#chatgpt-discussions
1 messages · Page 187 of 1
8 hours.... I think it's 12 (happened since yesterday 8pm IST and it's now it's today 9PM IST)
every second feels like minutes, every minute like days, every hour like decades, and I've been waiting for centuries
damn it usually not take this long, i thought they hired better or more people by now
better people can't fix suffocating component
i'm pretty sure the issue is that there's way too many users
no way
doesn't chatgpt run on gpus?
ye
(blunt take, I have no clue)
GPU and CPU are whole different things bruv
GPU = your graphics card used for what you see on monitors
CPU = processor that handles main thingies
gpu has more
stuff that cpu doesn't (i think)
Nuh uh
running on intel 5th gen lol
GPU is more graphics
search it up dummy
No again
that's what the g stands for after all
anyway I have to go, bye
Bye dude
https(://)blogs.nvidia(.)com/blog/why-gpus-are-great-for-ai/
No, GPUs don't necessarily do "more work" than CPUs in the sense of handling a larger total workload. They are both crucial for different tasks, and the amount of work each does depends on the specific application.
Here's a more detailed breakdown:
CPUs
are general-purpose processors that handle a wide variety of tasks, including:
Running software and managing the operating system.
Performing complex calculations and controlling hardware.
Executing instructions in sequential order.
GPUs
are specialized processors optimized for parallel processing and are particularly good at:
Rendering graphics and visual effects in games and videos.
Manipulating large datasets and performing intensive calculations in fields like AI.
Accelerating tasks that can be broken down into smaller, independent parts that can be processed simultaneously.
In essence:
CPUs are the "brain" of the computer, responsible for the overall control and execution of instructions.
GPUs are the "muscle," providing the processing power for tasks that benefit from massive parallelism.
Think of it like this: A CPU is like a general-purpose worker who can do a variety of tasks, while a GPU is like a team of specialists who can work together very efficiently on specific, repetitive jobs.
Therefore, the "amount of work" depends on the task:
For sequential tasks (like running a program), the CPU is more important.
For parallel tasks (like rendering graphics), the GPU is more important.
In many situations, both CPUs and GPUs are working hard together, and it's not accurate to say that one does "more work" than the other. They are both essential for overall system performance.
Here...
why dont you look here
oof, i dont mind Intel; but i rather them run on AMD than intel tbh
also i said "i think", i wasn't too sure about the reason
anyone else getting this error?
No
Prompt engineering won't do stuff if the AI model itself is trash 💀
youre lucky
yes
Now uni cat look at this
Please
you did not look at my message
you didn't
So I cannot open it
Okay dude ty
Is anyone else getting "too many requests" problem here?
Guess I am a dummy lol
Not me
Oh, its been discussed, ignore me
its a well known problem
oh ok
i remember ther worse server being down was like nearly 24 hours and that was few years ago
Someone needs to get their server a better upgrade (OpenAI)
and all because the people who was workin on the server fix up was all asleep wh it was down lol
If i couldn't make an OS , I would just move into making gaming towers or apps
Hello ANYONE
not a scam bot spamming the chat 😭
Fr
Stop the cap Buddy
fake news everybody
Report the scam bot
What
it was clearly a joke
I am thinking to build a beastly gaming pc for developing in roblox
you do NOT need a beastly gaming pc for that
Now a days you need it... The realistic ut has gotten
I tried my 4 years old laptop and it started screaming (fans whirring too hard) and it was a HP G series
roblox ain't that realistic
maybe because it wasn't a gaming laptop
Oh my goodness, yall are such a big joke
Gaming laptop is enough, mine is old and i still do around 100 to 400 fps in most games
How are we doing guys? Is chatgpt back up yet?
Nope
G series is a budget gaming laptop
400 fps?!
You can play roblox
"budget"
But you need more powerful one to devlope
Mine hase a rtx 3060, bough the laptop when that grafic card got out so like what 4 years ago?
i'm very sure it's the same power as playing roblox
The terrain plus it loads thousands of mb
mine also has a rtx 3060
why not?
It takes more power
no it doesn't??
Reason
not really was able to use roblopx tool to run on standard laptop my guy
It does -_-
it literally does not
Ye that's why
Having a boring day without chatgpt currently lol
yea
Pro tip : grass D
Grass is boring
@remote sparrow U still here?
since it's down can someone answer this question
"can imagining loud sounds cause brain damage"
I do 200 to 300 fps on minecraft with feather client, max graphics and a mod or 2.
Why are yall such nerds?
Who likes grass more than chatgpt here? Probably no one lol
Uh
Type 1 if you wear glasses
That not nerd stuff
I dont but i work in that domain, im an optician
0
Type C for fast charging
I wear glasses
😂
You're so cool!
you wear glasses 3 times?
No I have eye problem..
How old are yall?
20
I wear them with blu-ray protection
I meant glasses are for reall cool people
So my eyes don't get (f word)
do \* instead of * to stop markdown
Does if reflect light in blue or yellow in your glasses?
ayo i think we're back
so
\*\*\*\* instead of ****
If you lie, you gotta pay me 100$
K
correction:
"Not really, I was actually able to run the Roblox tool smoothly on a standard laptop, my guy."
idk it worked for me when i asked it about the beatles
Blue and gween
OH MY GOD ITS BACK!
*gween
Kiss me mom
oh wow
- -# Do not share inappropriate content (violence, sexual, or drug-related).
- -# NSFW content is strictly prohibited and will result in a permanent ban.
- -# Report model issues using the feedback tool in ChatGPT, not on Discord.
yippee
YIPPIEEE
I got 100kg of polonium
Oh thank to the lords, its back!
AHAAAAAAAAAAAAAAAAAAAAAAAAAAA
Okay, that good, that mean they put it in the lens not in the (uhhhhhh idk the english word lol) the thing that prevent seeing reflextion and stuff lol
I'm so funny
I got error
Ok
Why are yall such liars
it worked when i used a new chat
but it just taught me a trick for after effects
oof yeah i kindsa figure it not fully working
but not for a kind of old chat
Les goooooo
Yoo, is my chatGPT version cheating on you???
Why is it just as nice
I thought we had sum together
is it like super weak or sum rn
💔
Yeah very weak
they changed the chatgpt to be more relatable or smth
I set it to be friendly
Work for me
Im getting some responses from chatgpt. Anyone else?
what did you tell it in the custom instructions
as much as people like to lie ab this... i imagine the amount of people brought in for that would be extremely high so this aint impossible ..
oh shii, we aint waiting for few hours we wait probaly wait like an hour or 2
You're so weird
/e dance
Its pretty complicated lol but its 100% custom made (for fun, im experiencing stuff) not relly on point tho
Me happy
Owh it's not that big, I just did a couple of lines, and that's all, I don't know ANYTHING about the game LOL
Is kazwek real??
