❔technical-support

[prisma crypt]

ive been trying to get chat gpt to do things like generate something in powerpoint and make a graph for said presentation but it keeps saying things like "Since my plotting/file tools are currently failing" and other errors it claims are on their end is there any fix to this bc i was getting it last week too

[surreal spear]

hi, i have programmed an environment to answer on questions. The model is llama3. I have articles which i partitionize in chunks and save it in an database for the context i gave the model. Each chunk and the question have a vector created with mxbai-embed-large. Now i make it so, that i compare this vector of the question with the vectors out the database. I do it about pgvector. Also i gave in the keywords out the question that i get with phi3, a small modell, about json and convert it in a string which i uses to compare in the query with the vector (0,6 vector, 0,4 (single) keywords). My chunks have a size of 250 - 750 tokens. I ask nearly the sentence which is in the articles, but get not the right part. Have anyone an idea what i can do to get better answers? The chunks are splitted by headlines or new lines and every chunk have the headlines before or in the chunk. Sry for my bad english. Thanks for any help.

[shy ginkgo]

ChatGPT cannot create powerpoint presentations

[hybrid sparrow]

Where is the best place to ask for help setting up a new Jetson Orin Nano Super?

[slender wagon]

India

[hybrid sparrow]

India?

what channel is this?

[livid swift]

Indian channel

[molten vortex]

Chatgpt box where you enter suddenly gone on mobile

Also the log in issue where I press log in still have white box

[static lynx]

this is interesting

[gusty basalt]

who are you?

[gusty basalt]

do you have schizophrenia

[timid nova]

No, BUT you'll still need to sign in with your Apple ID. That's not too difficult though, it's just like how you probably use Google single sign on (SSO) various places

You can't switch from SSO to username/password though, once you pick one, you have to stick with it. Presumably your Apple account will still exist. Congratulations on making the jump 😎 Android and Apple phones tend to be pretty similar these days in every way but price, for better or for worse

Which meds? That's great to hear, keep it up 😎

[ivory ridge]

Thank you! I take Risperidone for schizophrenia and Vyvanse for ADHD. Fun fact : if ADHD goes undiagnosed the chances of developing schizophrenia increases by a factor of 5. I went undiagnosed as a kid and ended up self medicating with recreational stimulants which as you can imagine didn't go very well (it did at first).

The recreational stims led to schizophrenia. Don't do drugs folks! Or just don't abuse them...

[timid nova]

Lol that's probably relative to the non-ADHD population but yeah, I imagine that would be explained by stimulant abuse and maybe some sleep/ social mediating factors in pretty much every case.

FWIW, stimulant prescriptions don't prevent you from abusing them, I'm going through a list in my head of all my friends who abused their Rx stims 🙃

One of them did, in fact, develop chronic psychosis, though he had a lot of other things going on

Another one ended up becoming an alcoholic, just got out of jail for violently assaulting someone in their home, probably while blacked out. And the third friend is doing great, owns his own place and a small business, etc. Lol.

Really the entire range of outcomes there. I had a college roommate who also got addicted to stimulants hard while I was living with him, which turned him from weird to intolerable. He's a surgeon now 🤔

[lofty walrus]

hello is there someone who can help me here? 😢

my chatgpt keep saying ### ‘“Unusual activity has been detected from your device. Try again later.”’

and i didn’t even know what i did i also tried switching emails but it still doesn’t work i even tried clearing all the data but still doesn’t work and its been months

thankyou to anyone who can help me 😢💕

[shy ginkgo]

Look at this channel's pinned message for hints

[zinc folio]

my chats, keep disappearing and reappearing? But I’m not archiving them or anything.

Anyone else got the same problem?

[steel roost]

Is anyone else's chatgpt suddenly being really repetitive? I've been trying to regenerate answers and I'm not even exaggerating when I say they are almost exactly alike, usually starting/ partly word for word as the one before

[zealous heron]

That’s been happening with me. I use ChatGPT to write stories and it keeps rewriting the same chapter over and over and over again.

[steel roost]

I use it for brainstorming and character insight and it's been legitmately saying the exact same thing word for word, which it didn't used to do. I think its because of the new 5.2 or whatever. It's awful

I hope it stops doing that

[zealous heron]

Yeah same here I switched back to 5.1 but I’m gonna start a new thread and see if that helps.

[steel roost]

I'm sadly free so I can't switch. I miss the options so much, but $20 usd a month is crazy.

[steel roost]

I'm finding that it's even worse in new threads

In preexisting threads if you word it right you can call it out on it and it'll kind of fix itself, but not fully

[zealous heron]

That’s what I was just dealing with

[steel roost]

How is it that every new release makes got worse 😭

I legitmately believe that sam altman actually hates AI from the way he's always talking shit about it, but i think he likes being known as the frontrunner of it

[cedar token]

Is anyone else experiencing this, apparently now everything is locked behind the 200$ subscription

[gusty basalt]

show screenshot

[cedar token]

See? Because now I have nothing from before, even the freaking color for the chat bubbles.

[gusty basalt]

that just says you get unlimited access with 200$, not access in general

i have the $20 plan and i still have enough

[cedar token]

So I'm supposed to be happy that I no longer have access to the other models and stuff? Are they that greedy?!

[gusty basalt]

?? you do still have access

[gusty basalt]

oh on mobile

[cedar token]

See?! They took it freaking away!!

[gusty basalt]

i have it on mobile

[delicate dome]

same

[cedar token]

I'm so confused then...

