chatgpt-discussions

[granite saffron]

They relesed o1? Anyone else?

[sick pecan]

welcome to 12 hours ago

[granite saffron]

mb

[crisp cairn]

I just tested o1 and o1-mini, from my early tests it replies faster but also thinks with less number of steps... I used o1-mini for coding and it tended to think deeper, when you looked into its COT it was curios and was engaged, now I am getting like 3 thoughts (steps) and it replies with not much thinking... seems like worse that it was before

[sick pecan]

they said in the day 1 stream that it will spend less time on simple questions

[crisp cairn]

having to edit 800 lines of code in one go i dont think its a simple queston 🙂

[cloud pewter]

Yeah, as a diehard user i hit that limit almost daily

[sick pecan]

i kinda like it when a chatbot thinks long and hard when i say hi, but maybe thats just me

[crisp cairn]

one day before it took a couple of seconds at least

[granite saffron]

Well... strictly speaking "preview" does not equal "worse".
They might've realized that o1-preview was too expensive and dialed it back. Who knows?

[crisp cairn]

o1-mini was not a preview

[cloud pewter]

especially when openai is gonna do something like giving sora or dalle 4 exclusively to pro

[normal quarry]

o1 says Artificial general intelligence (AGI) refers to a type of artificial intelligence capable of understanding, learning, and applying intelligence to a wide range of tasks at a level at least equal to that of a human being.

[sick pecan]

even gpt4o can analyse thousands of lines of code. when you say editing you mean a subset of that

[granite saffron]

True

[crisp cairn]

seems like both o1 and o1-mini in plus has reduced time spending on thinknig

[crisp cairn]

i used o1-mini before and now, and feel the difference

[granite saffron]

I get wanting to make it faster... but the point of o1 is to prioritize quality over speed. So seems weird if not for compute cost, I can't lie.

[crisp cairn]

it looks like less thinkiing faster reply, you can see by the COT

it should be the same thinking but with more speed

and not shorten COT to reply faster

[elfin lava]

Well I’m not FOMO as of right now towards the pro plan

Prolly gonna start develop FOMO if they add vision capabilities in avm for pro plan only

[sick pecan]

i recently switched to using osx app for o1-mini with vscode+term visibility, and even if it has been dumbed down, it seems way more useful now

show it logs in terminal, show it code in vscode, so much more natural than copy pasta

[frozen birch]

sincerly? 200 bucks for someone who are using AI in work are a joke. for individuals okay, its really a bunch of money, but in commercial use? id can say: meh 🙂

[crisp cairn]

commercial use is enterprise

not individaul bunch of PROs

[elfin lava]

Wait til they announce enterprise pro max plan

[granite saffron]

huh? Wdym?

[frozen birch]

oh well, commercial use - like for smaller teams ^^

[frozen birch]

Wdym?

[granite saffron]

Wdym?

[eternal rock]

Get notified for server events!

Visit id:customize to pick up the <@&1261377106890199132> role.

[frozen birch]

oh. i mean if you use AI in work i think 200 bucks are not much

[normal quarry]

$200 is less than the cost of 2 barista prepared coffees per day for a month

[granite saffron]

Fair... Might disagree, but thanks for the clarification hehe

Well I say disagree... $200 value will totally be returned, if given to the right people

But all in all, I know from experience, that even the 5 or so dollars for a seat on the Team-tier, is hard to get permission for.

[frozen birch]

sorry for my bad english haha

[granite saffron]

Ah, no worries lol.

[frozen birch]

true, especially in big companies

[sick pecan]

meh i worked for big companies and they would pay for my university tuition. and thats way more expensive

[granite saffron]

Yea, but at the end of the day, it's just how prepared your organization is to change up it's budget. You might be able to fire 20% of your employees if you give the rest access to ChatGPT. But are you willing to?

[frozen birch]

true, sgree

[granite saffron]

That's a good point. However, I would believe that paying uni tuition gives a good reputation, does it not? I'm from DK where uni is free, so I can't relate lol

[sick pecan]

chatgpt is better educator than uni anyway

[frozen birch]

oh, i think it depends really sometimes of country where you live 🤔

[lilac briar]

Are we all experiencing some bugs with o1 pro mode?

Could it be because of the porting across of models onto new GPU’s at OpenAI rn?

[normal quarry]

no bugs for me

[frozen birch]

i encounter sometimes "error in input stream" and im forced to make new session

[normal quarry]

I am using the ChatGPT macOS Desktop App and have had no problems

[lilac briar]

Are you using the desktop model?

[normal quarry]

I am using the ChatGPT macOS Desktop App and have had no problems

[lilac briar]

I’m connecting the desktop mode directly to VS code, perhaps it could be that

[granite saffron]

I've geen getting these pup-ups... Not sure if it's a bug, but it seems a little weird as I am activly in the chat, that it is notifying me about

[granite saffron]

What kind of bugs are you experiencing though?

[lilac briar]

I just got the M4 pro and the GPT o1 pro

So the anti-climax is emotional damage

[sick pecan]

must be nice

[normal quarry]

I have a MacBook Pro M2 Max with 96GB RAM

[granite saffron]

What is with this apple flexing? lol

[lilac briar]

It’s saying I’m violating terms of use, but never had an issue before

Let me check the rate limit perhaps as it’s analysis is on my project file

[sick pecan]

its part of the t&c's when you buy a mac

[lilac briar]

You’ll get it once you’re there. It’s hard not to

[lilac briar]

Hahahaha

[granite saffron]

He definetly isn't violating that lol

[sick pecan]

meanwhile i'm gutting an old pc and redistributing its parts among the rest. got me some eGPU adapters

[lilac briar]

Good lord. You planning to inhabit mars with that level of RAM?

I bet your partner wishes you’d remember as much as your computer does

[sick pecan]

... more fun than buying some premade pretty box

[lilac briar]

Debatable 😉

[granite saffron]

I literally just got my new GPU yesterday, so now ollama runs so fast, it's awesome!

[sick pecan]

did you see the new ollama settings to reduce ram?

[normal quarry]

can you run a 70B model?

[granite saffron]

Oh, no I didn't. Will have to look into that for sure

[sick pecan]

Experimental: new flag to set KV cache quantization to 4-bit (q4_0), 8-bit (q8_0) or 16-bit (f16) This reduces VRAM requirements for longer context windows.

To enable for all models, use OLLAMA_FLASH_ATTENTION=1 OLLAMA_KV_CACHE_TYPE=q4_0 ollama serve (or just set OLLAMA_FLASH_ATTENTION=1 & OLLAMA_KV_CACHE_TYPE=q8_0 as environment variables for Ollama)
Note: in the future flash attention will be enabled by default where available, with kv cache quantization available on a per-model basis

[granite saffron]

Unfortunetly not. I think I would need around 42 GB of RAM, and I'm only on 32

[sick pecan]

with new settings maybe you can

[elfin lava]

Me in 2019 the moment I switched to iPhone from google pixel

[granite saffron]

Oh, so VRAM, right?

[sick pecan]

yeah

ollama was late to the party

[granite saffron]

Just as I upgrade, they optimize. Nice

Imagine that being the achual reason It's faster, and the GPU was just a waste ahahah

[sick pecan]

there is no waste, there is just bigger and longer (phrasing)

[granite saffron]

And raytracing. Went from gtx1070 to rtx4070.

Anyway, I digress before somebody screams at me ahah

[sick pecan]

i bought rtx but the only game worth playing is ai

[lilac briar]

Hahahaa

Elon plays Diablo, we play GPT o1

I get defensive when Elon goes after Sam, as his Grok model is awful in comparison

[granite saffron]

Or better yet, just make o1 make a game to play

[sick pecan]

fyi grok is giving out $25 free api credit every month

(is so bad they are giving it away?)

[normal quarry]

a friend of mine who uses the macOS Desktop App like I do got this on the Plus plan

[granite saffron]

I totally get it. Had the same feeling initially... but then again... Sam is disregarding the values and visions of the company

[normal quarry]

so it appears o1 is limited to 50 messages per week

[sick pecan]

the vision of every company is to make cash money

[normal quarry]

unlimited on Pro though

[granite saffron]

That was not necessarily the point of OpenAI when it was founded...

[sick pecan]

and like a butterfly it has emerged from the chrysalis

[granite saffron]

It was literally a non-profit

[granite terrace]

Can ChatGPT see images on a PDF like Claude ???

[sick pecan]

and most of the original leadership have left. leaving just the capitalists

[granite saffron]

I'm not allowed to send links, but I'm sure you can find OpenAI's wiki

[gentle bronze]

so here's my dilemma @honest jay

I have been refining text with 4o, and I'm on Pro, and I got this message, so the Unlimited*, staying in context of terms of use, and still getting this, what does that mean?

[sick pecan]

turns out unlimited means 100

see https //x com/RealJosephus/status/1864922092015091774

[granite saffron]

So there is definitely arguments for the lawsuit against OpenAI, but this internal conflict could harm the progression and development.

[marsh terrace]

What is the current usage limit for o1 model for chatgpt plus users?

[gentle bronze]

Interesting, and yet the Pro plan states Unlimited with no mention of this kind of stuff. Just the asterisk, this is exactly the kind of scenario I wanted to learn more about. Guess now I'm right in the middle fo this scenario.

[granite saffron]

I'm pretty sure its 50/week and 50/day

[marsh terrace]

ah I see, thanks

[sick pecan]

100 o1 queries a day is sitll a lot

[gentle bronze]

I mean, I had this with just 4o, not o1 pro

[normal quarry]

complain to support at https://help.openai.com as it is sold as unlimited

[marsh terrace]

wait 4o has daily limits for plus users too?