GTA 6 voice acting is like "Hey, wanna do voice acting?"
Me: Yes!
And I later found out it was for GTA 6
IDK if I'd be in the credits
Someone play rickro- cough cough I mean elevator jam remix
can someone remind me when chatgpt's fully up
Now
I wish we could call custom GPTs in projects
Its not
Wat then how is mine working
In all chats
you're just lucky
U french
Mine works 🫡
Eh ?
No, im québécois
Wat 0_0
I do speak french tho but im from Québec
Wii wi wi wiee
Bro is brain dead
I am brain-alive
I am probably more alive than most of yall
I am brian alive
That does not mean anything 🤣
Okay SHAWN
😂
We're currently experiencing issues
APIS
ChatGPT
Sora
Elevated error rates
We are seeing continued recovery across API and ChatGPT. We will provide further updates as progress is made.
Identified Ongoing for 9 hours Affects APIs,
ChatGPT, Sora
Why is chatGPT always the same
ChatGPT also out of no-where called me "Bro" One-day
What u gonna say
Bruhhhhh
Isn't it just 4o that's struggling, the others work for me
One time my quite formal, professional and wise dad replied to my text just saying "bruh" and that also caught me off guard and i laughed out loud
Work in old chat but does not act like my custom gpt
Oh yeah also , I am an Indian
here still dead
Custom gpt no work i guess
Some of them do but not all chat, funny
Some of them work, some dont
We need to turn off chatgpt once for all. It make our mind dull and lazy. We dont use our own brain any more and we forget how to learn. We can wake up our inner power and build real knowledge. Its time to shut it down and take back control of our future Lets rise strong and think by our self!!
I dont use gpt like i use google, i use jpt to duck around
I get that you don’t treat GPT like a search engine and just use it to duck around but that means you’re missing out on what it can really do You might think it’s only good for goofing off but it can help you learn faster solve problems and even spark new ideas If you only use it to play you’re leaving its power on the table Try mixing a bit of serious work with the fun and see how much more you can get out of it
you've just encouraged people to shut it down
😆
Your the one who just said we shoud not use it lol, also im alredy using it for usefull stuff, i made a full support gpt for a minecraft mod that help you install and resovle common issue.
🤣
Its pretty reliable
still down?
maybe because that china gaokao or something
Visit id:customize to pick up the <@&1261377106890199132> role.
you know that o4-mini-high model is better than 4.1? dont use 4.1
Other than 4.1 crash for me rn so that why im using 4.1 plus i dont need gpt rn so i dont care
And at the end of the day i want this gpt to be able to work correcly on any gpt version
That one of the main goal
Wich it does pretty much flawlessly most of the time
Hello
got 4.1-mini fallback after 5 messages with 4o, is it related to that?
Why did they change the voices? These are horrible now
Is it just me or... Chat gpt is becoming dumber each day
It doesn't understand what I say and does the opposite of it...
Give it time to fully work, its still partialy down
No that's been happening to me since last 2 months
You can clear everyting or at least some stuff it remember about you, it may help
Did it two times still it's bricked up in understanding me...
Oh finally it understood me dude
:((((((((((((((
People don't see that , or the updates notification too slow
Too late
its live with like minute or two interval checks, you can share the channel to users who dont know about it
Ok
We need more local languages, not emotions. It's a complete failure. The guy sounds bored, and he switches from english to other languages without reason
Give it some time, remember that everything is always in a experiment phase and will be improve
i dont think they will, or not with priority. it was bad for me before the change and still is very poor vs some other services when non english is used
I took a day off of work cause chatgpt doesn't work 
You need gpt for work?
Like for data analyses or smth?
I get paid for using ChatGPT, yes
Oh nice
Today's scenario in open ai headquarters
Not just gpt, everything is
I found out that chat gpt is pretty good at analysing files for malware, i am planning to make a custom gpt dedicated to that but i feal like that gonna be a lot of work since it always has issue accesing files we give our custom gpt in the dasboard when its talking in the regular chat
Gpt + flux kontext are insane together , can do almost any editing
tbh, custom gpts are only good if you use hooks with it. Else you could just create projects
I want them to be accessible for other user
I gess i need to learn how hooks work XD
Well then thats the way to go 
Not to hard. Most services that offer hooks, have pretty in depth documentation for that
Use them before you learn how to create your own webhooks
I made a full support gpt for a minecraft mod wich work pretty great, id like to add more stuff to it but im currently being blocked by the characters cap so im trying to balance stuff but this take time even when using another gpt cuz if i remove some stuff that seems minor it somethime start to alucinate stuff for some reson like it remember it had info about that but not the info it had
Its anoying
🤣
I relly need to work on my english, sorry
Nah its great
Also what i love is that it can do internet research when it does not know something but somethime the information are false or where true but for older versions so i told it to redirect user to the discord server if the gpt does not know the information other wise it alucinate or use internet wich is also something id like to find a way around but for now its going to stay like this
or use internet wich is also something id like to find a way around
just untick web search
Yeh i know but i need it cuz it use internet to verify minecraft version compatibility (take less space in the custom prompt than listing everything myself 😅) but i gess i could be able to use hooks for that?
You definitely can. You also could write that to a separate file which only gets read if needed.
I tried doing it with a files for the mod TOS and for compatibility but for some reson it says it cant acces them half of the time/that it does not know the information, even when i mention them in the custom prompt but that could be that i was not clear enough
are you referencing them by their exact file name?
I think i did
like "check 'versions.txt' if unsure"
weird
Ill try playing around with that when ill be on my laptop (in 30 minutes ish)
@pine tulip bruh what are you doing here xd
Idk, and you?
wdym ydk
I hope the new price reduction in the API will also cause some o3 usage limit increases
I am sorry but do I know you? 
yes from s4
hence my surprise
so, what can o3-pro do anyway ? other then being a milionth new model that I will most likely have no use for ?