[gusty basalt]

¯_(ツ)_/¯

[cedar token]

I tried refreshing it, still nothing. It can't be because of my Internet could it?

[gusty basalt]

probably not

try force quitting the app and reopening it

[cedar token]

I hope this is a glitch because I had so many pictures that might be gone now.

[gusty basalt]

why would your pictures be gone?

you can still see your old photos and chats lol

[cedar token]

Okay so either it updated and my Internet is just shit, or someone the AI found a way to gaslight me

[viscid fulcrum]

I've sent several clips before of me singing or speaking, for the AI to give me advice on how to do a specific accent, or on how to make a specific sound. Usually I just send a video clip with a black screen and it has been able to listen to it, but that didn't work today. Oh well - maybe it's better if I send an actual audio file, so I sent an m4a file - still doesn't work. The AI itself said that I should try making a new chat and send it from my PC, and that I should try mp3 - so I did just that. But, it still doesn't work!

What could be causing this? Is it a bug? Or have they changed anything?

[gusty basalt]

gpt I don't think can view audio clips other than the realtime mode (afaik)

[rapid zinc]

Just ask chat gpt what's going on 😂

[rapid zinc]

Yes. Also had that problem. With generating images it's suddenly very difficult

[neon cairn]

I need help guys can anyone tell how to build ai like pdf extractor pdf having images diagram and text just upload pdf ask questions ai give answer from pdf exact text anybody can tell

[gusty basalt]

that is a lot of words. you're looking for OCR (to extract text from PDFs) and RAG (to give the model context and info). ask gpt to help you and to explain those terms

[proven saffron]

I just wasted PLN 100 for a Plus subcription. Mid conversation it assumed i’m a minor and locked me in child safety mode. Now it refuses my prompts. The ONLY reason I bought plus it to get access to 4o and 4.1, since heard they’re killing it in january completely, and ever since 5.2 came out gpt has became completely unusable, it sacrificed its intelligence for extremely strict controls that flag you and spit psychobabble and legal code warnings even for using a hyperbole. Now I’m censored again! Shame on open AI

I’m also hesitant about uploading my personal info since I don’t trust corpos

[proven saffron]

ON TOP OFTHAT I CANNOT ACCESS THEIR SUPPORT PAGE

i fucking hate corpos

[green locust]

Try LLMs from other corpos. The're bit different. Not saying they're much better but... different.

[shy ginkgo]

I'm curious to know whether "being identified as a child" has inspired any change in behavior or writing style for you

oh he left

[neon cairn]

@gusty basalt does possible free open source

It's really helpful

@gusty basalt gpt issue with ocr not exact text came out when ask

[pine pasture]

Is GPT-5 what I have installed on my phone right now?

[gusty basalt]

what?

[gusty basalt]

yes, the default model for the ChatGPT app and website is gpt 5.2

[main path]

Broos Tried deleteing/reinstalling app, logging out/logging back in. No VPN or anything similar. Why is this happening and how do I get back my account brodied ? What am I paying for ?

[glossy osprey]

ChatGPT on the website sometimes at random seems to be capable of sending both images and text back in one reply, and sometimes not, as in in some sessions it is capable and in some sessions it isn't it seems, is there some kind of setting to just make it always capable of doing things/

[silent locust]

Is anyone else facing issues with their chat history on chatgpt web? I'm on the free plan and am unable to view my previous chats since ~1 week, but searching for the chats brings them up

[tacit shoal]

Anyone else have issues with ChatGPT lagging terribly with really long chats? I don't mean taking a long time to create a response, I get that it has to look through the entire chat to formulate a response. I mean just typing. Like I click in the chat box and it takes 20 seconds just to process the cursor being in the box. Then I start typing and get about 3 characters before it gets stuck again for 15-20 seconds. Each time my browser tells me it's not responding. It only does this when the chats get really long. It doesn't make any sense, the page shouldn't be processing anything until I submit my prompt. I usually get around this by writing my prompts in Notepad and then pasting them in (I do this anyway to avoid accidentally hitting enter in the middle of the prompt). I can't even rename my chat because it hangs in the middle of typing the new name for the chat, and when it stops hanging it has apparently given up waiting for me to type the new name and I have to edit again.

[glacial ravine]

Hey all! Very frustrated with 5.2 constantly disregarding Memories (which are on), all previous configurations etc, and I have to go back to my 4o conversation to ask why it has no knowledge about the Personas created

[lethal wing]

Hello. First time writing here so sorry if i break some random rule i didnt know of.

Chatgpt asked me twice to perform age verification through photo, and both times the identification failed. The next day, both accounts got terminated for no reason, or at least thats what the site says when trying to login. I also cant use the chatgpt support link the notification gives since i need an account to login (which got deleted) and i cant retake the age verification test for the same reason.

What do i do?

[gusty basalt]

was it the official ChatGPT site? I've never had age verification and I've never heard of any other users doing it

[lethal wing]

it was straight from the chatgpt mobile app, it said "you have 60 days to complete age verification". i even looked it up on the chatgpt support website, which confirmed i needed to take age verification.

[gusty basalt]

oh that's strange I've never seen that so unfortunately I can't help with that

[lethal wing]

welp, at least i tried

[steel roost]

Is there no longer an image library??

My library has been replaced with 'images' but it doesn't snow any generated images like the library did, it's just a seperate image generation thing, and the library was the only way I could view my images most of the time because for some reason they won't show up in the chat

[south bough]

I opened ChatGPT, and clicked onto one of my previous chat sessions from the sidebar, one from yesterday.