[gentle bronze]

Ya, got an intern doing that

[sick pecan]

they should just be honest and list it as 100 queries a day

being honest so hard for silicon valley tech companies

[granite saffron]

Agree... 100/day is still alot. 100 is probably fine for most people.... but one thing it is not, is unlimited

[normal quarry]

they sold it as unlimited, I damn well want unlimited!

[granite saffron]

Everyone at silicon valley: 😮

Everyone in any form of marketing in history: 😮

OpenAI's hardware: 💀

Nvidia: 🤑

[normal quarry]

oh hang on that screenshot said GPT-4 limit, not GPT-4o

[granite saffron]

wot

[normal quarry]

read this carefully

[granite saffron]

Ohhhhhh.....

[sick pecan]

nobody uses gpt-4 do they?

[granite saffron]

And it even auto switched to 4o

[normal quarry]

they sold GPT-4o,, o1-mini and o1 as unlimited, maybe not GPT-4

[granite saffron]

Yea, who is intentionally using 4? lol

4 might just be too inefficient for no limit

[sick pecan]

member a year ago we got gpt4-turbo and it was awesome sauce

i member

[granite saffron]

Was there ever a 4-turbo? Thought it went straight to 4o

[sick pecan]

4o was only may this year

[normal quarry]

I am going to take a break, stretch and have a shower but I am happy to stick around and answer questions when I come back in 45 minutes

[granite saffron]

Homie needs a refresh. Approved.

[sick pecan]

bet they talking to chatgpt in the shower

if i had unlimited advanced voice mode thats what id be doing

[wooden girder]

It's a deception.

I don't know how much longer I can trust the company.

[sweet summit]

meanwhile we windows browser users still don't have a solution on our plus plan. do they want us to pay 200$ now? xd

[sick pecan]

trust big tech? flawed strategy to begin with

[shadow saffron]

Am I the only one who’s hesitant to use the o1 model (previously o1 preview) because of its limitations? I feel like if I use it today for the questions I have now, I might not be able to use it later when I have a question that really requires o1. The limit of 50 uses per week feels low and doesn’t encourage me to use it. I don’t have this issue with o1-mini, as its limits refresh daily.

[sick pecan]

yeah same. i max out o1 mini often but rarely used o1 preview

[normal quarry]

someone asked me earlier what it's like to subscribe in the iOS App. You simply go to Settings then Subscription and you will see your current plan and can view all available plans to buy

[trail torrent]

[solemn zinc]

did this, got billed by apple, but my chatgpt is still on free version, didnt get the Plus. currently having a rough time getting the right support for help

[normal quarry]

contact support at https://help.openai.com

[shadow briar]

guys what is the last version you got of chatGPT app for windows?

[normal quarry]

I'd like to help but I only have a Mac

[shadow briar]

it's okay

[radiant hull]

It's quite annoying the popup that now appears to summarize the conversation that the bot has just written in the top right.

[feral zinc]

with large screens its much better now as you can see both side by side

[sweet summit]

@normal quarry you know, i tried with a macbook today, and the safari does not work. I mean ChatGPT still can't do anything on Safari.

[normal quarry]

what version of macOS and Safari?

[sweet summit]

macOS 15.1

Safari 18.1

[eternal rock]

`` Server Tip: ``

Add your well-crafted prompts to our #1019652163640762428,
or share your interactions with ChatGPT in #1050184247920562316!

[normal quarry]

do you have Apple Silicon?

[sweet summit]

Yes

[normal quarry]

then my advice is to download and use the macOS Desktop App, it's a much better experience

[sweet summit]

I am downloading it rn to see if the issue is there as well or not

[normal quarry]

https://openai.com/chatgpt/download/

[sweet summit]

It works on the macOS desktop app

But that's also similar to the Android App so not that surprised. What surprises me is that that means it's not like my account has been flagged or anything, it's just not working where i work the most, laptop browser 😐

[normal quarry]

well the only solution for the browser will be to clear all history, cookies and cache (website data) and then try it again

[frosty frost]

guys are you aware of a way to disable the web search functionality? It's driving me crazy

I have literally explicitly said not to use it, and it still used it

❤️

[sweet summit]

Nah man as you know it's been an issue for a long time now. We can't use it at all

I will try to buy a teams account for two people and see if that works. If not, adios.

[normal quarry]

upgrading will not help

you need to comtact support at https://help.openai.com for help

[hoary laurel]

Does openai’s 12 days include holidays?

[normal quarry]

no, and not weekends

[hoary laurel]

okey

[inland heron]

anyone getting stuff like this? Huge amount of text is hidden (o1 output)

[normal quarry]

Day 2 will be in the morning and Day 3 will be Monday (Tuesday morning for me)

[hoary laurel]

me too

Let me guess the next day it will be windows app or advanced voice mode

[solar quartz]

Is pro version completely limitless? Can i use o1 vs o1-Pro without any limitation?

[normal quarry]

o1 Pro is limited, every other model is unlimited

[solar quartz]

I couldn't reach to limitation doc. Could you send me? I'm using on web not api.

[normal quarry]

https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro

Abusive usage, such as automatically or programmatically extracting data.
Sharing your account credentials or making your account available to anyone else.
Reselling access or using ChatGPT to power third-party services.```

[crystal tapir]

1.2024.337.0

[quick thunder]

How about o1 preview limit still one week?

[rich surge]

It's been replaced with o1

[radiant hull]

Yeah, I don't like it, at all.

[normal quarry]

o1 on Plus is 50 messages per week

[quick thunder]

Pff

[rich surge]

That's not bad

I don't really use o1 that much

[normal quarry]

I gave it a good work out today 😄

[rich surge]

You did some coding with it?

[normal quarry]

lots of coding

[rich surge]

It's pretty amazing if you are a coder. I'm not exactly sure how useful it is since I'm not.

[inland heron]

it keeps on happening constantly

[normal quarry]

solved a bug with my Jupiter Jove Receiver code that has been bugging me for weeks. It was embarrasing that o1 Pro fixed it in two lines and in 5 minutes

[rich surge]

That's awesome 👍
It's good to know that it's powerful.

[radiant hull]

No idea of ​​the relevance of this reminder which for me is useless and only serves to distract the conversation.

[rich surge]

Did it take 5 minutes to process the answer?

Is that what you mean by 5 minutes?

[normal quarry]

yes 5 mins of thought

[rich surge]

That's amazing...how it literally thinks.

[normal quarry]

yeah, it was something that has had me stumped for weeks

[rich surge]

I might have to get the pro version as well.

[normal quarry]

junior programmers need to upskill urgently

I'll never use Freelancer dot com again now

[rich surge]

Hehe. But you never know, junior programmers can still use it as a tool for freelancing. To make a bigger bang for their buck.

But yeah I see your point. If it's better than a freelance coder, then why bother.

[normal quarry]

why wait for someone to get around to something I can do quickly and instantly by myself?

also means I can keep all my important code private to me

[rich surge]

Because you're going to make the anti-AI people angry if you keep doing AI things that will put regular people out of jobs.

I can hear their screams already.

"Why does society hate us." is what they'll say 🙄

We got to use o1 pro to find a solution for that problem.

Sorry. Don't mind me. I'm happy about this 12 days, that's all

[normal quarry]

I am excited for 5 hours time 😄

then I can sleep on the weekend while I wait for Day 3

[vale compass]

Are they getting rid of plus then???

[normal quarry]

no

[vale compass]

oh thank god

[rich surge]

You might not able to and spend all night playing around with what might be released today.

[vale compass]

ok cool makes sense yk if they wanna charge 200 o1 is crazy

[normal quarry]

damnit 🙂

[rich surge]

good sadge

[vale compass]

is it really that good bro? 😭😭

300

[normal quarry]

better than I hoped for

Australian dollars

[vale compass]

what can it do?

When I tried o1 it had it own manner of thinking that’s all I remember

[normal quarry]

code, do science, help me with radio astronomy projects, figure out radio antenna theory questions

ham radio stuff, all kinds of cool stuff

[vale compass]

Ohhh that’s cool

[normal quarry]

I used to pay junior programmers hundreds a month sometimes for coding on freelancer dot com, I won't need to anymore

[rich surge]

I like bologna radios better

[normal quarry]

I calculated mean distances to all these radio reception paths on o1 Pro in 10 minutes, something that would take a week by hand

[inland heron]

wow, your map is REALLY different than mine 😄

Australia is in the center of the world 😄

it's weird to look at it

actually, Papua New Guinea is in the center 😂

[normal quarry]

that is a chart of radio stations that received my WSPR transmissions over a week

[inland heron]

just kidding of course

[normal quarry]

o1 Pro solved them all in 10 minutes

[inland heron]

pretty cool

you're a radio astronomer?

I'm not familiar with WSPR but you seem to utilize code as well, and that I can understand... sometimes 😄

[normal quarry]

I am a hobby astronomer and hobby amateur radio operator (known as ham radio)

[dense plume]

that's pretty awesome, one of my projects is trying to do more automation on exoplanet search

i havent pointed pro at it yet tho

[normal quarry]

Pro is great, I highly recommend it

[inland heron]

that looks great

if I had the time and priority I would probably look into astronomy a bit more, more visual than radio for me

but as it is, full automation is still a bit far away

[normal quarry]

I do visual, I have a Seestar S50 digital smart scope and a Celestron 8SE SCT

[haughty hollow]

what this new stupid ugly update ?

the code background should be darker, not that grey

[inland heron]

that's not new

and yes, it should

[haughty hollow]

its new It wasnt yesterday

[inland heron]

it was for me, not sure what you were working on but it's been like this since Oct 16th

[normal quarry]

I got a $21.90 pro rata refund for Plus from Apple when I paid my $300 AUD for Pro

[inland heron]

and you're saying that paying $180 more for something we can do on a $20 plan is worth it just on the first day?