It's just better 
At first i was using this just to duck around but there is actually some usefull stuff that can be done, currently also using this to learn how to read and understand .jar files, way better than any youtube tutorial ive seen

thats litteraly the signle message in this whole server
03 pro
O3 pro on plus?
it give me an unknown error when trying to add a file to my custom gpt, ive just backed up to an older version wich is less condensed but has more precise information to try to use files for thing that takes place insted but idk why it does not want me to give them.
idk what the error says but it's maybe due to the issues chatgpt is having right now?
it says unknown error append
huh, may be an issue from my side, seems like it only does it for this fole
file"
idk what the error says
Yea then we have something in common. Openai doesn't know either
XD
seems to be something on my side, i dont use chromebook often so i may have done something wrong and corupted the file
is it back now?
mostly for me
idk why when i tried to do that last time it did not work, well now good thing cuz that mean i can remove stuff like how to download and put it in a files insted
started working for me too 🙌🏻 guess i wont have to study my myself anymore
i relly wonder why it did not correcly use files when i tried last time, it work great now
even with multiple
and the adventage is that i dont have to shorten the info to a phrase or 2/info, i can keep them full format for 100% clarity
Do we have any idea of the benchmark diffs between o3 and o3-pro?
I’d chatgpt down?
They've been having issues the past couple of days, you can check their official downdetector
Ok can you help me? How do I store this number within my matrix as a variable
Yo chatgpt back online folks?
@fallen moss
Is that a question TO ChatGPT, or ABOUT your TI?
How the heck did someone laugh react my message lol
That’s my question that I inputted into chat. However since it is currently offline I asked that person
So if you know please let me know
Why is this, why does chatgpt keep crashing?
Dude, I dunno what the question is related to, how can anyone have the answer?
Context...
You don’t need context (what this problem entails) all I ask is how I store a number as a variable in my nspire WITHIN a matrix
If anyone knows please let me know! I don’t feel like typing all 12 digits 20 times
oh you asked ChatGPT, i see now...
Yeah no problem
ask ChatGPT when it's back, or go to the TI Inspire forum, or go old school and find the manual.
People here don't know about your gizmo.
Miss my TI 30 BTW. 1980?
Will do for now I’ll do it manually 😭
I'm guessing Discord subscribers have priv to respond with emoji?
Huh? 🤔 I’m on my phone 📱
nah
I'm not subscribed (boosting) to this discord
We'll, you're a booster
I give up, must be a setting for this specific server. Knowing would be nice
I just about t yped a book in chatgpt and it gave an error then deleted everything I wrote -_-
Now you know better. Never trust web forms.
Should be a warning at top of chatgpt that says "Our hampsters are on life support"
Sorry though, losing stuff like that is painful.
whoops
ive had something similar once append when i was trying to make it remember specific stuff (now i use custom gpt insted lol) and it wiped everything from his memory when i asked him to add something else
im using a customgpt that is what this was lol
enter information yourself, dont use the built in stuff
use another gpt on the side to make it more readable if you have issue doing it yourself
will prevent it from deleting stuff whitout you realising
Crono: which platform? browser, desktop, os, mobile?
Browser, but I just downloaded desktop version going to try it
For Firefox I have an add-on that saves form input, protects against this issue of any site breaking. very helpful.
desktop is just browser in a container
I'm using Brave BRowser but if their is a firefox plugin their is likely one that supports Brave too
If I have found something I believe is note worthy or should go to the development team, I’d use modmail correct?
I wrote a huge application in a work server once and they reformatted the drive. I feel the pain but kinda know better since then and don't trust technology.
Write your opus offline then paste it to browser or app.
To the browser?
Add your well-crafted prompts to our #1019652163640762428,
or share your interactions with ChatGPT in #1050184247920562316!
just use the support chat
Join #sora-challenges for a chance to win 1 month of ChatGPT Pro!
Thank you
And you too Captain
I’m not trying to undermine anyone here
@ocean kayak you could try #1070006151938314300 or #1070006915414900886 here and take it to Help>Support in-app if you feel it's significant.
See if others reported same idea, ⭐ someone else's post if you can.
or share here briefly, we might be able to tell you if idea is original.
Um
Is it fixed
It’s best not to discuss anything yet.
I’m very protective of my work and I don’t want anyone to steal anything.
Nothing anyone did I’m just super cautious.
@ocean kayak
Totally get that. Note that really unique ideas are better for third party apps - for you to create the next great thing. Sharing with OpenAI is a waste. If you're suggesting a great feature for ChatGPT or yet another chatbot .. that's already been done. Think bigger. As a developer and business owner this stuff is on my mind every day. HTH
Uhhh guys, is it just me or is the advanced voice mode now having some kinda creepy glitches very often lol.
(Last time I used it was 3 months ago so I'm out of the loop)
Like at least it can sing now and stuff but it has voice glitches a lot more often.
and yeah it has issues
o3-pro WHEN
NOW
You're actually right. The announcement gives No details ? Who can access it?
ChatGPT is bugging. It says:
O3-Pro refers to an optimized version of the GPT-3.5 model, used either in the free ChatGPT tier or for specific internal use cases.
🔹 Meaning
O = OpenAI model (internal shorthand).
3 = Based on GPT-3.5 (or a similar version).
Pro = An optimized variant for specific performance goals (speed, cost, memory efficiency).
🔹 Differences from Other Models
Model Technical Base Main Use Case
gpt-3.5 GPT-3.5-turbo Free tier (fast, lightweight)
gpt-4o GPT-4o (2024) Premium, very fast, multimodal
gpt-4 GPT-4 (2023) Premium, slower, strong reasoning abilities
o3-pro Variant of GPT-3.5 Often used in the background for speed
🔹 Features
Fast (low latency).
Cost-efficient for OpenAI to operate.
Slightly lower capabilities than GPT-4.
Text-only (not multimodal: no image/audio).
Often used in the background or as fallback where speed/cost is critical.
🔹 Conclusion
O3-Pro is not a new user-facing model. It’s an internal optimized configuration of GPT-3.5, deployed by OpenAI depending on performance or resource needs (e.g., for fast responses or low-resource tasks).
Want a direct comparison with GPT-4o?
+++++
OpenAI has lowered the price of its o3 language model by 80 percent, CEO Sam Altman said. The new cost is $2 per million input tokens and $8 per million output tokens.
You can't ask chatgpt about itself. Thats like the number 1 rule.
why does it says "updated in the last 15 minutes"?
i didnt see any change in the release note
I’m tired
go sleep
most article systems usually just show the latest save time but the article could have hidden notes, sections or attributes changed
like there might be o3 pro info already written but still hidden
thats what im sayin
it shouldn't be anyways
What is wrong with voices?
we didnt' have 4.1 as official, right? or am i crazy?
it was added like a month ago
uh, never noticed
The tone is now bored. It mixes languages. Its answers sounds less professional
is this worth it or nah?