Except it now says "0 messages remaining. Start a free Plus trial to keep the conversation going". I haven't bought any thing from ChatGPT, I've only been using the initial stuff when i created my account with google. This is the first time I've ever had this happen.

The send arrow button is greyed out too. If i click "New Chat" the same message pops up again. Is my chatgpt broken, what is this?

[mighty goblet]

Hi.
My chat is glitched.

In the update it was in a project and then in the update they ringed the ability to edit prompts in projects.

But now when I take it out of project I can not edit that prompt It is the newest updated version and the best iOS version on the newest phone.

Why is this happening

[trail falcon]

Im not normally in this chat so sorry if its been asked a lot (i didn't see any messages about it) but has chatgpt just taken a massive crap for anyone else? its like they completely disregard anything in the "personlization" or "memory" sections. like no matter what I put in there, it just starts making stuff up like a toddler with the brain of Einstein.

before it was super efficient and would give me exactly what I asked without all the long winded completely unrelated stuff. it doesnt even understand context anymore

[primal perch]

Yeah, mine's crapping out. It corrupts conversations every time it has the automated feedback thing to choose between 2 images, and any thread that involves more than 1 image fails to gen any more.
I just want my spaceships ;-;

[trail falcon]

lol I feel that. something major with me is that if I upload a file, edit it, then reupload it; GPT will disregard the file I reuploaded and just go based on off the old memory and be like "this issue is still here, change ....etc". Im going through and trying a cleanup/wipe of the memory and such because I know thats caused issues in the past

like just now I asked it a simple question about a software that is a "frontend" for data and I asked it a question about my code and it starts rambling about how this frontend software isnt "hiding" any data...I never asked or even suggested it was

[trail falcon]

*incase anyone else has this issue, cleaning up the memory and personalization settings then clearing out the instructions for projects seems to have helped. I also put a prompt in the personalization section that says to prioritize context

[near scaffold]

can't log into sora on my brothers openai account, he pays for it and wanted to generate a video, it just says "account already created, log in" when you put your DOB in, i have sora on my account but can't generate videos because i don't pay for it

[near scaffold]

can't generate any images on mine, is it down?

[hasty gazelle]

Hello , I’m giving away my Sony ZV-E10 camera along with the 18–105mm f4 lens for free, both in their original boxes and in nearly brand-new condition. I purchased them during Christmas 2023 but only used them for a few test shoots before keeping them stored. Since I recently got a drone, I decided to pass these on to someone who truly needs a camera but may not be able to afford one.

This is strictly on a first-come, first-served basis.
IF YOU ARE INTERESTED DM.

[gusty basalt]

scam

[near scaffold]

sir we don't need cameras here

[zenith dirge]

Im getting this only in one Chat

[zenith dirge]

When I go to the site, it's fine, but it's on mobile and only in this chart.

[opaque flower]

stop spamming

[fervent socket]

Hi

[broken blade]

Anyone knows why this is happening to all pictures I try to create?

[flat trout]

Whenever I upload files in a Chat, I can only send like 1-5 messages (or more, I don't even know) and then i have to wait a few hours for the next messages, now I want to know, which plan has no limit, or what limits do the plans have? It often just says "more"...

[vivid shadow]

Okay so this been happening for a while and I email already and tried to use what they gave me but each time I send a prompt It keeps saying internet offline despite my internet on BOTH ON WEB AND APP.. is it just me or... but is there a way to fix it if it just me

[marble flax]

Hi, has anyone meet the issue in GPT today that, when uploading a PDF file, it shows "Unknown error occured".

I also tried to compress it into a zip file. But GPT can not even detect the existence of the zip file.

[marble flax]

Oh I already see four posts discussing similar issues.

https://www.reddit.com/r/ChatGPT/comments/1qppw3i/chatgpt_cant_access_my_uploaded_files_today/
https://www.reddit.com/r/ChatGPT/comments/1qpzwp5/cant_upload_any_files_on_my_macbook/
https://www.reddit.com/r/ChatGPT/comments/1qpv577/problem_w_uploading_files/
https://www.reddit.com/r/ChatGPT/comments/1qpwmse/unknown_error_has_occurred_when_uploading_file/

[haughty river]

Hello everyone!
I’ve got some GPUs like 4090, A100, H100, A6000, etc and I’m looking to rent them at a very cheap hourly price, cheaper than any anywhere.

If anyone’s interested or has contacts, it would be great if you could reach out!

[tulip glen]

hi i have few doubts about rag can any one expalin
i have one form multiples users can sbmint eg: loan request form there are nearly 1L submisson in that form record . i want to get some info form that reconds how the rag works here or i am going in right path or i need to fine tune the llm for this or what i need to do.

[crimson tundra]

Heyyy, mine says there's a prob on server, is that normal or?

[coarse rivet]

Is there anywhere I can request a trial for chatgpt plus?

[gusty basalt]

there are no trials

[coarse rivet]

Open AI site says you can get a trial via referral

[gusty basalt]

thats not a real thing. i have plus and others have plus and none of us have seen it

[shy ginkgo]

where does it say that?

[gusty basalt]

its some openai forum post thats been around since gpt4 turbo days and occasionally people think its real, idk why they still have it up

[coarse rivet]

Does anyone have any pro tips for bypassing restrictions on a custom gpt? I found one for my review cycle at work. I'm dictating via voice and I want it to refactor and improve my input.

I actually did this successfully on an old account but now I'm failing to and that old account is gone!