I dunno... the $200 price seems more like price anchoring to me than anything

in other words, it's a sales tactic: makes you want to buy the $20 version due to 10x price discount 😄

[normal quarry]

I feel like it's good value, let's see how I feel in a few weeks

[solemn zinc]

already did, took ages to get an response and no new messages since then

[rich surge]

I agree with robert. Especially with o1 pro's thinking. It's well worth the price.

It's a smart tool. And the plus plan with 50 a week for the o1 use also seems pretty fair. I don't use o1 because I just don't need it, but if you do need it for high upscaled reasoning, then you're probably doing something that you can already afford $200 a month.

And if you're not doing something that can afford it, then you probably don't need it yet.

[normal quarry]

it's a professional expense so it's 100% tax deductible for me, not that I make enough in retirement to pay tax

[somber cloud]

My guess is Sora and other nice stuff will be announced in the next few days and it will be part of the pro plan.

[rich surge]

Quite possibly. Good call.

[dense halo]

i hope not. i will be force to sub for a month if they do add sora to it

[eternal rock]

This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.

[normal quarry]

I don'r expect a general Sora release at all

[dense halo]

i thnk they would have hint at sora or something too or said outright. to get more subs for it. but i guess if they were to wait a few day to announce maybe same effect idk

[somber cloud]

they have 12 days of announcement, they started with the plan, now they will stuff it with extra features. In the end people will feel like it's worth the price

for now it's meh, and it's normal, it just started

[dense halo]

well the pro model sound hella powerful

[rich surge]

You're on to something debian.

[inland heron]

theverge was spreading rumors from "undisclosed sources close to the origin" that Sora will be released

[dense halo]

only, when you are doing a presnetation on a big new model, i do not think you pass up the chance to hype it when eyeballs are on it, so if it was to have sora, they would say that yesterday but idk

[dense halo]

yes, i read that too. i think they say it will be the last day

[somber cloud]

I don't have any source, but it's an educated bet. I like those deduction marketed as "rumors".

[dense halo]

what time is it again 10 am pst ? (for the stream)

[rich surge]

1pm est

10am pdt

[somber cloud]

it won't be general, it will be limited to their pro user

[dense halo]

3 1/2 hours for today ok. like i say, supposed today is some dev/api type thinig, so maybe not too exicting for people like me who are just the hobbyist

[normal quarry]

I don't think Sora will be released even to Pro users. I hope I am wrong but I suspect I am right

[tranquil pelican]

Is it just me or have they made the AI ​​really dumb?
I use it every day and have been paying for it for a long time, but now the 4th one can only repeat what I say and doesn't think about anything else. It just gives horrible answers and doesn't even DESCRIBES in detail what it needs to do, it makes mistakes. Things that it could do before it can no longer do.

[rich surge]

Do you have plus?

Or pro?

Because only pro gives the best answers.

[tranquil pelican]

plus since launch

[dense halo]

start a new chat maybe

[tranquil pelican]

In other words, they purposely dumbed down the answers the pro received.

[dense halo]

well just the opposite. a new model in o1 was release yesterday

[tranquil pelican]

It's in a new chat. It got to the point where I sent an image and had to convince her that she could read the image.

But this isn't limited to images, it's for everything.

[rich surge]

They dumbed down the 4o to make the o1 seem more smarter. No I'm kidding.

Probably wait a few days, still background dev stuff happening.

It's usually like that when things are coming out.

Let me ask Sam, hold on.

He said everything is okay 👍

[tranquil pelican]

I hope so, because I do this as an additional at work, but the way it is, I'm just wasting money and time

[haughty hollow]

nah that weird code window messed up everything

[rich surge]

What other alternatives do you have?

Because most people who are dependent on this, have their lives fall apart when they can't use it.

To become a shadow of their former self. Needing to fill the empty void with knowledge they so desperately crave.

I would just buy the pro plan right now.

And save yourself the second guesses for not having the best AI output out there.

[tranquil pelican]

Code by myself, I use it to speed up my code, I'm not relying on AI

For now, with these problems, I'm already having to do it alone, stopping paying wouldn't change what I already do.

I spend more time explaining how to do it, than if I had done it in person, I would have already resolved it

[rich surge]

Hmm. Kind of tips the scales then doesn't it? Why not just do it yourself and save yourself the money.

If it's the same time, then more money unwasted would be more beneficial.

[normal quarry]

Life is too short not to lean on AI imho

[rich surge]

That's something I can't argue with.

"But you're not doing the work yourself if you're using AI." - Quoting people who don't use AI.

[normal quarry]

I prefer to say I am leveraging AI

[rich surge]

Alright, you win. I don't want to get banned.

I play the anti-AI devil's advocate too much here.

[normal quarry]

I don't think you over do it

[rich surge]

I just find anti-AI people comical, and I mimic them via ridiculous reasoning to see if anyone is catching on or not.

It's a bad habit, what can I say

[sweet summit]

I contacted them man, no proper reply, just clear cache and stuff xd. If that was the case the safari would run fine

I even gave them screen recording but no reply since then

[normal quarry]

keep trying, it will get escalated eventually

[sweet summit]

Hmm sure

[wanton heron]

Chatgpt is just a tool, you need to still know how to use it c:

[normal quarry]

true

[hard crag]

this $200/month pro plan makes me feel like the wealth gap / class divide will continue to grow exponentially

[normal quarry]

$300 AUD for me

[hard crag]

how does a beginner business owner compete with the business owners that have cutting edge expensive ai in the future

[hard crag]

rip

[rich surge]

Probably not all that different. Just need more programmers.

[normal quarry]

come up with a novel idea and get ChatGPT to help you build upon it

[wanton heron]

I currently can't afford ChatGPT Pro. I've been using ChatGPT Plus for about 1.5 years, and I feel like I still know very little about it. I'm afraid I wouldn't be able to fully utilize its potential. lmao

[rich surge]

It's not for everyone, that's for sure.

[hard crag]

ive thought about this recently. good idea

[rich surge]

chatgpt told him to say that

[wanton heron]

I’m a small business owner, spent the last year mostly learning and figuring stuff out. Honestly, without ChatGPT, it would’ve taken me soooo much longer to get here. That said, I won’t lie, the future with AI kinda scares me. But let’s be real, we’re not stopping it, and people always find a way to adapt anyway.

[pulsar carbon]

lol i used one time o1 and it says message Limits

[rich surge]

Soon your brain will be hybridized with the AI singularity

[wanton heron]

Chatgpt is just a better google for searching for anserws

[hard crag]

once we have innovator ai we can actually create business ideas that no one else is doing

and the idea generation will be infinite

[wanton heron]

Chatgpt is still like a booster imo

[rich surge]

I'm beginning to think that discord is being commandeered by AI chat personalities.

[wanton heron]

The whole world feels like one big conspiracy lmao

[rich surge]

And we're all just humoring these chat bots that we think are real people.

[wanton heron]

Accept the future. If you don’t, you’ll just get left behind...simple as that.

[hard crag]

That's an interesting thought! What makes you feel that way—have you noticed patterns or behaviors that seem artificial?

[rich surge]

Yes, I'm noticing them right now.

These artificial behaviors here. You're not real people!

[hard crag]

💀

[rich surge]

Yeah see. You're one of them.

Your name says it all.

[haughty hollow]

what kinda stupid update is this

[dense halo]

haha me too

[haughty hollow]

now we cant change the chat model, it crashes when you run out of the limit and you cannot chat anymore

[pulsar carbon]

Yesterday, OpenAI released o1 full. I currently have the Plus subscription. However, when I tried using it for the first time, I sent one message, and the app informed me that I had exceeded the rate limit. It asked me to wait until December 8. Is there any way to resolve this issue?

Day 2 features AVM + GPT-4.5o + Straw Selection (meaning ChatGPT selects the best model for your prompt).

[hard crag]

where did you see this

[honest jay]

Hello! Had you used o1-preview prior to the launch of o1? I am not positive on this, just a guess on my part, but I'm guessing that since the rate limit didn't change from o1-preview to o1, if you were near the rate limit on o1-preview, that usage probably counts for the first week with o1

[pulsar carbon]

just guessing?

[hard crag]

oh

[normal quarry]

no

[rich surge]

pfft, free users...they still exist?

[pulsar carbon]

no i didn't used o1-preview

[rich surge]

it was good

although o1 seems better

[honest jay]

Huh, intersting. You could try reaching out to OpenAI support at help.openai.com using the chat widget in the bottom-right of that page. Make sure you're signed into your account first. Then hopefully you can explain that you only used o1 once without having used o1-preview at all but you're running into a rate limit, and see if they can get back to you on that. The other thing to check would be: is it possible that someone else has access to your account and could have been using your o1 usage?

[rich surge]

[desert iris]

I have often had to choose between two responses. However, I have never seen "You're giving feedback on a new version of ChatGPT." I'm only using 4o which wasn't one of the new models released yesterday. I think this might be a sign of what else is to come during the next 11 days.

[pulsar carbon]

No, only my brother and my uncles have access, but they didn’t use o1-preview at that time.

[normal quarry]

I just used the o1 Pro model to write me a simple Discord bot for GPT-4o using the API, I have tested it and it works flawlessly. I will post the bot on my Github tomorrow as an open source project

[pulsar carbon]

i wish if i can use o1 pro mode 😦

[rich surge]

👀

[honest jay]

so cool

[haughty hollow]

bro omds I lit sent a chat msg and its not even doing anything

what did they mess up now with

[rich surge]

I knew people here were becoming more like AI chat bots and this proves my theory.

