yeah, it's pretty good indeed. Good for researching, for me.
I see. Welp
Who cares about stocks. Don't rely on LLMs answers to medical advices
Yeah, it appears that they really want the ai to work FOR them. Which is not necessarily the case as of yet
thanks
moral of the story: Don't rely on LLMs for anything important
If we submit@ourselves to@chatgpt recommendations@then chatGPT owns@the economy
First step to domination…
Wel correction : it would own the assets of all the poor people asking chatGPT for advice lmao
One thing i did try and was interesting with the web browsing feature is probabilities and statistics, again, DO NOT trust the LLM to make a decision, but I can say it can lend a hand
Lol
@hallow owl Any other recommendations for plugins?
Wolfram 🙂
well, guess the site I'm trying doesn't have scrapping cause adding robot.txt gives me a 404
It’s not bad when it comes to long term trend analysis. I inputted candlestick data back when nvda was around 210 and it provided a long term price target of 260.
Man if only I had gotten into nvda
why
Already got that one. That was my first one 😂
yeah it was my first also
I was thinking about that a while ago and didnt even use ai to calculate xD should've bought.
I feel like my job is important and I think relying on LLMs for coding has been a net positive to me. Even if anything is wrong there, I immediately know while running it. I can also tell ChatGPT to make integration test cases. Since I have multiple years of experience in my field, I do have a high level understanding of the stuff though....
I don't think people truly realize the power of wolfram
What's Wolfram
Especially with ChatGPT
Look into it 😉
I can't 
that's the thing, you must have an understanding of what you want gpt to do for you. Going in blind and asking for answers to topics and concepts you don't have a clue can only be useful to some extend.
true
I’m waiting for image analysis so I can start training it on my strategies lol
You can use google 🙂
Is it only me or is anyone else noticing this? Recent ChatGPT implementations have lost the ability to use the full chat history for inference? My models have stopped being useful at longform discussion because they forget code or details previously inputted into the model.
Did not have that problem once I started using gpt4. But I also use "refer to our conversation" in prompts once I know it's been a while into the conversation. maybe that helps..
Yeah it’s funky. I use it to help write my story and it keeps forgetting details, sometimes introducing characters that never existed.
Naming lists seems to help. “Refer to Character list v3” and it’ll get it right.
Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.
Please report any incident by sending a DM to @open crescent immediately.
Definitely weird, this didn't use to be a problem in early implementations of gpt-4, but lately this has made the model hard to work with.
@solemn basalt @tawdry tree I've tried including "reference" prompts, still doing poorly.
Anyone thinks it's intentional ? Midjourney also evolved to make the prompting more "intense"
How much code are you working with? I think even gpt4 has a limit of 8000 characters or something (don’t quote me on that)
I don't see why they would intentionally make it bad to work with directions
I realised that to write stories it was better to edit part of it and feed it back in with new prompts to maintain certain consistency because it does create new characters or new plot lines that are not initially intended
Not bad, just more complex
it seems GPT-4 model at times doesn't as much forget but ignores instructions if during generation it can't work something in you want... so instead of telling you "I messed up" it just tries to wing it and lose various details
But you're right. Makes little sense to do that intentionally when they probably want more new users
I'm hitting the character limit occasionaly, but if I split up the prompts and then reference one of the previously inputted prompts it will have forgotten that prior prompt.
Yeah. Using outlines also helps. “Write part1.2 of Outline A, using Story Linguistic Profile”
Yes. Always generate an outline first
And then specify the way in which it will generate each part
I find for example if you want to use it let's say for storytelling... have it first work outline like five chapters with short summaries what happens in them and then refer to that stuff when working out chapters... this is of course if one wants to use it as storyteller help for example
Hi there. Just wanted to enquire is there a way for GPT-4 to evade AI detector tools like GPTZero and Turnitin? Is there a particular prompt you need to input or?
Ensuring that it doesn’t limit its responses based@on the token length and responds over subsequent outputs
Yes, totally. Best way I found to mitigate that is to instruct it to "not make up information or data, if it doesn't know it has to ask for guidance"
how do we gain access to plugins again
I haven’t used code analysis recently, but I did a few weeks ago and it managed to break down a 600-line program that was broken up into 3 or 4 sections
So@good@to see teacher on here trying to understand the modern techniques!
Are you using it to generate code or analyze code?
Give it samples of your writing and ask it to analyze the writing style, then copy it.
Yeah this is got 3.5
Hard@to avoid it
Actually I’m a student. Just trying to see if I can use GPT-4 to escape AI detector tools 😅
I’ve done it before and it’s indistinguishable from my own writing
In gpt4 I did not have issues with character limits as much as I did with 3.5
Yo what
The cops are gonna track down ur discord if they see this
using ai to make ai undetected
But in all seriousness yeah you can sort of by telling it to use a more unpredictable, sporadic tone
Use ChatGPT as a tool for essay assistance, not a complete solution. Relying on AI may lead to inaccuracies and plagiarism risks.