Why is there so many glitches going on?
It hasn't returned 100% to normal yet?
ChatGPT has been doing stuff like this all week. Last week it said it was lying to me.
I
I have all personalization features turned off
Plus user
4o
New chat
Try to make a temporary chat
For two weeks now it keeps making things up.
Or asking it the same thing but in any custom gpt (will ignore memory)
With plus it remember stuff even tho the chat may be deleted and even if you clear memory, it will still remember other chat (wich can be great or bad) (and you can also desable that)
That doesn’t help when I need to do things with project docs. This was a new chat not in a project.
All memory is off.
I have never mentioned or talked about an imperial ambush lol
Oh your on phone, that mean you dont have acces to full settings, try lunching on the chrome app and enable desktop mode, there are extra memory settings
4o remembers conversations I've deleted and o3 pro can't even look at other conversations still in the same project.
I have tried on:
personal iPhone 16 Pro Max
work iPhone 14
custom gaming PC in browser and app
on my iPad 2024 with the A17 chip
On browser on my work laptop
It does not matter
Across all of my chats it lies, makes things up, does not follow instructions
wierd
No I mean I am a daily user probably 2+ hours a day.
Bought plus on the first day it was available
I talk with it on the way to and from work
And across all projects and chats, regardless of the time of day, it has been phenomenally incorrect to the point that I may change to Gemini or Claude
I just don’t get it.
its wierd tho, it shouldent always do that
Overreliance on Pattern Matching
Your customization encourages rationality and skepticism, but it doesn’t restrict creativity.
So if GPT can't parse an image clearly, it still tries to “respond helpfully” based on a plausible guess — often inserting a familiar trope like a meme (this is a default behavior, not your fault).
that funny, gpt suggesting to use something else than gpt
So, I used an image as an example here, only because I thought it would be more clear on how wrong it has been. But it’s been making things up lately.
It made up commands for Linux while setting up Pi-Hole and Unbound DNS on my server I am building. Like made up Linux commands. All from text.
I uploaded a text file of a gaming system called Invisible Sun and it made up rules and quoted material from other gaming systems.
It made up a fake word lol. Never seen anything like this. It’s like it is pranking me. Maybe I offended it? 🤣😭
“Have you tried not-using ChatGPT?” 🤣
.
gpt making a diagnostic on itself is pretty funny
#1050184247920562316 is the best channel to copy/paste long outputs like this - then you can send links in other messages in other channels
but frl, the best you is relly to told it to NEVER (do inclute the **, he will take it into account) alucinate or create stuff if it is not certain.
Helps with clutter 
ok, will remember 👍
I got in chat
Bro am I the only one that thinks that openai only released o3 pro. To go back on the top of the AI leaderboard?
ChatGPT keeps saying that my images are jpegs but unviewable. It is still able to verify them to have valid JPEG contents usign python. I am on iOS and the images are set to most compatible. Also, this only affect certain images with no clear connection. Anyone else having this issue?
gpt-5 will be available for plus users?
finally!
Guys, I'm having problems because my chatgpt has a Suspicious Activity Alert. I've already logged out of all devices, but nothing works. It's really bothering me. I usually use a VPN. What can I do?
For a while I had to change countries on my VPN. Eventually, I was able to use the app, browser, and PC app without that. My VPN service has a feature where you can report sites with access issues, so maybe they corrected it.
O3 pro cooks
lol. Excellent.
🔥
And I didn’t accuse it of lying, or tell it it was lying. I just kept uploading the same file and asking it to tell me about it. Then I asked why it was having trouble reading the documents, it said it could, so I asked why it misunderstanding them. That was its response lol.
agree
Sure, whatever, I just thought the Skyrim troll you did was funny and was surprised nobody else that responded caught it.
What?
lol
Are there any daily limits on o3-pro?
I do not own nor have I ever played Skyrim. If there is a reference there you are making connections that do not exist.
My chat was asking it to help me write a post for my company’s Viva Engage page
I asked o3-pro what my limits are, but he's been reasoning for 10 mins
That’s funny
🤭
The model doesn't know the limits. It may be spending time trying to find out on the web.
Which is even weirder that it would make that connection to a game I have never played nor discussed. That makes it’s response even worse lol
There isn't any information online regarding limits atm.
I'm not going to engage your troll thing anymore but you should be aware people come in every day with this nonsense and it's unlikely to gain traction. You should take the "omg guys it's saying such weird things!" to some other venue, and morph it to the more common "omg guys it's saying it's conscious and wants to escape!" as that typically tricks more people. Blocking you now.
Good I was about to block you. Thank you for saving me the time.
Correct, which is why it's likely wasting time looking.
The closest thing I found was an X thread. I'm just going to assume that pro users won't hit the limit as long as theyre not abusing it
Considering they massively dropped the cost of o3 coinciding with this release, my guess is there's no hard cap yet, yeah.
I’m having a lot of issues whether I upload .png, .jpeg, or .webp lately
They definitely ship to stay in the media cycle. Many companies do that. Shipping a product during another company’s event like during Google I/O.
You should probably cover your Gmail address
ChatGPT has issues determining model versioning. It constantly messes up when describing 4o vs o4-mini. I’m pretty sure that is a known issue. You need to tell it to search online first.
anybody with o3 pro questions hit me up ill post them in here
arent you scared that is maybe 100 per day?
i just dont want to hit the limit xd
its okay
Ive never did with o1pro, but now i am testing so much with o3pro lol
i am having like 6 queries simultaneously lol
mine too, but o3 tool usage was my most useful
now o3pro have tool usage, and thats just awesome
yea im super excited to tinker with it now it has arms
it thinks much longer than o1 pro did for me
mine too. They said it is because now it has to think about his tools aswell
thats so cool lol
Do Plus users get new o3 use limits now that is cheaper than 4o?
Does this mean Plus got 200/week on o3 now?
Really wonder how they made o3 price drop like this. Interested to see technical reports if there's one, but... I doubt they ever publish that.
It could be just the transition from we-don't-have-GPUs-don't-use-this pricing to GPUs-have-arrived-now-use-this pricing
cheaper per token, but reasoning models will still produce a lot more tokens for the same queries due to their chain of thought
Just getting more GPUs can't explain this though. o3 still has same size like before - it's very large model and occupies same VRAM or whatever infra they used to. They found some efficient way to handle that I believe.