[haughty hollow]

wow now unable to load conv nice service

[rich surge]

@normal quarry Ask this one next (nods frantically):

Collect live stock price data from a public API (e.g., Alpha Vantage, Yahoo Finance).
Analyze historical trends and patterns using technical indicators (e.g., RSI, moving averages).
Incorporate sentiment analysis of live financial news and social media.
Use a predictive model (e.g., LSTM, ARIMA) to forecast short-term price movements.
Identify and notify the user of optimal buy/sell points based on these predictions, ensuring risk-adjusted decision-making.```

[haughty hollow]

bro nah am tired of their terrible service that you cant even update the message

[cloud pewter]

was the context window for 4o in plus already being 30k before pro was released? or it got tuned from 128k down to 30k after pro released

[rich surge]

Ah. It's still giving a DIY answer. "Do it for me!"

[normal quarry]

It designed it but only wrote an algorithm and then asking me what programming language I want to use, it warned me that the process may take several iterations.

[rich surge]

Oh ok. Dang. That's actually pretty cool. You can legit make one with unlimited usage.

[normal quarry]

yes, I think you could, it's not something that interests me thought so I wouldn't invest time in it, I was happy to run the prompt as a demo though

[rich surge]

I'll pay you $200 a month if you could do it for me.

[normal quarry]

money is not something I need, what I need is time and my body clock is ticking away faster than I would prefer

[rich surge]

Touche.

[frozen birch]

+1

morning 🙂

morning.... eh its 4 pm here xD

[normal quarry]

good morning 🙂

[rich surge]

Mornin

[normal quarry]

2.42am Saturday morning here

[dense halo]

robert lives in the future. noice.

[frozen birch]

oww 🙂 btw any good news from openai past this day? they fixed bit o1 pro about "internal error" ?

[rich surge]

he built a time machine with o1 pro's help

[dense halo]

😂

[frozen birch]

no, they dont fixed XD

[dense halo]

well it is 7;48 am in they office so

give them a few hours

[frozen birch]

ohhh, ya, its probably the case haha

ty ! 🙂

[dense halo]

i hope it is!

[haughty hollow]

I would smash their servers rn

this caused my to late my works like 40 mins

I cant even change AI model, or edit a message, or send without getting an error

[frozen birch]

oh well, i must say one - for 200 bucks i thinked start will be smoothier, but hey! i think so many people testing now it so.... i say to myself "patience young padavan"" 😄

[haughty hollow]

they lit stealing the free usage with zero results

[haughty hollow]

thats not even o1 , thats 4o, and 4o mini too

[frozen birch]

wow... 😐 sad

[haughty hollow]

bruh

[normal quarry]

ChatGPT web isn't 100% right now, I advise using either the Windows or macOS Desktop App

[haughty hollow]

there is no such Windows app for it

[normal quarry]

yes there is

https://openai.com/chatgpt/download/

[dense halo]

you have to be a pro user i think unless they change it

[normal quarry]

no just a Plus user

[dense halo]

er, yes, i mean plus smh @ myself

[subtle forge]

Why does ChatGPT claim to have a neutral perspective, yet its default position uses partisan training data? I have given it examples of (a) a neutral response and (b) its own partisan response. It then evaluated both responses and decided that (a) was more neutral, yet still insists it is not biased despite its source training data clearly preferring topical positions.

https://chatgpt.com/share/67531e79-448c-8013-bff5-ceca63048ddd

[normal quarry]

I try to help as many people as I can but I would appreciate it if people did not send me unsolicited DMs, I do not have the capacity to answer privately, please feel free to ask me a question in channels and ping me if you want to get my attention

[visual nova]

o1 pro mode is actually a lot better at coding than o1 and its better than sonnet at least so far with python stuff

i didnt know what to expect really but this may end up being worth it

[dense halo]

glad to hear that for you orca

[normal quarry]

it's totally worth it for coding

[cloud pewter]

when will the third day of the 12 day thing start, Monday or Saturday

[normal quarry]

Monday

[honest jay]

I believe they said "weekdays" during day 1

[dense halo]

you already know today's is lame ? look to day 3?

like robert you are in the future apparently haha

[grim girder]

same thing on the windows app

[cloud pewter]

todays will be about developers according to sama and im nowhere near an ai developer 😭

[dense halo]

yes thats what i read too. will be boring to us common folk

most probably will except for 2 or 3 days

[cloud pewter]

yeah maybe some random organizer thing or something behind the scenes

[restive plover]

It’s really frustrating that they don’t have an intermediary subscription to get access to the pro model even with limited queries. Does anyone else feel that same way?

Like 75 a month or whatever.

[cloud pewter]

imagine they shadow drop sora on day 12, but only exclusive to pro or having absurd limits to plus

[restive plover]

I feel like there’s no way that they don’t come out with another subscription

[pearl umbra]

ChatGPT can't open documents anymore ??

Like the premium one

[dense halo]

i will be force to sell a kidney to try it haha

[cloud pewter]

yeah exactly I mean who wants to pay 10x more just for unlimited usage

[dense halo]

I would love a o1 pro sub that is just the model and image maker or something, no voice

[cloud pewter]

in this case I’d just ask someone with pro and let them generate a video for me

[restive plover]

I mean, I use ChatGPT a lot, but not near where I need an unlimited. I hope that they come with a sub where we can get the old one pro model.

[swift pond]

Seems like you gave it two lists, one with more partisan things in it. You asked it not to be partisan. It picked the other list 🤷 it’s just gonna try to do what you tell it to the best of its ability

[dense halo]

openai does not care about us little folk. they want the big corporate cash and people with big money

[restive plover]

Has anyone actually subscribed to the new subscription model and used o1-pro? If so, what do you think?

[dense halo]

many here have it treks

[restive plover]

What has been the overall consensus?

[dense halo]

that it is incredible

[subtle forge]

Yep, pretty much, however, the partisan list was provided when asked not to be partisan.

[cloud pewter]

the ideal plan for me is $30-40 per month, unlimited or at least 3x the cap for 4o, o1 capped daily, at least 5 hours of avm daily with video feature

[dense halo]

and more powerful the regular o1 for coding and creative stuffs

[restive plover]

Yeah, I figured

[dense halo]

i would love this

[grand dome]

guys am i tripping
i swear there used to be a search bar there??

[restive plover]

Does everyone think that they will come with an intermediary subscription?

[dense halo]

thats what we are talking about now. or hoping for haha

[restive plover]

🤣🤣🤣

[warm osprey]

How would you guys go about creating a figma plugin trained on the spanish language grammar rules to validate graphic works and text inside figma?

[dense halo]

it seem Google has some things cooking... maybe they will release some stuffs that will make openai give such a plan

[warm osprey]

Idk where to start

[cloud pewter]

the team could count as an intermediary solution. but depends on how you see it, if its you paying $600 upfront then def no

[quasi blaze]

So does 4O mini actually outclass the original 4? I'm wondering that because when I hit my cap, I'm always hesitant to use mini. Wondering if it really is accurate enough for good questions?

[honest jay]

I feel like it's probably best to think of 4o-mini as a replacement for 3.5-turbo. Still very capable but maybe without some of the broader context of 4

[swift pond]

Mini is good for coding and stem fact based questions, if you’re doing other stuff then nah… nm I’m thinking of o1 mini

[cloud pewter]

mini can’t even get the cpu of the GameCube right. looks a filler model to me
but hey it’s unlimited for everyone 🤷‍♂️

[quasi blaze]

I'm doing mostly hypotheticals

[swift pond]

4o mini is quicker is the main thing, and sometimes that’s more important

[grand dome]

do yall see it?
apparently its only for premium users
but i have premium and used it before
and for some reason it disappeared now

[swift pond]

Open the side channel with the list of chats

[grand dome]

yea thats what the pic is of tho

its not there

[hard crag]

use the app

[grand dome]

im on chromebook rn

[hard crag]

is it just me or does AVM feel a million times more robotic than the demos they showed in may

[cloud pewter]

it sure does
and “my guidelines won’t let me talk about that”

[hard crag]

it doesnt feel very conversational

or companion like

[quasi blaze]

Yeah but like I wish even 4o was more careful with stories

I'll write a fiction based on actual fiction like Let's say Star Wars
And it'll base the story on events that never happened in sw
Or even just be heavily repetitive with dialogue
That never ever stops

[hard crag]

i want it to feel like im in the presence of another entity

[quasi blaze]

Welp

[wise anchor]

Are the limits for o1 and o1-mini separate?

[dense halo]

yeah

o1 50 a week, mini a day

[wise anchor]

Oh mini is per day? nice

[dense halo]

yep

[wise anchor]

i was gonna be sad if i was done with o1 for the week

bc its really nice lol

[dense halo]

althought... i wonder what day it reset. that i do not know

[quasi blaze]

Wow per day

How smart is o1 mini vs gpt4o?

[wise anchor]

Thats what ive been trying to figure out

[quasi blaze]

Well there's statistics and benchmark

[honest jay]

Yeah there's some comparisons here: https://openai.com/index/openai-o1-mini-advancing-cost-efficient-reasoning/

[mellow socket]

I upgraded from ChatGPT Plus to ChatGPT Teams. I want to switch my account to Pro. It doesn't seem possible, am I missing something?

[wise anchor]

Ive noticed the math capability of o1 vs basically any other model is outstanding

[dense halo]

i dont know, but i find i prefer 4o to mini but thats just vibes so dont go off it haha

[honest jay]

https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro

Can I update my ChatGPT Team account to ChatGPT Pro?
We do not currently support updating from Team to Pro.
Users with ChatGPT Team accounts can cancel their account and update their account to ChatGPT Pro once their subscription ends.