It repeats something about analysis only replies

[runic vector]

hey all, i am not sure if this is the correct place to ask but my chatgpt 5.1 (thinking) went sideways since yesterday. it is not following custom instructions and all. is this a common problem?

[ashen fossil]

Maybe an unstable system? 🤷🏻‍♀️ often OAI "fixes" technical failures in real time, this affects users as well.

[molten vortex]

Why have one of my accounts, when generating message there is a white box saying message ? other accounts don't have

[gusty basalt]

can you show a screenshot?

[molten vortex]

Like this

[gusty basalt]

oh 🤔 that's strange I've never seen that before
do you have the mobile app?

[molten vortex]

No this is the website

It only starts generating like this recently

[gusty basalt]

that's strange let me try it

yeah it doesn't show on the website for me

[molten vortex]

It doesn’t show on my all accounts as well only one

[simple ocean]

Not sure if this is the right server, I don't believe a server for Meta AI's website exists, but can somebody help me figure out why I can see the names of the chats in my chat history, but cannot VIEW the actual contents of those chats? Have they just straight-up been erased because I haven't used Meta AI in about 2 years? I needed to access a conversation I had where I was doing some brainstorming for my novel encyclopedia and had a fight scene between a psychic boy with the power of the stars in his core and four evil spirits.

[near radish]

Your message has been deleted, sorry!

yonskrt

Rising timeout

You have been muted for 13 minutes for the following reason:
flood score of 1.0

This is an automated message. If you believe this was a mistake, please contact a moderator.

[gusty basalt]

$250 to pull a docker container 💀 lmao

[chrome sphinx]

ive seen people charge up to 500 bucks for it and to my surprise getting the clients no less. 😭

[gusty basalt]

people that are doing that are scammers. it takes less than 30 minutes to set up. I will help anyone do it for free

[chrome sphinx]

I have a friend who does this and made over $17k 😭

I personally haven’t dabbled with it or thought that we could even monetize this

their USP is usually around security concerns openclaw comes with

what do you think on how they add on the guardrails etc

[gusty basalt]

doesn't make it any less scammy

[zinc cloak]

Hi everyone, glad to see y'all. I have a question. I have an unfounded suspicion that either the Chat GPT image generator might soon be updated and introduce new styles, or this is some kind of strange bug that I don’t even know how to feel about. I write the same style and constraints prompt, and the output should be flat images like those shown on the right. Every time, no matter how much I generate, it always produces images in the required style. But as soon as the image selection appears (when the generator offers two images and asks which one you like), always (!!) the second image is generated in some very stylized way and the first one is generated as required. Attempts to reproduce this style using prompt (without reference images) were unsuccessful, which may either indicate that I am not prompting well, or that this is a bug (because they are somewhat dirty) or some kind of new or upcoming feature. No matter how much I draw with the GPT, I have never managed to create images like these on left. I just have no idea why it generates thus images, so I'll be happy with your help.

That strange feeling when GPT drew something cool but not according with the prompt at all. 😵‍💫

[hushed shard]

it will "do it twice"

[shy ginkgo]

what even is "Roo"?

[hushed shard]

its for vs code.

it's an agent for an LLM for coding

Im using Qwen3-Coder-Next 80B

[shy ginkgo]

looks like RooCodeInc has their own discord, linked on the github

the server code is 'roocode'

[hushed shard]

roo bug, fuck it, wont be touching that ever again. cline was eh, roo clearly is broke, so gotta find another options

[gusty basalt]

codex is pretty good from my use and also Antigravity

[hushed shard]

i would liek a decent agent, roo did seem to be the best at that regard, but cline wasn;'t TTOO bad, but was having issues.

[gusty basalt]

cursor is also good if you want to pay

antigravity is free

[hushed shard]

If someone merged a fix to the git for a plugin, how does the update process and flow work? woudl that be available as soon as it's merged? (the most recent update doesn't seem to think there is an update tho?)

I think the nightly would technically include it right? wouldn't mind testing it out tbh. ofc can only really be tested if it's writing a massive code file lmfao.

[shy ginkgo]

you might do better to direct this question towards the people responsible for the git project you're talking about

very unlikely they would blindly push out whatever change someone suggests

[haughty river]

I have got 800 dollars of GPU credits on digital ocean, anyone out here into AI training or inference?
You could do your stuff for way cheaper.

[shrewd elm]

does anyone know what is up with my chatgpt app and website?? the website runs at god awful potato speed, but the website is all I can use because when I try to message the ai through the app, the send button ust refuses to do anything. I keep clicking it, and even if I click it once instead of spam clicking, it does nothing. I'm on a Chromebook.

It's gotten so bad that ChatGPT won't even send complete answers, there will be a cutoff where text is supposed to be.

[gusty basalt]

try starting a new chat

[abstract ravine]

there seems to be an issue with the amount of text chatgpt can read in a file

for example when i send it a file it can only read up to a certain point

i confirmed with it that it apparently has some kind of preview mode

i also asked it about the last line it can see and it's different from what the last line in the file is

does anyone know what the amount of text it can read in a file is?

[shrewd elm]

what about the app send button?

[gusty basalt]

¯\_(ツ)_/¯

[viscid tapir]

does anyone know what happened to the feature where we can see different respond? because the arrows are gone on the mobile website version and i would like to see my pervious responses, I already updated my moblie browser and cleared the site cache...