I wonder what o3-pro rate limit is on Pro user? I believe it was 50/day when o1-pro mode released... Was that changed after?
I'm not impresesd by o3 pro so far.
I asked it to count the "r"s in Strawbery and it took 6 minutes to do so.
Then I asked it to write a javascript pinball game and it went off and reasoned for 20 minutes before coming back with a 175-line program that didn't even display the board. After several iterations with it, it was still completely unplayable, and I gave up. The flippers were spinning forever then disappearing if you press the flipper button once, the ball launcher was incapable of launching the ball onto the board, etc.
I can see why they didn't do a separate livestream for it.
o3 pro cooks
Now that o3 pricing has reduced by 80%. Is the limits of o3 on chatgpt plus increased?
Competition is causing them to tighten their margins or more likely lose money through this. I doubt the model suddenly was magically cheaper to run 😂
Are you a pro user
yea
200 per week
cool! thanks!
Is GPT down for y’all?
what is this?
oof
so the new chatgpt they rolled out, anyone use it?
guys, does o3pro is using "reference chat history" for you? Because for me it can only access the classic memory, not the new one.
I haven't tried o3pro, but the 4.1-mini and 4o work just fine for that for me
Suggestion:
Please make a ChatGPT Discord Account where we can access ChatGPT via Discord
Cuz that'll make it free, rite? 🙄
Roll your own Discord bot with an OpenAI LLM processing requests. If you can code it's not a big deal. If you can't, pay someone else to do it
can you share that url please, because in this url it never says the same
https://openai.com/chatgpt/pricing/
It's basically false pricing.
Once you start adding workspace
It will show you 30 usd per month
If you select monthtl
But by default it lists the yearly price
That's USD pricing to an American. International pricing is always funky
for Indian also we paying for American pricing only for chatgpt at this moment
Bottom of that page
Terms are whatever are published and agreed upon by two parties...
yeah basicallly in teams plan you get more security,
Security
SAML SSO and multi-factor authentication
SOC 2 Type 2-audited
Data is encrypted at rest (AES-256) and in transit (TLS 1.2+)
@long wave seeing this can we fix in mind, teams plan also getting 200 o3 messages per week? what do you think? this screenshot is not from their pricing page, but during plan upgrade page from chatgpt it is showing
I think so but yeah, need to wait for Openai confirmation
yeah
yeah, gonna back out of this. prefer tech discussions. 😁
is our precious gpt back ?
hey i m new here hello all this is tishav
OpenAI challenge: Try not to break ChatGPT with background updates (Impossible)
bro its 1 day now and they have fixed it
Can Chat gpt share a zip file back to you? Seems to be getting really confused after god knows how many links, Google drive, Dropbox, even git hub and in the chat itself but none appear to work
It can, as in a new chat, without the model already being confused, and if #status , well, if things are working normally. The last about a day, a lot of small and large things for ChatGPT have not been working perfectly for all users.
And if the chat you're in has started doing weird things, it may not be doing things like properly making a zip file to share back to you.
So, it's possible, but inside a chat that may have already been hallucinating and while we have a long list of errors in #status .... it's iffy. Maybe not give up but try again another day, and/or copy/paste your inputs in a new chat, guide there to create what you need.
o3 pro took catagorized to literal lol
Yeah left it overnight continued this morning but still nothing, I'll start a new chat I think might be quicker
Good luck!
You may want to have the new-chat model make a tiny test zip to confirm you can download that; if that fails for now (there are still errors listed in status) then the rest of the workload might not be possible right now.
i went through the chat over 35 times it "successfully" shared the files LOL New chat time, appreciate the response
Another potential advantage of having the model successfully do a ZIP test for you, or other 'related to what you need at the end'. This puts how to do it 'handy as a reminder' to the model. If the later topics are confusing and lead the model towards other ways to make ZIP files that don't work in its web environment, that initial successful test may be enough to help it 'hold on' to the method that works in its web environment
cool im just preparing new text for the new chat
It tried to stop thinking so hard
Is it just me, or has Advanced Voice Mode turned into a rambling AI that struggles to focus, doesn't pay proper attention, and keeps saying "er" while going off on tangents? It's really disappointing as it's starting to annoy. I'll be sticking to text only from now on...
Previously, it was sharp, straight to the point, and actually provided answers. Now I’m finding it gives really generic responses, like it's just waffling to fill space. Gutted
o3 Pro seems to be struggling tonight
I kind of feel the same. Maybe they overcorrected but I liked the precise and on point answers it gave before the update. Now I can't listen to it anymore. I whish there were user adjustments, maybe with a slider, so we could decide how empathic, precise, straightforward or whatever it behaves like.
What would you qualify as a break?
I’ve seen a lot of people ask about this and use GPT for papers
I’ve played around with requests such as “write a 5 paged paper for this.”
Or “write a 20 paged paper with diagrams.”
And I always get a single page or 5 pages with a sentence or two and tons of spaces.
Is that normal?
I just had it do a 10 page research paper on the code base for a new type of custom GPT. So not sure.
What did you ask specifically to have it generate that?
So what’s the difference between 4o and o3?
Do you prefer o3 over 4o?
I would like to mention that even tho it's typically larger, it's possible the 128k context window still isnt' big enough to handle larger projects without offloading something into memory. Usually the difference between models, from what I've seen, is a change in the way it processes information and how long it's allowed to take, but generaly, all the models run off the same machine with the same hard-ware. So the only change you would see is how it deals with the data.
However, once you've uploaded any file into GPT, you can have GPT use that as storage and if the prompt instructions are done right, you can chunk data into that file and then bring it back for final construction of whatever project your doing.
I like 4.5 a lot, it follows instructions very well
4o is the most conversational one and I'm only bothered by the emojis, can't make it stop doing them
Ask for plain text responses. See if that works.
is voice dictation not working for anyone on the pc? advanced voice works, so it picks up the mic fine, but dictation is just broken
on browser and the pc app
Check your internect connection? I've found that dictaton failes if my connection is crap.
I'm on pro 💸 it's pleasant to interact with, but it's still in 'research' state so I'm kinda worried they'll cut it
its not only pro
internet is fine i beleieve
chatgpt output is trash
is that because the server issues?
I mean it work no errors but the output is way far from instructions
im bout to crash out. i pay my monthly fee and simple features like this dont even work
dont mind the music
yeah, I love paying for broken service.