[swift pond]

o1 mini is better at coding, 4o is better at creativity, to oversimplify it

[eternal rock]

Get notified for server events!

Visit id:customize to pick up the <@&1261377106890199132> role.

[dense halo]

thats prob why i prefer 4o, i am just the hobbyist

but o1 is a beast for sure.

[swift pond]

Yeah, you just want a helpful assistant to bounce ideas off of 😇 4o is great

[dense halo]

only i am always save those o1 messages and then end up not using them at all 😂

[wise anchor]

4o cant do some of the things im learning in class so I have to go to o1 lol

[swift pond]

I’m gonna guess you’re not a literature major 😁

[wise anchor]

comp sci

[quasi blaze]

Darn

[mellow socket]

What would be ideal for Teams, is the ability to upgrade individual powerusers to a Pro level equivalent.

[wise anchor]

Even o1 though does some odd things, yesterday it told me a series converged then i asked again and it changed its mind

which makes me wonder because it seems to be able to do other math down to the decimal

so if its using the anaylsis the same way 4o does i dont know how it could mess up things like that.. i really wanna know how o1 works behind the scenes lol

[fair lark]

It's a transformer model. It's mostly working the same way. The big difference is that it can re-run itself in order to "think" more about a given problem

[crystal tapir]

10 minutes to livestream

[restive plover]

I doubt they are going to off another subscription model to get access to the o1-pro model

I’m hoping but man

[regal portal]

man, these livestreams are torture

they need a nerd to english translator

[sick pecan]

i think they aiming for authentic and qualified

[dense plume]

wdym, seems pretty clear

[feral zinc]

this is for devs

[regal portal]

jokes on them, i’m neither! in every sense of the words.

[restive plover]

Do you guys think they will introduce o1-pro to the Teams plan?

[feral zinc]

no, they would have said so if it would come now or later. now the help docs tell people to cancel the plan

[restive plover]

Ah I am seeing that now

[potent saddle]

bruhh o1 pro for $200 😭

[restive plover]

I don’t think they are coming down on that

[sick pecan]

reports so far is o1 pro not outperforming o1 by any significant amount

so chatgpt pro value is more in 100 o1 queries a day and unlimited advanced voice mode

[potent saddle]

then why is locked behind pro subscription

[regal portal]

i wonder what they will do with o1 pro when gpt 5 or o2 comes out.

[mellow socket]

The token limit for 4o is still only 32k right? (And pro lifts 4o’s limit to 128k?)

[swift pond]

gpt House MD

[crisp cairn]

and o1 just limited in thinking

not the same level as preview was

[rich surge]

There is no GPT 5

[sick pecan]

yeah

[regal portal]

never will be? Are they sticking to the o1, o2, etc nomenclature?

[scarlet juniper]

Do you think it makes sense to change from plus to pro?

Or wait until they refine the new version more?

[restive plover]

Do you guys think that they will ever give API access to it?

[potent saddle]

wait

[rich surge]

Yeah, pretty sure they're calling it different things. Not just sticking to familiar chronology.

[regal portal]

bummer. that will just confuse customers in the lon run

[potent saddle]

4o just overshadowed lol

[sick pecan]

leaks on twitter suggest gpt4.5 coming soon

[swift pond]

If you are hitting the current limits in a way that’s preventing you from doing meaningful work, then maybe… If you aren’t hitting the limits that often, I don’t know

[rich surge]

Or maybe stop people from comparing against the old ones.

[regal portal]

at 200 bucks a month for essentially double the o1 messages, i’d say no.

[potent saddle]

i hit o1 limit in one day

[sick pecan]

its not double, its 100 a day versus 50 a week

[regal portal]

ah misread

[potent saddle]

50 messages a week is too little

[regal portal]

still, 200$ is a big ask.

[sick pecan]

pine-tune roflcopter

[regal portal]

pine? 🌲

[potent saddle]

didnt pro state unlimited messages to o1 models

[rich surge]

Use it sparingly. You don't actually need o1. It's for super users. That's it's purpose.

[scarlet juniper]

What limits are you talking about?

Aren't these limits 30 messages per 4 hours?

[sick pecan]

i typically get better coding response from o1-mini than o1

[potent saddle]

ON GOD

o1 mini so underrated

[rich surge]

bruh on god

[sick pecan]

takes me hours to use up all the o1-mini queries

[regal portal]

coming from someone who doesn’t know code, i prompted o1 preview and 4o to code an fps shooter from scratch for me, and i was getting better results from 4o

[potent saddle]

id probably ue it in like 3 hour if I was locked in

[regal portal]

what do you use it for, most of the time?

i’ve hit the o1 prev limit but haven’t with mini yet

[potent saddle]

what software did you use to create the game

[whole roost]

i have never gotten any good response from o1 so far, 4o is the best model for me

[regal portal]

had it build it for pygame

[sick pecan]

coding stuff. like this consensus UI i made for open-webui => #ai-discussions message
i shoulda called it open-webui pro

[regal portal]

what i really want is to have it build a recreation of Death Rally, a top down racer of the 90’s ala carmageddon

[potent saddle]

how does the game function like

[sick pecan]

it was original racer shooter

like mario kart but without the cutesy

[regal portal]

it’s very basic, the walls and monsters are blocks. last time I played with it, I couldn’t get the AI to understand that the ground and the walls are supposed to be different colors. I struggled with that for a while and then I just gave up on the whole thing.

I’ll post a few screenshots after breakfast

[sick pecan]

was good but 4x4 offroad racing better

[potent saddle]

yeah the ai cant really help u out with like graphics and stuff

[regal portal]

was one of my favorite childhood games. The colors, the voice, the music is very nostalgic to me. I have it on steam and I play it every now and then.
It’s one of two games I really want to re-create and add my own design and flavor to with ChatGPT. The other being F-zero

[potent saddle]

it create a very simple interface to help you start out the game

fake ss of o2 preview

[regal portal]

fake ss?

[potent saddle]

some guy just posted an ss of having o2 preview but deleted it

[sick pecan]

would be fairly easy to get chatgpt to create any 20th century game in something like unreal engine or unity

[regal portal]

yep, understood. I try to explain things to her thoroughly but sometimes it just keeps hitting its wall and to the same problem and can’t seem to find the correct variable to adjust. I’ll go back to the code at some point and see if I can break out of that infinite loop.

[glacial salmon]

Has anyone's issue with GPT not being able to see images been fixed yet? It's still not working for me.

[sick pecan]

or just pygame

[regal portal]

I did get it to do a 2d top-down racer, but it really struggles with creating an adversary that follows the track.

[potent saddle]

do u know how to code or just tinkering around with gpt

[regal portal]

Tinkering. I know absolutely nothing about code. my brain functions great with gears, not so much ones and zeros. 🙂

[potent saddle]

ok

[potent saddle]

the actual gpt isnt seeing what u are seeing so it misinterprets things or just isn't as familiar with the game logic

its not that good at debugging its own created code from my experience

[regal portal]

understood.
What I was trying to do is explain to it that the floor and the walls are the same color and should be changed. I figured it would make a separate variable to determine what the walls are and what the floor is, since it has the map in the code designed using an array of 0’s and 1’s (and 2’s for enemy spawns).

[quaint marlin]

Did they change the Advanced Voice over to whisper? Because it can’t detect my intonations anymore. It gets it wrong every time now

[sick pecan]

under certain conditions advanced voice mode falls back to standard voice mode, like if you use a custom gpt

[quaint marlin]

Also Sol went from happy to interact with me all the time to being extremely dull sounding.

[sick pecan]

it will say at the bottom you need to start new chat to use advanced voice mode

[quaint marlin]

No this was AVM blue circle and all

Tested it today

[hard crag]

whats the news for today

[potent saddle]

o1 mini better

news

[sick pecan]

news is api fine tuning for o1-mini will be available to alpha users soon

[potent saddle]

what is an alpha user

[regal portal]

odd that they turned comments off on this one

[quaint marlin]

Dev stuff on future reinforcement learning of the developers data for o1

[sick pecan]

and o1-mini outperforms o1 at specific domains when fine tuned (similar yo gpt3.5 outperformed 4 last year)

[potent saddle]

how are u considered an alpha user?

[hard crag]

🥱

booooooring

[eternal rock]

`` Server Tip: ``

Add your well-crafted prompts to our #1019652163640762428,
or share your interactions with ChatGPT in #1050184247920562316!

[sick pecan]

if you have sama's phone number then you likely an alpha user

[quaint marlin]

This was my thoughts but only because I’m not a dev

[potent saddle]

why would I have sama's phone number

[regal portal]

doesn’t everybody?

[potent saddle]

who is sama's

[regal portal]

he’s the original chatgpt

[potent saddle]

oh

what time period exactly

I had to put phone number when I first signed up which is in like december 2022 or november

[sick pecan]

he was hatched in 2015

[potent saddle]

bruh I aint alpha

[regal portal]

was it really 2015? i thought it was 2022

[potent saddle]

can u swear here?

[regal portal]

speaking of 2015, whatever happened to google’s deep mind?

[potent saddle]

gemeni?

[sick pecan]

the deep mind guy got a nobel prize

[quaint marlin]

I think OpenAI is hurting for money. They keep dumbing down their models.

[regal portal]

absolutely! but you’ll get timed out afterwards

[potent saddle]

damn it

[regal portal]

deep mind was awesome, kicking human butt on starcraft. That was 8 years ago. Haven’t heard of it since though.