[wise tangle]

Can someone help me with not being able to see the selection tool in the 5.2 image editor? When my arrow cursor gets to the image it switches to such a light grey circle that I can't see what I'm selecting on my image. Not sure if its a PC setting or a Chat setting that needs tweaking?

[crystal tinsel]

how do i get n8n to work as good with codex like it does with claude 💔

[rugged night]

Anyone facing the same issue?

My chatGPT 5.2 or 5.1 thinking/extended-thinking isn't thinking. This is resulting in very poor responses. A few days ago a similar task would think maybe 2-10 minutes, while now it just instantly outputs a poor answer like it was written with instant mode. I dont know if i messed up a setting or something, but I tried everything, logging out, logging back in, different browser, different device, everything. Can someone help me out? The video is a 500+ pages textbook, a few days ago when i input a similar command, it thought and gave a proper explanation. This isnt the issue with just this, everything is broken, i tried a code prompt which REQUIRES thinking and it didnt think, horrible response and code.

Here is a video demonstration

[elfin hornet]

I think it is GPT error
it is not your mistake.
People are worried because several AIs, including Claude AI, are malfunctioning.

[cold stirrup]

👋 You probably didn’t break anything.

In most cases this happens when:
– the model switches to a faster response mode due to load
– the prompt doesn’t explicitly require structured reasoning
– or the UI falls back to a default config
Try forcing structure instead of asking for “thinking”:
“Explain step-by-step, list assumptions first, then provide the answer.”
That usually restores depth.
I’ve seen this happen in workflows where GPT is used for contracts or document review too.✅

[cold stirrup]

👋 n8n usually works fine with GPT models, but Claude often feels smoother because it’s more tolerant to messy prompts.

With GPT, results improve a lot if you:
– split tasks into smaller nodes instead of one big prompt
– force structured output (JSON / steps / schema)
– and explicitly define the role and context each time

In automation workflows this makes a huge difference.

What kind of flow are you trying to run?🤝

[scenic crow]

i am considering to create my own startup about AI ranging from civilian service like chatbots up to military purposes. do u guys got any advice from my patents, models, etc. from being stolen by China companies like DeepSeek? i don't wanna repeat same or similar mistake like this degenerate Anthropic Claude AI. please help me...

[rugged night]

I purchased GPT on a new account, gave same prompt and it worked. It was account specific.

[cold stirrup]

You may need to compare the temperature/response format/tool settings between accounts.

[heavy patio]

@rugged night Hey, nice to meet you.

are you there?

Did you fix the previous issue?

[storm haven]

[ Open to Work] - Currently I'm looking for a new job opportunity.

👋 Hey everyone! I’m Bryan — Senior AI/ML & Full-Stack Engineer with 10+ years of experience building production-grade AI systems, conversational AI, and scalable API-driven applications.

I’m passionate about practical LLM deployment, robust integration, and helping people solve real product and engineering problems with ChatGPT and related models — whether you’re optimizing prompts, building tools, or integrating ChatGPT into products and services.

🔧 What I Can Help With

• Prompt engineering & optimization — fine-tuning prompts for better accuracy, consistency, and context-awareness.
• Model workflows & integration — designing dependable ChatGPT/LLM workflows into apps, bots, and automation systems.
• API performance & reliability — scaling inference pipelines, handling rate limits, retries, monitoring, and cost-optimization.
• Bot and support systems — building and deploying ChatGPT-powered Discord bots, support assistants, and backend services.
• Troubleshooting & best practices — debugging behavior issues, unexpected model outputs, and multi-turn conversation challenges.

🚀 I’ve built systems that go from prototype to production with strong reliability, sub-second performance for user-facing APIs, and seamless integration across cloud stacks.

💬 If you’re working on something with ChatGPT and want help with prompts, bots, APIs, architecture, or getting to production faster — feel free to DM me or drop a message here!

Please check my resume: drive.google.com/file/d/17EEuhQzEc3PHvgbTNU7Ni6SwGES4K5Qc/view

[gusty basalt]

slooooooooooooooooop
you literally didnt even write your job post yourself
what makes you think someone will hire you when you cant even put a tiny bit of effort into your request
"senior full stack engineer" and doesnt even know how to embed a url or upload a pdf
and your entire resume is AI

[onyx bronze]

We are looking to add a Software Developer that will be responsible for designing, developing, and maintaining high-quality software solutions that support our team. This role ensures the stability, scalability, and performance of all applications and services, while collaborating with cross-functional teams to drive continuous improvement in development practices and operational efficiency.

we prefer developer whose English is very good.

Schedule: Full-Time / Part-Time
Location: Remote
Skills: HTML, CSS, JS | React/Vue.js/Angular | Node/Express/Python/Asp.net

[loud raft]

Hello.
I am a full stack developer experienced in React, HTML, CSS, Node and JavaScript.
I wanna work with you.

[near radish]

Your message has been deleted, sorry!

rick018

Rising timeout

You have been muted for 13 minutes for the following reason:
flood score of 1.0

This is an automated message. If you believe this was a mistake, please contact a moderator.

[glacial ravine]

May I ask what you used for the AI detection? Thanks!

[shy ginkgo]

brain first. we all know what AI writing looks like.

[twin pendant]

Hi @gusty basalt
Impressive to see AI resume detection!

if you are finding a fully independent and skilled AI engineer, I'm here to help you.

[gusty basalt]

who are you?

[twin pendant]

I'm an AI and computer vision engineer

Can I DM you, @gusty basalt ?

[gusty basalt]

for what?