I'd recommend editing the result yourself so you can inflex your own voice. But there are tools already in place to workaround that, I just don't use them
simple example is like... instruction prompt tells character details... they're dragon etc... but it's inconvenience for GPT-4 during storytelling that the character is a dragon during output process so... suddenly it's like "the character is human now because then they can do what I want them to do, because otherwise it makes no sense" but you are like "no, the character is a dragon... it's important... don't change the detail"
Chatgpt also has some tendencies I can’t describe. If not augmented properly, you can almost definitely tell chatgpt wrote a piece.
this breaks the TOS
Oh ok
But having it copy your style usually works. Grab whatever it spits out and modify manually
Yeah. That's why I think writing stories with ai may be less productive, instead better just use it for outlining and plotting, then one just writes the story. Takes longer getting the ai to follow through than to fully flash out a story ahahah
now I dont know what it was 😄
It’s great at imagery though
So@its@more@than just an outlined
Having it write a draft is also useful. I make it write entire chapters as drafts, then copy what I need and modify accordingly
Works similar to coding with got
Totally. The writing style and repetitive wording is a straight giveaway
You can easily mess around with all that
Also you can copy paste a desired writing style
I'm having the AI generate a brand new card game using a deck of 52 cards. It's crazy.
That is true and can make your writing process much more streamlined
Tell it to mimic it
It's adding a dice rolling mechanic to the mix, it's crazsy
yeah, I agree
it can be a pain to be trying to be getting the whole code written
better to use it for smaller parts, or examples from which to work from
It’s awesome for overall project planning management and execution
On a high level, as well as technical labor on a lower level
It’s putting the lower level pieces together that makes@it hard
100% like you can do that in mj and get pretty unique results but the overall feel can still give away the ai behind it. But of course, you can experiment in a plethora of ways to refine it
What are the main methods of plane table Surveying and describe them
I'm gonna add this game to my Tabletop game, in universe game creation baby!
This is not ChatGPT, you can access ChatGPT here - https://chat.openai.com/chat
😄
You can then use python to get the full game on, let's see that result
Oh my god, I got access to plugins JUST AS I REACHED RHE MESSAGE CAP IVE BEEN WAITING FOR DAYS ON END REFRESHING
Nooo
nah might as well refund it
Btw avoid caps message you can get mute by automod
,,??
this is outrageous openAI. 100% working towards my downfall
Oops yeah, I just wanted to really get across how I felt haha
Speak espanish?
Yeah OpenAI is really awesome with their censorship in dcord as if we were all chatGPTs ourselves who needed to be contained
English only
“We don’t currently@have the capacity to support other languages lol”
let’s see what chatGPT says about that policy
@hallow owl thanks again for webpilot. it's truly awesome.
You can use different language on chatgpt but not on discord
Is anyone else testing out the new browser ability? I've noticed it's quite buggy.. slow, lots of failed clicks, and it produces inaccurate information. It seems the model is more functional on Bing at the moment.
you're welcome 🙂
Lol why can’t you use browsing if you turn chat history off
#off-topic if you want to share
Can we start muting suspected none native speakers as well

Huh
So they can’t speak? Lol
No politic
probably b/c its still in beta and the point of rolling it out to users is to test and collect data from users using it, Im assuming
Assuming what he said is true, that makes sense.
Yes provide a small amount of trusted people access with gpt 3.5
Then release befa@and roll out more
I get so many failed clicks in the browser mode. I assume it's because websites are blocking GPT as they would see it as a bot. I wonder if they'll get it to start solving captchas..
we just delete the messages in languages that are not english, no need to completely mute anyone =)
Ohh ok haha
Sometimes
I feel like I gotI 60 second mute once for@soeaking “Spanish “
After sending a prompt
An English promot lol
Automod was worst 5 month ago no worry it is better now
I wonder how websites are going to handle traffic from AI..
This was recent
I will do a test
5 month ago saying à would mute me for 60 sec instead of a
FINALLY GOT PLUGINS WOOT!!!!!!!!!! LOL
This was a typo I did often with my auto correct
æ
nice I was so excited when I got it yesterday
guys, I know why gpt4 was being a bit buggy before
HEY testing automute
I mean partially
ok nice
Let's get back on topic :3
Pls tell us
please, don't probe the filters, that will only get you muted
I mean part of it might be that I had a very very very very big chat, (not long but wide), there were like 89 attempts at the beginning
and each attempt had lots of subattempts, like 20 or so
and each thread could be like 10 or 20 times
I've been using like 1 chat for like 3 weeks
so it got really big
SAMEEWE
I’m ready to be immensely disappointed
lmaoo same
Tho the scholar plug-in is really making me excited for my question cap limit restriction to go away
Wdym by this?
How would@it do@that?