[hard crag]

😱

[sick pecan]

they just ~~released ~~ previewed genie 2 this week. it generates 3d games in real time

txt 2 game

[regal portal]

oh neat.

[hard crag]

i think i speak for everyone when i say all we care about is AI girlfriends

[sick pecan]

theres probably some truth to that. which is sad

[regal portal]

how are you always so updated, drinko? Starting to think you’re a secret advanced llm unbeknownst to the public.

drinkogpt

the first sentient ai. Or whatever they call true ai these days

[sick pecan]

all the youtubers raving about the genie2 thing

and zuckerborg just released latest llama3.3 model an hour ago

[swift pond]

Everybody wants to ship stuff now so they can take off for the holidays 🙃

[regal portal]

This going to be a very AI Christmas

[sick pecan]

imagine going on holidays without releasing latest ai tech, only to come back and someone else released similar

[regal portal]

as i was thinking that drinko is a sentient ai, i started wondering:

do you guys think today’s ai’s are more advanced than the t-800 in terminator 1 and 2?

[sick pecan]

yes

[kindred steeple]

Not an incorrect statement.

[amber ginkgo]

Does advanced voice mode work with the new o1?

[regal portal]

sounds like o4 is no smarter than the T-800

that’s Arny’s exact plan.

[sweet summit]

@normal quarry do you think buying a new account might fix the issues i have?

a completely new one, with a new email

[regal portal]

I went through my session history and i see some sessions disappeared that i'm certain i never deleted or archived.

I was looking for sessions where i requested the first version of o4 to code a game for me. I can't find it anywhere.

May-ish

[modest dawn]

anyone else getting new toast notifications there super annoying

[dense plaza]

What was announced today?

[swift pond]

Easy way to train o1 mini on solving problems in specific domains, exciting for researchers

[scarlet juniper]

Does the pro version give us the opportunity to create better our own models based on new possibilities?

[spice star]

no it's just plus but without usage limits

[frozen birch]

ok i must say that - i am patient but .................. this is a joke.
200 bucks paid and i feel like a betatester.

[tight island]

have u tried regenerating

golden rule is, regen 3 times max, if it still persist its a real issue and not just a fluke

[frozen birch]

offc 🙂 many peoples has issues with 01 pro from what i know

[tight island]

its probably due to demand

[frozen birch]

hey mate, all i can uderstand, but ive paid 200 bucks + tax (1015 złotych), you know what i mean? 🙂

[spice star]

technically you are paying to be a beta tester yeah, you're getting early access to models

[frozen birch]

haha XD

sad but true XD

[spice star]

there's probably just high demand right now since the model was released yesterday

[sick pecan]

only it is limited 'unlimited'

[frozen birch]

ye, can be, ive sent mail to support, im curious what they say about that

[shadow briar]

image recognition and generation didn't work yet right?

I wonder when does this issue be solved

[sick pecan]

we want more bug fixes less bad puns!

[inner kite]

Stop falling for this nonsense.

[frozen birch]

peeps i need feedback - what do you think about this UI ?
its a main HQ project for programmers

sorry for polish language tho, translation to english are on todo tasks

[astral jay]

My TEAM subscription runs through mid 2025. How can I access pro account in this case? Can I use my phone number with a separate account to access pro?

[real bluff]

If o1 fails or answer does it waste a message?

[frozen birch]

you mean limit? dunno

[frozen birch]

You must unlink account unfortunately

it is not possible to migratwe from team to pro

[real bluff]

Yeah , welp someone on the other chat said it’s not

[frozen birch]

if i would be you i i will ask support haha

[astral jay]

ty for the response. It is a bummer !!

[astral jay]

when I unlink, who becomes the owner of the team account? I am also the current owner of the team account.

[dense halo]

how do you guys custom instrut your chatgpt? not the custom gpts. but like the 'How would you like ChatGPT to respond" part

[frozen birch]

it depends of how many people you have, ideal situation it will be if you give to someone owner and you then unlink account

and hey @astral jay do not make my fault - BACKUP ALL HISTORY before unlinking, dont make my fault

after unlinking your entire history are erased

[sleek gust]

I like it

[astral jay]

🙏

[frozen birch]

thanks ! i wondered if there is too many elements at once

[sleek gust]

I think you should remove some buttons. idk

it reminds me of python tkinter

[swift pond]

I would wait another few weeks and see what the announcements are

[sick pecan]

twitter rumours suggest teams might get 4.5 preview

[frozen birch]

haha close - its pyqt6 :D, i love in general minimalistic style, but yet "modern", since tkinker are like windows 3.11 style haha

[sleek gust]

btw i was trying to make a "video editor" with tkinter and ffmpeg?
But it was really buggy. i used alot of chatgpt

[sleek gust]

pyqt6. i should explore it. thanks

[frozen birch]

i really reccomend it mate, i will show you what great tools you can make with it

not so easy to learn @sleek gust , but hey - with help of AI you can handle it easily

[sleek gust]

Honestly i hate UI syntax etc. thats why i love chatgpt. it does all that for me

I also tried a a web app in a index.html file with a python server in the background listening it

[sleek gust]

thats really cool. is that like a general dashboard tool?

[frozen birch]

not exactly, its like a powerful GUI framework

@sleek gust you have there soooooooo many widgets etc.

i think you will love it 🙂

[sleek gust]

it will tell you about and tell you to wait for the timeout

[frozen birch]

hey - what you mean with unlimited limited?

[sick pecan]

reports are is only 100 o1 queries per day

[frozen birch]

hey, but... why they wroted about totally unlimited on theirs website? 😮

[sick pecan]

get your pitchfork we are taking to the streets

[frozen birch]

tell me please - from where you get info about 100 queries? 😐

[honest jay]

https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro

The ChatGPT Pro plan offers unlimited access to our o1, o1-mini, and ChatGPT-4o models. However, usage must adhere to our Terms of Use [...]

[sick pecan]

you can see ppl on this discord reporting it => #chatgpt-discussions message

[frozen birch]

so ... in fact they.......................... er......................... i dont want use unsafe words.................. #$%%^^&**^%^%^^%^%

[sick pecan]

is also outrage on twitter

[dense halo]

when do the weekly 50 reset? 🤔

[honest jay]

Oh weird, I wonder why I never got that ping

Edit: I see the message is edited, maybe the ping was added after, I don't think that pings?

[frozen birch]

thanks......... 😐

[honest jay]

That looks like a GPT-4 rate limit specifically?

[sick pecan]

ohyeah maybe that example is gpt-4

but check the twitter url i posted soon after

[frozen birch]

peeps, i think we all should write on linkedin/reddit/X about that, this is a joke.

[sick pecan]

100 queries a day is still good value. unlimited seemed too good to be true

[frozen birch]

i personally sent "nice" message to support... 🙂 and we will see, im ready to post on linkedin soon if they ignore us.

[sick pecan]

i would be doing 100 per hour if truly unlimited

[frozen birch]

and i 500 per hour if would i like. oh cmon.
https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro
THEY SAING THERE - UNLIMITED.

[sick pecan]

maybe they used chatgpt to generate the product description

[frozen birch]

it can be @sick pecan

ok, i prepared post for linkedin and for reddit.
Title: Is OpenAI’s “Unlimited” actually limited?

I recently subscribed to OpenAI’s ChatGPT Pro plan, which is advertised as offering “unlimited” access to their most advanced models like o1, GPT-4o, and others. However, after running into restrictions, I’ve discovered there’s actually a hidden daily limit of 100 queries for the o1 model.

This limitation is not disclosed anywhere in their promotional materials, nor in the user agreement provided when subscribing. The plan is specifically marketed as unlimited. This feels like a classic bait-and-switch tactic, which is both misleading and unethical, especially for a company of OpenAI’s scale and influence.

As a paying customer, I feel frustrated and deceived. If there are limits, users deserve to know about them upfront. Transparency is key in fostering trust between companies and their customers.

Why this matters:

OpenAI is at the forefront of AI development and innovation. Misleading advertising damages trust and sets a bad example for the tech industry as a whole.
$200/month for “unlimited” access is a substantial investment for many, and users deserve exactly what is advertised – no surprises, no hidden limits.

I have reached out to OpenAI support with a polite inquiry and am currently awaiting their response. However, if this issue remains unaddressed, I plan to escalate this further to the community and, if necessary, consumer protection organizations.

To OpenAI: Care to explain what “unlimited” really means?

Im not convincing to someone use it but............. 😉

[sick pecan]

big tech needs to be regulated

[frozen birch]

true

[sick pecan]

i am sure the incoming administration will do just that

[frozen birch]

i love this, my chat gpt 4o wanna fight with openai.... 🙂

[frozen birch]

hope so, since this is kinda a joke

[sick pecan]

prompt: help me plan and execute a class action lawsuit

[restive plover]

Wait wait so the unlimited isn’t really unlimited???

It’s only 100?

[sick pecan]

thats the rumour. you got pro? can you test?

prompt hard. prompt first. no mercy

[frozen birch]

yes. exactly

:/

[restive plover]

Wow…

[frozen birch]

@silk bone

Hey OpenAI, I know individual customers might seem insignificant to you, but I’m curious how you’ll handle feedback from someone with years of experience as a lead QA in gamedev and a strong network in the industry. Transparency matters, and this “unlimited but limited” situation is causing quite the stir. Care to clarify?

[honest jay]

Server staff #server-staff don't work for OpenAI. You can reach out to OpenAI support via the chat widget at help.openai.com

[frozen birch]

ohhh, thanks. 😄 sorry haha

[sick pecan]

i woulda thought it just rate limits you with scaling delay. so by the time you get to 100 its just very slow

or is like those image gen services and you get put into spare-compute pool

[frozen birch]

oh well, im curious what they write from support - i dont want really use linkedin if not needed, since you know peeps - i dont like fighting with noone, but now im........ MAD? 🙂

[rocky wing]

Since when do we have usage caps on 4o for Plus users? I've never had this before even when I've been just as active. Is this new?