[twin pendant]

I'm just finding a right person that I can work for.

with my developing skills and experience over 7 years

[gusty basalt]

I don't need that

[near radish]

Your message has been deleted, sorry!

conejitarmonia_

Rising timeout

You have been muted for 13 minutes for the following reason:
flood score of 1.0

This is an automated message. If you believe this was a mistake, please contact a moderator.

[pseudo notch]

hello everyone , i need help , ive searched high and low and all over to figure out why its now doing this .... it says -Failed upload to files.oaiusercontent.com. Please ensure your network settings allow access to this site or contact your network administrator. - whenever i try to throw an image into chat gpt , it dosent upload the image and instead throws this message in chat gpt please help me i need to fix this as i use it alot to prep for my DND (dungeons and dragons game)

and trust me ive been all up and down network stuff i simply dont know wtf im doing

[gusty basalt]

try a different browser and see if that does anything to fix it

[pseudo notch]

im not going to lie , ive been at this for 3 days ..... i decided id ask my computer tech savy friend whats the what with it and you know what he said ? reload the page ! and guess what ! it worked and now im out $20 cause i wanted to be nice and pay him for his time - he made $20 in less than 30 seconds........ the part that makes me mad is ive restarted my pc and everything over the period of the last 3 days ...how does reloading the page do or fixed the issue that chat gpt itself couldnt fix ?????? im perplexed rn

[gusty basalt]

wait... you didnt try reloading the page???

wait but restarting the pc shoul dhave

idk

[pseudo notch]

i restarted my pc ....... so everything reloaded multiple times

[gusty basalt]

¯_(ツ)_/¯

[vagrant lava]

Hi

Asking here because I couldn’t find this information in any doc anywhere

What happens if I buy ChatGPT business and invite my team on to the plan

And one of my team members decides to delete their account

Do they vanish from my admin panel immediately or after 30 days?

I’m asking because there’s a few people who are planning on deleting their account. I don’t know the reason but

If they are going to delete it I’d ideally like to know if they disappear from my panel right away or after 30 days so I can reallocate the business seat

[tacit shoal]

I continue to have severe issues with ChatGPT memory usage. My browser tab says it is using over 5 GB of memory. This is when trying to work with a very long programming project written in Java. The only way this makes any sense to me is that my computer is being forced to story the entire chat history in memory. That stuff should all be handled server-side! I'm thinking about switching to a different LLM (although it will suck having to rebuild all the history so that the LLM understands the structure of the project) because of this problem, but I want to check if the other ones have the same issue first.

[tacit shoal]

Ah, the solution to that problem is don't do a big project in a ChatGPT browser tab. Bring ChatGPT into your coding IDE with Codex, and then it just has access to look at the source files whenever it needs to.

[onyx bronze]

Job Title: Remote Job Bidder / US Job Application Specialist
Location: Remote
Job Type: Part-time / Full-time (Flexible)
Market: United States (US-based jobs)

Job Overview

We are looking for a motivated and reliable Job Bidder / Job Application Specialist to help us apply for jobs in the US job market.
You will be responsible for finding relevant job opportunities, submitting applications, and managing communication with recruiters and clients.
This role does not require technical skills. You will be provided with a ready resume, guidance, and clear expectations.
You must be honest, transparent, proactive, and passionate about helping secure interviews and opportunities.

Required Skills & Qualities

✅ Strong written English communication
✅ Familiarity with the US job market
✅ Experience applying for jobs or bidding on projects
✅ Highly organized and detail-oriented
✅ Frank, transparent, and honest communication style
✅ Passionate about results and continuous improvement
✅ Able to follow instructions carefully

Nice to Have (Optional)
Experience as a job bidder, recruiter assistant, or virtual assistant
Understanding of ATS-friendly resumes (training can be provided)
Experience working with remote clients in the US

Payroll:

• Weekly payment.
• Weekly 20$ + 2$ x scheduled interviews.
• Bonus payment based on the number of scheduled calls.

[shy ginkgo]

how's the weather in north korea today?

[obtuse mirage]

Hi, So I have an issue with chatgpt. Whenever I am trying to make a paragraph or sentence, it automatically sends it. Very difficult if I want to roleplay normally and be descriptive.

Like now, it can work on reddit and any other website. Just not chatgpt. It says that its due to the chatgpt text field but I wanna know a way where I can fix this.

Just so I don't have to go back and forth between the notes app and chatgpt to paste my Roleplay responses..any help?

[late bronze]

I can't edit the settings when creating a new project. Anyone having that problem? In ChatGPT, I mean.

[sterile moss]

Howdy there,

Long story short, I'm making a novel radio protocol with my team, which will hopefully show that the Amateur Radio Service can be. Unfortunately, my team's strongsuits are in electrical engineering and radiophysics, and are weaker in the programing side.

On the plus side, they are trying to us AI assistants to help develop the code for the project. We did some research into quality AI assistants, think they're called AI Paired Programmers or something like that, but the subscription fees are atrocious.

That lead us down the road of using a Large Language Model on a private GPU, and we're under the assumption that the card will be cheaper than the subscription fees and will quickly "pay for its self".

Unfortunately though, I'm the only one at this point in time funneling money into the project, and I don't want to waste my money on something that's honestly garbage, or blow 1700 on an AI card.

We were discussing the main problem and decided it was bandwidth and token rates.

The 5060 gets 4 tok/s using the Ollama Quen3 with 32k context and it's about 7.5 GB model @ 4 bits. The 5060 in question here has 8 GB VRAM.