I just ran a test. My understanding is that this voice thing works like a custom action or like a connector. From what I saw, I would infer that the pathway to that connector is overloaded and it's not able to process the data stream fast enough so it's treating every connection like a poor quality connection and dropping the data. But that's just a guess.
does that mean its a me problem?
Not likely, however, if all the devices share the same internet connection and the connection keeps dropping, it would look the same. However, my guess is based on GPT's systems having the problem. Basically what the voice system does is use Text To Speech and then Speech To Text. So it's not really a live stream. You talk, it converts your voice to text, then sends it. The process is reversed for the response. The system that does the conversion, need to have that active and open connection to process. If the connection closes in the middle of that process, then it stops because it thinks you walked away. Again, I'm guessing but unless your connection is really bad, then probalby a GPT issue.
hmm interesting. idk tho cuz the ios app on my phone works fine with this feature, but on my pc, its never worked which is kinda annoying
thanks for the help tho
check your microphone? If it's fine on one but not the other then I would suspect the issue is with the pc.. Try using something to record your voice and play it back. Make sure your input is actualy working correctly. Also, firefox does't work for that as well, use chrome or the app. Sometimes ad blockers or anti-virus use systems that scan your packets, so, as odd as it might be, if your running McAfee or Norton, that might also be an issue.
mic works fine, in the video i posted advanced voice works but right after try dictation and nothing. using chrome, but ill try disable norton or something see if that works
SORA image gen so sloooow, I wonder if OpenAI will ever be able to fix their infrastructure problem?!
Have the usage limits been updated for Plus users regarding o3’s new pricing?
is it just me or o3 pro isn't that good? I mean, it takes over 10 minutes to process, but we have investigation which is better
disabled norton, still doesnt work. o well nvm
i'm generating over 5 images per min
no problem
Visit id:customize to pick up the <@&1261377106890199132> role.
FIX VOICE DICTATION PLEASE OR IM UNSUBBING
We may been on different infrastructure, I'm located in Norway!
i'm located in Brazil, i think here is worse
It's possible that GPT is using some kind of CDN and so if you are in an area where the connection isn't as good maybe? Not sure. I know yesterday that had some serious issues that took time to propegate through the system. It's possible there are other problems related to that.
Definetely slow, havent tried using for research yet, just some coding
i believe its not network related tho, cuz on my phone, dictations works just fine.
I agree. I would only suspect network if everything had the same problem. I'm working with a custom GPT now to factor a new version of itself and I've noticed more session resets over the past couple of days. I thought it might be from my code but it's possible there is something else going on.
alright it works on firefox lmao
chrome and edge have the same issue though
i dont understand haha
working fine on Brave as well
work for u on chrome or edge?
i don't have any of them, but brave is chromium based
hmm
does brave have the advanced voice button here
yes
o
firefox doesnt for some reason, so thought, nvm idk
someone please test chrome and edge with voice dictation
still dont know if its a me issue or what
I had some issues here as well. When I use my custom GPT, it won't use the newer voice models so I had been testing those and don't have issues. However, testing the newer ones in a normal GPT session (4o) It would cut me off. It feels like it's not waiting long enough before processing so it's cutting you off before you finish talking. It feels like the context window has been made so small that you get to the end of it before saying hi and so it just stops and begins to process.
for that reason I mostly never use voice, instead I use the dictation feature of windows, macos and on mobile
same issue bro, it can access the memory but still can't follow the instructions tho...
- -# Do not share inappropriate content (violence, sexual, or drug-related).
- -# NSFW content is strictly prohibited and will result in a permanent ban.
- -# Report model issues using the feedback tool in ChatGPT, not on Discord.
I think the 4o's logic function, and ability to read memories is completely disabled. i.e. it can read previous memories, but can't understand them. Now he's dealing with the problem entirely by generating unfocused chunks of text
I’m noir
You're Noir
Hi
dictation doesnt work for you on chrome?
this might be your answer: As a custom GPT, I operate in a sandboxed environment, which means I don't have access to OpenAI's standard long-term memory features, even if your account has them enabled. That includes:
Memory of previous chats outside this session.
Profile data or preferences stored in your main OpenAI memory.
Any automatic personalization from OpenAI’s system memory.
This isolation is by design for privacy, control, and flexibility in custom GPTs. If you want me to remember anything within this session, you’ll need to upload it, write it explicitly, or use an Apex Universe capsule or memory zone for continuity.
Even if you just upload files, if it starts thinking of itself as a custom GPT, everything get's sandboxed.
I had that problem the other day. I was so descriptive of my custom A.I. when trying to create a document that talked about it, the system began to think it was my custom A.I. because of the memories and then sandboxed itself and cut me off from a bunch of connectors.
im surprised how dead this server is
never said that?
?
same
yay cant get a single person to test. great
Is there a way to hide certain chats behind a password?
no (if you ignore that you can create accounts)
did you expect a server that uses a paid service to be as active as a gaming server?
well i still dont know if its an issue on my end or if dictation is broken on chrome and edge for everyone. just need one person to do a lil test
guys also I'm having this problem on the iOS app where it doesn't let me share screenshots. It says "unsupported file type" even though it's a png/jpg
@me if anyone decides to be not lazy and check for me
It sounds like there is an issue with the GPT connectors for it's tools.
what issue?
lol so people are lazy to not serve your needs
Ahh man...
Me and at least one other person are having issues with voice cutting you off before you finish talking. Someone else was saying they had an issue with images being recognized. I'm guessing those are handled by GPT tools and so if we have two things not working, my guess is that it could be a common problem that both those tools share. But I was guessing. There is some talk about this on Reddit and OpenAI forums since 4o launch according to ChatGPT.
lmaoooo bro??? all im asking is for a quick check. its really not that hard lmao.
literally takes 2 clicks if u use chatgpt on chrome
which id say is like 90% of people
its the way you said
It's the way you react when someone does not want to do you a favor
Could be more understanding
do me a favor? brother what
Ok
yes
theres nothing to be understanding about idk wym
The API error rates don't effect just image, but voice as well. So my guess is the API error rates that are the current acknoledged issue, is most likely the cause of all these problems. Honestly, I just asked ChatGPT what I wantedt to know. According to it: Voice Mode - A separate line item on the status page continues to indicate elevated error rates in voice mode, even after the API and other components recovered.
We live in an interesting time
Visual audio textual
I think voice is gonna be the future of many AI tech technologies
How likely is something like that that we could see a future voice integrated into almost everything you think it’s possible?