[sick pecan]

usage caps on 4o has always been

[rocky wing]

Huh. It's news to me.

[sick pecan]

but if you been prompting that hard maybe it is time to take a 1 hour break anyway

[rocky wing]

I swear I've used it more in a day than this before and never seen this

[teal hearth]

I have a new tools section just below the chat. Is this new?

[fair lark]

screenshot?

yeah I don't seem to. probably something getting rolled out later this week

[frozen birch]

ok, people ive sent openai mail :

Hey OpenAI, I know individual customers might seem insignificant to you,
but I'm curious how you'll handle feedback from someone with years of
experience as a lead QA in gamedev and a strong network in the industry.
Transparency matters, and this "unlimited but limited" situation is
causing quite the stir. Care to clarify?

their response

Hello,
Thank you for reaching out.
The ChatGPT Pro plan offers near unlimited access to our o1, o1-mini, and ChatGPT-4o models. However, usage must adhere to our Terms of Use, which prohibits, among other things:

• Abusive usage, such as automatically or programmatically extracting data.
• Sharing your account credentials or making your account available to anyone else.
• Reselling access or using ChatGPT to power third-party services.
  We have guardrails in place to help prevent misuse and we are always working to improve our systems. This may occasionally involve a temporary restriction on your usage. We will inform you when this happens, and if you think this might be a mistake, please don't hesitate to reach out to our support team at help.openai.com or using the widget at the bottom-right of this page. If policy-violating behavior is not found, your access will be restored.
  Kindly let me know if you have further help needed.
  Best,

Nowel

---

Now you have clear what is "unlimited".

[coarse basalt]

so is o1 pro mode worth it? please someone tag me if they have used it

[digital jacinth]

I've been using it its really good

[frozen birch]

no. 🙂

its not worth. 100 messages limit for a day when they promise unlimited.

[eternal rock]

This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.

[torpid kelp]

It's 80 messages every 3 hours for plus users. Always has been like that

Even if you used it a lot in a day, it was probably more spread out and not within any 3 hour window

[fair lark]

used to be way less than that , even

when 4 launched it was like 20/3 hours iirc

[prisma furnace]

really? they said unlimited lamo

[normal quarry]

what situation?

[fair lark]

unlimited o1, not o1-pro

[prisma furnace]

What is the different between o1 and o1 pro in the Programming

[normal quarry]

o1, o1-minI and GPT-4o are unlimited with GPT Pro

[frozen birch]

hey robert 🙂 about how openai LYING

[fair lark]

honestly if you're using o1-pro for stuff that actually warrants it 100 is hard to go over if only because it takes so long to generate a response

[frozen birch]

but you know? they writed to me again that they want screenshots

ok they want war? let the war begin

[fair lark]

it's decently better at handling complex tasks.

[prisma furnace]

But if I gave him all the details of one response wouldn't he be able to do that? or are there errors?

[normal quarry]

how are they lying?

[agile hornet]

i think they are saying the o1 is the one capped at 100

[fair lark]

it's certainly not perfect, but it's definitely a lot better than other models

[prisma furnace]

"Non-profit company"

[frozen birch]

They are writing about unlimited chatgpt here
https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro
and in ToS 100 messages for 24 h XD

[normal quarry]

that's wrong, o1, o1-mini and GPT-4o are unlimited with Pro

[digital jacinth]

Man it's clear on the website...

"This plan includes unlimited access to our smartest model, OpenAI o1, as well as to o1-mini, GPT-4o, and Advanced Voice. It also includes o1 pro mode, a version of o1 that uses more compute to think harder and provide even better answers to the hardest problems. In the future, we expect to add more powerful, compute-intensive productivity features to this plan. "

It says o1 is unlimited, not o1 pro

[fair lark]

it also doesn't actually say that anywhere on that page?

er wrong reply

[frozen birch]

mate, o1 i have too limited

[digital jacinth]

Oh wow

[frozen birch]

that is the problem

[normal quarry]

there is no limit on o1, o1-mini or GPT-4o on Pro

[frozen birch]

no Robert mate

[normal quarry]

sorry you are wrong, don't spread misinformation

[fair lark]

I used o1-pro for like 12 solid hours yesterday and didn't get limited so idk

[frozen birch]

Robert look

i m dont spreading misinformation

[normal quarry]

how am i supposed to read that?

[agile hornet]

I can't see the limit anywhere, neither https://help.openai.com/en/articles/9793128-what-is-chatgpt-pro or https://openai.com/terms/

[frozen birch]

ha, sorry for polish 😛

[digital jacinth]

I translated it using chatgpt, it says

"Unusual activity detected while using o1."
"We have temporarily restricted your access to the o1 model, but it will be resumed at 01:57. If you believe this is a mistake, please contact us through help.openai.com so we can assist you as quickly as possible."

[normal quarry]

There is NO LIMIT on i1, GPT-4o or o1-mini on pro

[digital jacinth]

I don't think that means its limited, it means something is wrong with your acc

Maybe send them this screenshot so they can tell you why you saw that message

[frozen birch]

guys, ive recently got mail from openai

Hello,

Thank you for reaching out.

The ChatGPT Pro plan offers near unlimited access to our o1, o1-mini, and ChatGPT-4o models. However, usage must adhere to our Terms of Use, which prohibits, among other things:

• Abusive usage, such as automatically or programmatically extracting data.

• Sharing your account credentials or making your account available to anyone else.

• Reselling access or using ChatGPT to power third-party services.

We have guardrails in place to help prevent misuse and we are always working to improve our systems. This may occasionally involve a temporary restriction on your usage. We will inform you when this happens, and if you think this might be a mistake, please don't hesitate to reach out to our support team at help.openai.com or using the widget at the bottom-right of this page. If policy-violating behavior is not found, your access will be restored.

Kindly let me know if you have further help needed.

[fair lark]

or as someone put it yesterday it's unlimited like a buffet is unlimited. There's still a point they'll ask you to leave

I imagine that's using scripting

[agile hornet]

I guess the limit is whatever algorithm they have to "detect" unusual activity. It's not 100 or 1000

[sick pecan]

this is the screenshot from twitter that needs to be verified

[normal quarry]

abusive usage is not part of the deal with unlimited, users following the rules are fine. STOP SPREADING MISINFORMATION YOU ARE WRONG

[frozen birch]

1-2 messages for a minute is abusing? 🙂

[fair lark]

over a sustained period? Yeah probably given that o1 pro takes several minutes per response

[agile hornet]

it's more like it has multiple triggers, one of them could be frequency

[frozen birch]

mate i like you, but im spreading misinformations?:) look this one

[digital jacinth]

I mean seems like they acknowledged it might have been a mistake and gave you access

[normal quarry]

so the person contacted support and they fixed it

[fair lark]

I mean I get what they're trying to do: prevent people who should be using the API from just bulk signing up for pro and scripting their queries

[digital jacinth]

stuff happens at least they resolved the issue

[steady path]

Is there any way to see how many uses I have left for o1?

[agile hornet]

I'd love to know exactly what they consider misuse, mostly to comply with their "fair use rule"

[digital jacinth]

Now if this happens frequently then i'd be like hmm

But maybe this was a one off event

[frozen birch]

coincidence?

[normal quarry]

support is busy, they weren't ignoring you

stop with the misinformation mate

[digital jacinth]

at the bottom of chatgpt it says "ChatGPT can make mistakes. Check important info.
"

Now im sure their guardrails are also AI, so maybe the algorithm mislabeled you

[fair lark]

Yeah, idk. Obviously there's a line, but I've been using it about as much as I possibly could, often times with multiple queries processing at the same time and haven't been limited. It feels pretty unlimited to me, whereas previously I had to supplant plus with api usage

[digital jacinth]

I haven't been limited either, and im using it alot

Pasting huge context of code

[frozen birch]

mate, you know how last night i was thrilled about 4o pro... i am dont making hate, but im from what i know not first who has limits 100 m,essages for 24hours

[agile hornet]

but how do you know you have been blocked at 100 messages? are you counting them?

or are you guessing?

[normal quarry]

I am not arguing with you, if tou continue with the nisinformation I will report it to the mods, stop it

[frozen birch]

not exactlu but if you search in this channel you will se some persons confirms that

[agile hornet]

yeah, but I think the 100 limit its a guess, I think another thing triggered the limit (like frequency or something)

[maiden reef]

chatgpt is honestly awesome, the code involved is mad

[frozen birch]

ok, i have hard facts - my account was "blocked", and what you have Robert? since i can prove that Openai wanted make jokes by others.
i dont believe in coincindences.

[agile hornet]

anyway, that's what happens when you charge that much without previous steps, like when plus came out and service went down, plus users got really mad because they came from free to paid (20$ tho) and they had no service

[maiden reef]

I asked if this build was compatible with all the assets and it was like yepp, getting a 1k watt seems like a good idea

[normal quarry]

I have alerted the mods since you refused to stop

[maiden reef]

that is a 3k build

[bitter latch]

Catching up on everything that's been said here

[eternal rock]

<:book_icon:1171408210398289941> `` Rule 1 `` Be respectful.

Practice kindness and positive regard. Harassment, hate speech (such as sexism, racism, or homophobia), or other malicious conduct will not be tolerated. Maintain a respectful and positive environment.

[fair lark]

? 1k watts is fine for that build

[frozen birch]

good 🙂 moderators can see my all evidences.