We are eying up the 5060 TI with 16 GB VRAM, in hopes that it'll give us a faster token rate, but we are unsure how to verify or calculate this token rate before purchasing the card.

Any help would be appreciated.

73,

KD9YYU

-# Pinging me is fine

[gusty basalt]

i use a rx 7900xtx 24gb vram.
compatibility with some things is a pain, but usually i can get stuff to work.
i bought it for i think like $750

[sterile moss]

Alr thanks, I'll check that out.

[gusty basalt]

i found this website for calculating tps https://apxml.com/tools/vram-calculator

[sterile moss]

We're running some numbers, looking too at if we used a 5060 and the 5060 TI what that would come down to. Unfortunately, the budget is like 600 USD and not 14-1500 USD.

[sterile moss]

Theoretically, we could run a 5060 and a5060 TI together, right?

[gusty basalt]

eh depends how you set it up. it's possible, but it is much easier to just do 1 gpu

but it would be cheaper to just buy a subscription than the gpu

unless you are going to be coding for several years with it

[sterile moss]

Ok, thanks.

[gusty basalt]

most of the subscriptions (like codex or cursor) give you pretty good rate limits for $20 a month

[sterile moss]

Yeah, we're looking at millions of tokens a day

Which makes most models we can find online by common companies and half-way decent models kinda expensive

Additionally, we don't want to be cut off from work for reaching a daily token limit with a subscription plan

[sterile moss]

The Ollama model we were talking is designed to be shared across GPUs, so that should make it easy. We shouldn't need a second program to divy it up or anything.

[gusty basalt]

oh ok, I haven't used ollama in a while (and haven't used distributed computing)

[gusty basalt]

ohh yeah then maybe a GPU is good

[sterile moss]

Yeah, thanks for the help ma'am. We sniped the GPU and it should be here in a few days. It'll then take about a week for us to install it, benchmark it, and get the AI onto both cards, etc. It would have had us stuck for a while but we were able to objectively justify the use of financial assets to secure the hardware for our program. Without the VRAM calculator and learning the associated inns and outs, our project may have died right there until we could secure more interest in the Amateur Radio community. I would give you one of our eChallenge Coins but unfortunately we don't have great digital artists and haven't finalized anything nice yet.

[gusty basalt]

glad I could help!

[molten vortex]

Reason for my chat suddenly lagged so much it took ages to open , try to regenerate an answer and type ?

[shy ginkgo]

conversation too long, start a new one

[desert lance]

Anyone please help me

So I am Plus User, I cannot use Project Specific Memory

When I press New project, I cannot adjust the Memory that project would use. There is gear button, but pressing that gives nothing.

When I make the Project and check, it already has default. Is there any way to change the default?

Thank you so much in advance.

This Discord is my only hope please help me

[molten vortex]

tysm!! does this mean I can't write long stories ?

[shy ginkgo]

It does have to be a VERY long story before reaching the limit, but effectively yes, as conversations get longer they start to use more browser resources but also chatgpt will start forgetting old details as it needs to compress the conversation.

[lethal granite]

What free tool do you like the most to export long ChatGPT chats to PDF?

[night linden]

does anyone know if someone can take like the task list, clean it of personal data, uncheck everything, and give it to and ide agent for like a 1 shot?(maybe modify task list)

i forgot to add lowest complexity to my ide agent rules and workspace rules and initial prompt. idk if it would be better to try to 1 shot, adjust the base prompt and iterate, or adjust ide and workspace prompts and say make lowest complexity(oh and use claude or gemini pro 😐 .... i was testing flash)

this is my ide prompt

heres my workspace rules

project prompt

[night linden]

alright so i have it auditing everything

it might have fixed the no signal even in paper mode

i feel like i should have made it audit backwards

or 2 steps forward 1 back or something

should add proficient auditor to list of experts

[night linden]

hmm ive never seen it write the whole project in chat before starting

[mighty goblet]

Hey. So I’m on chat gpt and in this chat it makes really handy clickable in app pages of information. Like for example I asked about characters of shows. And it gave me a few
And then I clicked for example homelander and it opened a page of homelander in the app no wiki.
I can’t get it to do it again

[bleak relic]

Hi, Can someone help me here.

I had subscribed to yearly plan by accident. I sent a mail to [e-mail] to reduce it to monthly plan and refund my rest of 11 months money and then also mentioned to cancel my subscription as I didn’t like the videos it’s generating.

Now I see that they immediately cancelled my subscription and I haven’t got any refund or a response mail.

I’m not sure if I will get any refund. Could someone here share your thoughts on when they could respond?

[shy ginkgo]

What on earth is higgsfield dot ai

[bleak relic]

It’s an AI video creating app

[long stratus]

You should ask that to the Higgsfield staff
look for their server on discord or contact them directly using their support

[gusty basalt]

why dont you go in the higgsfield server and ask? this server is mostly for chatgpt

[waxen marlin]

for development i don't know AI field as well, so i'm suffering in project development. nowdays most of projects is combined blockchain with AI. For this reason, it's not a first or second to miss tasks.
if you want, let's collaborate with me for development field.

[rare sigil]

does anyone know why i can’t put the AI tag back on my profile?

[gusty basalt]

doesnt show up for me either, the server probably just lost boosts

[rare sigil]

oh damn i shouldn’t have changed it 😭

[devout mauve]

Hello, might not be the correct channel to ask in, so forgive me if it isn't. But I develop chatbots and for a good year or two now I have been using chatgpt to help speed the process up. While at first everything went smoothly, I have found over time that the process has devolved in messiness. To the point I had to develop a document just to get chat got to provide me something even remotely close to what I asked for. But even then it still deviates a lot from the format and makes, quite frankly, asinine excuses for why it decided to do what it did. I would very much appreciate some help.