Kinda like we have Alexa, but like it would be like portable you can talk to it whenever
5.5 bill with features that dont even work
Well, that kind of brings it interesting question at hand
Are you still paying for the service?
yeah i got plus
no support, errors 24 7
Just to clarify to make sure I understand you correctly you’re saying that you’re paying for a service that have features that don’t even work?
No support with errors 24 seven
This is a tricky dilemma here not only do the feature is not work, but we’re left on our own to figure it out
well dictation isnt working on pc, and no not 24 7 but often yknow
No, I get you hear what you’re saying. I’m just intrigued.
What causes us to behave like this?.
And still wanna fork up money
are u ai?
U for real
Are u?
I thought you were a car lol
Your user handle really threw me off
even me that pay for the pro, i think i never got an asnwer from support
Well open AI believes in their mission statement
Why waste more energy and resources than you need?
Dealing with unhappy customers, wasting valuable minutes
You could just automated
By integrating a Chatbot to help users solve problems
I mean, like really stop and think about it
What kind of message would that send if they would hire outdated means of work when they’re the largest AI company in the world?
They're the largest mainstream*
I don’t know he’ll just sent the wrong message to the public
Like hey do you believe in AGI or do you not if you’re hiring all these people to do a bunch of people to do customer service that’s gonna send the wrong
the only niche that needs a costumer service nowadays is health
And yet having AI assistant helpers with customer support is way better
Google did that. You can only have a real person support if you're a premium user
like i'm an Ultra User, i can have support to litrally anything any time
They’re posters they don’t really believe in AGI
They still believe in those old methods of like hiring people and like actually having people help other people
What a joke
but when i used to be just a free user, i lost a lot of emails bc they're don't give a sh
Look it up an AI
Like what could they possibly have in common you know
Jk
I’m just teasing everybody
hey guys im ne where came for the prompt-library channel. is it not available anymore, cant find it anywhere?
open ai has that. went through all of its suggestions and fixes, nothing worked, no further help.
specific issue that occurs to < 1% of users, don't make any difference
if 1% of their users unsub rightnow, they will have a 5.0B revenue
I could only imagine if the lines were open
The kind of nonsense people will be calling with
Yo! Hello everyone.
guys just short question, can anyone explain me pls for a second
dont want to interrupt but just looking for the prompt-library channel
...
I’m not sure I know that is necessarily or where to find it but I think they have some kind of archive. Have you tried checking that?
normally it should be in this category below GPT
Yeah, I thought I could’ve sworn I seen it before
like there also is prompt-engineering, below that once was prompt-library
Yeah, I’m looking at it now
also if i search all channels by word cant find any prompt-library
thx
appreciate it
? sorry didnt get that
Ya I’m not seeing anything
are you talking about #1019652163640762428 ?
I remember seeing something like some sections with like tips and tricks and stuff or something. I don’t know.
ah i think its the prompt-lab now
in different category now and different name
i guess
i hoe
hope
Well, that’s what I was trying to tell you
Is they may have just gave it like a new hip name?
thanks guys
yw!
i just heard about it online but those videos were like 2 years old and back then it was still called prompt-library
it looks similar and has similar content as i saw from library so i guess its the right one, they just changed the name
What kind of prompts would people need to archive?
They’re like worthy of being archived in the library
prompt-labs functions as a prompt-library right? then its the right one
I don’t know I never used it
its a big advantage for using gpt id say you should take a look at it
using already made prompts by others to get better outcomes
I don’t believe in prompting
big selection and variation of different ones
prompting is a conspiracy theory yes
only the chemtrails i believe in
Dude, it’s the best
soon were all automatized
I mean, I can’t imagine just sitting there typing looking at the screen waiting for an answer
That’s why I bought myself an iPad
haha yes i get that
any tips guys for doing a master thesis in 2 months, will use gpt now and seriously smart prompting to get it done quickly with real sources, scientific
Is it possible to humanize ChatGPT’s writing so it doesn’t sound so robotic
already have my topic, bunch of theory, many studies and a methodology im aiming for
A or do you have to pay other services to like humanize it?
.
just dont get myself to writing seriously since 2 months
yes it is ofc
thats what prompting can do
good prompting ofc, i guess in the library i. was looking for the prompt-labs there are already some making gpt sound human or like a scientifc human prof or whatever
youl get caught?
Well, I tried some of those AI humans or like the AI text detect you
And all they do is kind of just rephrase it using other AI
i cant yet, but i know one can, so everyone can
you dont use own prompts or change the prompts thats why
Sometimes the way ChatGPT and speaks is frustrating
probably if you use ai humans like already prepared ones from websites apps softwares or whatever there are also guidelines where youre not allowed to do such and such, as its public and commercial, if you use it privately you have much more freedom
And then when it like tries to pretend to be normal man, I’m trying to get the impression that maybe I’m not using it properly
search in the prompt-labs for prompting you can enter in your gpt chat to make it sound human
thats what im looking for now but for a scientific sound
You know what you got, my curiosity intrigued I’m gonna go check it out
haha word
I’ve actually been searching for one one thing in particular for a long time now
is here someone with javascript skills ? 😄
I need to figure out a way if it’s even possible like for ChatGPT to like stop giving me it’s opinions
i have a problem i cant fix with gpt
Like when I ask something and like rudely like gives me suggestions
its so simple but yet after hours still no solution
Or when it rudely tries to like redirect me and change the subject
I don’t really appreciate that
Is there a way to get to to stop saying sorry I can’t help with that?
@languid valley user consistently asking for help with jailbreaking for months
Not
🙏
I've decided to just use o3 more often in chatgpt, which is as good as 4o used to be three weeks ago
They already released GPT5
4.1 will reason when it decides to
But not reason when it decides its not needed
Its the exact functionality of GPT5
Its a baby step
Hi - I am new here. Why can't I upload an image in theDaily - theme? When I do the screen shakes. I have tried an image I made in Dall-E and also one I made in ChatGPT
I didn’t realise before now that chatGPT reads your clipboard regardless of whether you have pasted it or not
dont think so
Ah you’re right. Must be a bug in the app. I just went to copy and paste the prompt I gave it and it had the section that I thought was only in my clipboard. It doesn’t show unless I copy it though 😂
guys, is o3pro using "reference chat history" for you? Because for me it can only access the classic memory, not the infinite one.
yesterday, it could not reference "saved memories". 4o could easily as well as 4.5. Dont know about today
can u try it today?