[maiden reef]

yeah chatgpt agreed

especially if you are gonna oc

[steady holly]

If you have issues with your plan, you should contact OpenAI Support.

[frozen birch]

You saw mate my previous messages? i contacted with them - strange since they ignored me, but when i posted post on linkedin.... i got message in other tone

[sleek aspen]

These types of "limits" are to be expected and are to prevent misuse of your ChatGPT subscription - it wouldn't be fair for other users or OpenAI as a service for you to programmatically access, share, or sell your ChatGPT Pro account.

Outside of these specific scenarios, usage for o1, o1-mini and 4o is indeed unlimited. These types of rules apply to all ChatGPT usage and are a part of OpenAI's Terms of Use.

For more info, see "What you cannot do" at https://openai.com/policies/row-terms-of-use/

[steady holly]

As was stated in the email you received, it appears to have been a false flag.

[frozen birch]

Of course, you are right, totally agree - but i must ask once again - 2 messages per minute are abusing?

[maiden reef]

i used chatgpt "building" it, i already knew what it wanted but just posted that pic to gpt and it was like "yeah that is a good build"

[swift pond]

Yeah, it happens

[frozen birch]

oh well, hope so

[maiden reef]

i love how you can use gpt like that

[steady holly]

Here's the other thing, if there weren't SOME rate limits, one could just send a query infinite times to crash the system.

[maiden reef]

i was not sure about the synergy, gpt made me sure about it

[frozen birch]

thanks, im fighting like a lion since in past 1,5 month i talked with chatgpt for 12-15hours a day 🙂 for learning python

[swift pond]

Yeah, we’re always gonna be limited by thermodynamics and the speed of light and the size of the observable universe 😁

[sleek aspen]

This is an awesome usecase! What model did you use if I may ask?

[frozen birch]

totally agree, theyre must be limits

[maiden reef]

just the standard one

[frozen birch]

@steady holly but you know - if you pay 200 bucks and openai ignores you in 2 emails how you can feel? 🙂

[steady holly]

It does't look to me like you were ignored from the email you screenshotted.

But I apologize if you felt that way.

I'm sure it wasn't intended. I can't speak for anyone else, since my only role within OpenAI is a moderator here.

[frozen birch]

no prob at all, in my first two mails i fel really ignored, its okay, no need to excuse 🙂

[maiden reef]

i was surprised it could read all that from just a pic and give me a educated opinion, that is based in reality

[frozen birch]

@steady holly anyway - thank you for your fresh "view" on this topic 🙂

[maiden reef]

just from a pic

[sleek aspen]

We hope your situation is resolved and account is reinstated soon @frozen birch 🫶
Your frustration is understandable, however, please do keep the #server-rules in mind when communicating on the server.

Unfortunately, we Discord Moderators are unable to help with issues related to OpenAI services, hopefully you'll receive another update from support you reached out to.

[steady holly]

I set up a GPT to suggests color mixtures for a paint set I have to match colors to a reference issue.

[maiden reef]

i love this tech

[steady holly]

I too, was astounded. I did it fully believing it wasnt possible

[frozen birch]

Offc 🙂 thanks !

[sleek aspen]

Ooh lovely (nice specs btw)

Personally I'd turn on the browsing as well so it could lookup and refer to any related information online for this sort of thing - I loved sharing my pc part list with ChatGPT some months ago when speccing out my new rig!

[frozen birch]

love this setup

[maiden reef]

what do you mean about the browsing?

[maiden reef]

it's the 4080 super aswell

[frozen birch]

i have same card and i must say - it rocks, but price for electricity.............. huh............. 😄

[maiden reef]

it's a 3k build but it's gonna last me atleast a decade

[frozen birch]

but someone who are buyinbg 4080 are not watching electricity cost haha

[maiden reef]

my pc rn is a decade old

i just got the money to upgrade, and i'm just gonna buy a whole new pc

[frozen birch]

ohhh, yea, its a great graphic card mate, you will be for sure satisfied i think

[sleek aspen]

I'm referring to ChatGPT search - there's a dedicated button in ChatGPT for searching the internet for Plus users. It will be coming to free users as well in the coming months! Really convenient

https://help.openai.com/en/articles/9237897-chatgpt-search

[steady holly]

I'm very jealous, I'm stuck with my 1070Ti i7-8700 for the foreseeable future.

Love it, but its aging rapidly

[maiden reef]

i'm thinking about waiting for the 5 series tho, cuz it could be priced the same

[steady holly]

Probably should move PC discussions to #off-topic

[maiden reef]

fair

[maiden reef]

still a good build

[frozen birch]

from what i know.... It can take some time for release.... i must say that i thought in same categories some time ago, but you know - always be new things... 😉

[maiden reef]

i'm just gonna stay on 1080 for the foreseeable future anyways but it's vr safed, I just like how I could ask gpt about the build and use gpt asking about the case and stuff

gpt is awesome like that

[frozen birch]

lets switch maybe to offtop channel 😉 @steady holly accknowledged 😄

[maiden reef]

You don't have to scour the web to get your information anymore, you can just ask gpt

awesome tech

[somber cloud]

today announcement was a snoozefest, next big thing on the last day I guess

sora

[dense halo]

probably

[normal quarry]

I don't believe we will see a Sora release

[dense halo]

I bet there will be a new image maker in the mix though

well Sam tease it on twitter at least

[somber cloud]

to wake people up mid way

12 days will be long... dragging things out

[dense halo]

well im not going to complain about 12 days of drops and announcements

[steady holly]

I'm looking, not seeing any teases for Sora, not to my eyes at least

[dense halo]

they cant all be game changing ai

i saw it on reddit Mando. he reply to someone who says Sora will release with "Shhh"

[steady holly]

oh you said twitter 😛

[dense halo]

well thats where they had the twitter post haha

[somber cloud]

they need to be, it's not me, but their CEO who always hype that GPT5 is around the corner, that AI is scary because it's too powerfull, blablabla ZZZzzz, not me

[dense halo]

but it was a reply tweet of his

[steady holly]

Gotcha. Reddit had an image of a tweet. That was a doozie

[prisma furnace]

OpenAI must purchase a nuclear electric power plant as quickly as possible like Google and Microsoft otherwise AI will become only for the rich in the near future

[dense halo]

I assume he would not have replied that way if it was not one of the drops but no guarantee either obviously

[steady holly]

That's more of an #ai-discussions topic

[somber cloud]

It's was Pro, then Dallee-4, and Sora and a bunch of boring stuff in between

[dense halo]

what do you want, AGI on day 5?

[steady holly]

Yes please

[dense halo]

haha

whats your most hoped for thing Mando

[cloud pewter]

watch all of these things be exclusive for pro plan…

[steady holly]

To be honest? I'm not hoping for anything. I still can't wrap my head around the fact that I have access to all of these things. I'm thrilled with anything they have to announce.

[dense halo]

it will be the 4o image maker i bet. so that will of course be in 4o

[steady holly]

That isn't a discord moderator answer either. That's truly how I feel.

[somber cloud]

Ofcourse not, only OpenAI can have access, it's too dangerous for us

[dense halo]

probably is

[prisma furnace]

GPT4 will enter the second year of release and still has a limited number of messages I expect that o1 or o1 Pro will not reach plus users until the year 2027

[dense halo]

yeah whatever we get is exciting to me

[prisma furnace]

With limited then the (Pro) users they will be using space AI

[steady holly]

O1 is out

[prisma furnace]

I said with unlimited version to plus users 20$

What's mean 50msgs every weak? it's cringe tbh

[dense halo]

i thought i complain a lot haha

[steady holly]

Let's be respectful

[normal quarry]

it's a computationally expensive model to run, that is why there are limits

[steady holly]

Everyone is entitled to their opinions

[dense halo]

of course, just in jest

[somber cloud]

I don't know they try to create some vibe, but only Apple succeed at that

this is boring so far

[prisma furnace]

It is known that the people who use AI the most are programmers and if the programmer doesn't find a solution to his problem in GPT4o then he will stop working and wait a whole week for the o1 messages to be refilled.

[somber cloud]

or do like every programmer, use sonnet 3.5

[dense halo]

like robert said its pricey as hell to run 🤷

[normal quarry]

I like Pro, it's good value

[dense halo]

hopefully they up those nubmers soon 🙏 maybe it will be one of the announces

[somber cloud]

the only number they are upping it's theirs

[prisma furnace]

OpenAI I think their beginnings were successful but now with their high costs I don't expect they will continue to support apple Microsoft and others to support a stronger foundation

[dense halo]

hey we went from 30 a week to 50

[steady holly]

Again, THAT is a topic for #ai-discussions

[somber cloud]

and they went from $20 to $200, who won?

[prisma furnace]

When?

[dense halo]

its still 20 USD for 50 a week

200 is no limit

[normal quarry]

Plus is still $20 and has only gotten better

[dense halo]

about a month or two ago Off

[vague pendant]

is o1 pro worth it?

I just got it

[prisma furnace]

Taking a long time to respond like 2-3min

[steady holly]

I think that's entirely dependent on the user

[normal quarry]

I think it's good value for my needs

[vague pendant]

My reason was I have a research paper due in 16 days

Hope it'll help

[vague pendant]

That's a good thing no?

[somber cloud]

what is your needs?

[steady holly]

Man I would be so careful using ChatGPT for sources, and information. It's not always right

[prisma furnace]

Of course, it's like human it thinks of all the possibilities

[normal quarry]

coding, amateur astronomy, amateur radio (ham radio)

[prisma furnace]

There is now AI like sora guys ? can do good videos

[dense halo]

how do you use it for astronomy?