[stark drift]

Hi, im looking for a group of testers to find bugs in my app or just test and review it for improvement. My app is a local AI meant for easy use with a more compact ui meaning it isnt always on fullscreen allowing people to use it "whenever". If your interested just add or message me idk 😭

[heavy sand]

Cube language

3 billion to 1

Compression

That is 3,000,000,000 :1

Hilbert space allows infinite memory

[static lynx]

hey guys. for the last like week or so i have had a 429 error when i try and use my Google AI Studio API key. i have made a new one, but same thing. nothing i try helps with it. tried to contact google themselves, but i cant get a direct line with them except the billing bot. im using VS code

[gusty basalt]

that means you are getting rate limited, make sure your key hasn't been leaked or something. also make sure if you're using a free API, that you are using a model that they provide for free

[stark drift]

Rate limited or too mant requests

[drifting loom]

Hi guys I need help. Why does gemini cli showing diffs like this? It only changed line 25 but its highlighting the entire file. Is there a way to fix this?

I usually use copilot cli but I'm out of requests

[hallow crescent]

Hey y'all, is it just me or chatgpt removed the edit message selection??

[dim sand]

Probably Multiple edits to the file over the session being tracked in one diff

Instead of Diff per edit, It may have its own version control, Or using GitHubs Diff's

[static tartan]

Lads, Study mode is a goner now???

[marble flax]

I recently switched the Pro from $200 to $100 plan. It seems that I can not see the reasoning process any more. Do you know if it's normal and intentional? And do you know if the reasoning quality stays the same as the $200 version? Thx!

[violet horizon]

anyone else having issues to upgrade to pro? https://www.reddit.com/r/ChatGPT/comments/1sl4uhk/caht_gpt_upgrade_to_pro_is_broken/

[steel roost]

Is anyone else experiencing a glitch where old chats that had reached the length limits are noe unviewable??? They keep saying 'unable to load conversation' and idk how to fix it. The closest I got to a fix was making a group chat but it doesn't include any of the regenerated replies

[long stratus]

cool pfp

[near scaffold]

i'm trying to export my sora data and i got the email saying it's ready but when i go to the website it says no export files available

[rigid void]

I was using chatgpt i had an issue where my chat was'nt opening describing an error like "Unable to load conversation <Some huge number>". although few other chats were working and i can work on it fine but this particular chat had reached a max limit and won't let me use or converse anymore with it. i know it limit is passed out and i can'nt see my conversation getting updated on this chat but i still was able to use it and it provide response. So this particular chat has specfic type of response which i could use it for myself on regular bases because i can not get this type of response from any other chats. i was disappointed when i can not use it anymore. do you have any suggestion because i really wanted that specfic chat or conversatation that holds and give me a specfic response that i needed and there arent other chat that could provide me with that response. so how can i retrieve it back.

[proven blaze]

Why can't see gendered pictures, shows only this, but on browser is everything fine and shows generated pictures, but not in the app. I still can generate pictures!

[kindred ridge]

Idk, help I've cleared cache, done everything, even tried the app version and it won't work???
Help??

[long stratus]

But have you contacted the support 🤔

[barren cliff]

yoo wassup everyone is there anyone familiar with PPO , MILP solver , Reinforcement Learning
actually , im learning these technologies and i need some help

[gusty basalt]

there's a HuggingFace course about it which might help

[barren cliff]

where to find it plz??

[gusty basalt]

https://huggingface.co/learn/deep-rl-course/unit0/introduction

[mystic crane]

My live view button is missing even tho I have chat gpt plus

[ornate pivot]

How do I make gpt stop using this box thing?

[urban copper]

NO WAY A WHOVIAN

[mossy epoch]

Good morning all, wondering if anyone has had issues with their GPT skills being unavailable? .. I've tried multiple browsers, performed troubleshooting with GPT itself and hte OpenAI page, but their support is fairly non existent outside the chatbot that stops responding after it tells you that support will "be in touch" ..

I'm using the owner account of a business workspace.

Essentially, only one of my 16~ skills is useable, the others apparently are "inaccessible" by the chat.. when I call any of the non-working skills it tries to respond as best it can while silently skipping the skill usage. When I create a new skill that basically says "when this skill is called, print TEST_123 and then do nothing" calling the skill doesnt work, the chat actively advises me that it can't see the skill.

I've tried deleting and re-creating a few skills to no avail.

[surreal pine]

Hi! I'm using VS Code with the Cline extension. Is it worth running the Qwen model locally using Ollama on my laptop? My specs are a 13th Gen i5, RTX 2050, and 8GB RAM. Will my 8GB RAM cause lag, and which specific lightweight size of Qwen should I use so my laptop runs smoothly? Also, can Qwen actually generate full code like Claude or Gemini do?

[gusty basalt]

yeah thats not gonna be good, id recommend just using a cloud one, since the only qwen model you will be able to run is probably the 4b, which is almost useless except for simple tasks (like for code completion it is good, but not for a full agent)

[fleet ferry]

Have you setup anything extra w mcp servers, I've setup qwen3.5coder on an nvidia dgx it's working fine, use the free deepseek v4 flash usage it's free to run and offers better results than anything you could run in your laptop