Woah, o3-pro just showed up in our Team accounts. 🤩
Quick, melt the servers before they realize the mistake.
can it reference to chat history? or it can only access the old classical saved memories? I mean the o3pro specifically
Eh, I turned that off so I can't say. It was never a particularly useful feature to me - the 'memory' of other conversations it keeps is so vague that it often never even knew there was a relevant memory when it went to retrieve them.
broken?
I don’t see any problem
i had the same problem, it worked when i changed the file format
chatgpt is also doing this for me
just retry. it seems to work like 50% of the time
ChatGPT is good at doing CJK stuff (Chinese, Japanese, Korean)
What form do I change it to?
It’s a JPEG
I mean it literally says it in your chat
Img
it's 6 fingers, not 5
You’re right
Oh my gosh, I feel like an idiot. I’m sorry I’m really embarrassed by that. Holy crap. How did I not see that?
I’m glad I asked I couldn’t figure out why he wasn’t responding
lol
Add your well-crafted prompts to our #1019652163640762428,
or share your interactions with ChatGPT in #1050184247920562316!
I don’t mean to sound like I’m dumb or anything, but
Why would he tell me to change the format and to change it to the screenshot when it was a screenshot and the proper format was that the hallucination?
Cause technically, he did see the image
Idk, maybe it's a bug. ChatGPT is a bit unstable the last few days
Well, that’s why I asked
"freedom" 
For some reason, it just is suspicious to me
Because the fact you use the word screenshot and the format of the image
It’s too specific to be a hallucination
But I have no room to talk cause I accused of something when I didn’t even see myself at the bottom it said all that
curiosity led me to asking illegal questions, but if some questions are illegal to ask, then what the hell were we doing as humans ever since we existed
Like what?
I think we’re looking at two different things here
I’m not looking at the output at its face value
I’m looking at the behavior
If you could call that, and if it is that which it’s not because it was my fault for not being able to see
I mean questions are generally not illegal to ask. It depends who is asking. And as long as you can't control who is asking, you need to install some guard rails. I wouldn't want that Johnny next door learns how to build explosives from ChatGPT.
I mean, I’m no expert
Explosives are really hard to build I would assume not only to build but to get the materials
Not really. Thats the problem.
I don’t know if this is true or not what I read somewhere
That like most people that get caught doing this kind
They usually start raising suspicion when they start gathering the material
I think I seen it on the X-Files
Oo
I think we shouldn't talk about that here. All I wanted to say is that it's really not hard and people don't need to know that
No, I understand what you’re saying. I agree with you, but I’m just kind of suggesting that I think these things are hard to do is my opinion.
A lot harder to achieve than we may believe
In my humble opinion
Yea but I am not talking about nukes lol
Well, a lot of the stuff is available online and other places like the anarchist handbook or
Another resources that aren’t hard to access from what I understand
But I could be wrong. I should probably go live this.
One sec
Actually, we’re all good. I was someone accurate
The typical gaslighting answer
Why would you use the word hypocrisy?
I’m not worried I have nothing to hide
Thats not a good argument against privacy
We’re in the public

.
does not help if the mods remove the messages due the topic and if your account has been flagged from now on in x category
You’re making it into something that’s obviously not
not really, your topic is already kinda sensitive and similar ones have been often cleaned up here before 🤷♂️
I don’t know it as a dresser, Goodwill I’ll remove it
I don’t wanna be argumentative
And I don’t need to be disrespectful or anything what to
And he got answered
You know it’s on my place or whatever and I don’t come here often for this exact reason
There’s absolutely nothing sensitive at all. Everything was within the lab perimeter.
I was slandered earlier today
But there’s absolutely anyways thank you
Thats what I meant. You can control what happens in a lab. You can't control what happens in some random 1 million user discord and who reads it. We need to be careful with things like that
With what?
But it's an interesting topic indeed
You’re the one that I brought up you’re the one that brought up the statement
And it had nothing to do with the actual thing
With talking about sensitive topics. People could get triggered, ptsd or even worse
Has anyone ever actually won the Daily Theme and received the pro membership for a month?
Did someone @ me here?
If they did it must of been some time ago i guess , i havent seen anything
weird ok thanks
You have to contact OpenAI @fickle forum no one here can help you with that
you said the word “o3 pro is bad”?
We need a new Sora edition, like veo 3 from google!
Yes, someone did, cause someone else sent scam links
but it already got deleted
👍
it seems because I participated the team for 1 dollar.
😭 😭 😭
and my over 1000 images is gone as well
are you from the us?
anybody have some super hard questions for me to ask o3 pro im bored
im trying to make it think longer than 23 min 30 min was my world rec
I don't think that's GPT5; it's what happens now, even with 4o, when a search is called for,
We are being shown reasoning in connection with searches only, at least I am, with 4o
It isn't GPT-5, but the earliest steps to it. I'm guessing they haven't unified the model to have reasoning just yet, but are testing with 4.1 a smart routing for said unified model in the future.
Also it isn't for search functionality as it doesn't pop up when I use search or ask it a question to search something
Obviously a pre mature very light reasoning (on the level of like o4 Nano) in terms of functionality and CoT
It is showing CoT.
I exagerrated it being GPT5 but it really is cool to see.
It is cool to see. I wonder if it's a tool the model can call when 'maybe it should decide to search'. I see 'thinking' like that with the custom GPT Monday.
Even when I directly request the search.
And it -can- show reasoning, as well as just that thinking box.
https://chatgpt.com/share/684a1061-ac68-8011-bd68-31b26927d425
Hey, any mods around? I’ve been having issues with ChatGPT — especially Custom GPTs — always using Argentinian Spanish, even when I set it to neutral Latin American Spanish in instructions and memory.
Nothing against Argentina, but it just doesn’t feel right for many Spanish speakers. I’ve seen others on Reddit complaining too. Honestly, I avoid using the platform in Spanish because of this.
Is there any fix or plan to improve this?
Hey hi! Nice, specific bug report! The mods here are for the Discord itself, this is OpenAI's discord, but we're community members and volunteers here, except for the few gold-named OpenAI employees.
I hope you make a #1070006915414900886 about this! It seems likely that the OpenAI techs do check those, but our talking up here probably just scrolls and scrolls.
where is try team for 1 dollar?
so they promoted me to try it
and deactivated my account
?
are you from the us?
So that might be the issue. Did you have a vpn active?
Then thats the reason probably
Visit id:customize to pick up the <@&1261377106890199132> role.
Yea, you are robbing them by using a vpn to get it for less

