[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: 1772399227090.png (116 KB, 512x512)
116 KB
116 KB PNG
>send 3 messages
>you have used your usage limit for the next 12 hours, please buy more credits
When are they going to fix this shit?
>>
>>108272511
Never, running good AI costs money, 'specially if it gets overloaded.
>>
they lost all their income because they needed to be woke so now they get fucked hard kek
>>
>>108272511
upgrade your plan, switch to api, or switch to another AI. the cheap claude plans are a scam.
>>
>>108272531
>woke is when you refuse to support global jewry
>>
>>108272511
They have to use rate limiting to avoid resource congestion, otherwise their server infrastructure might become overloaded.
>>
>>108272511
There is a very simple fix. Pay.
>>
>>108272511
>the best AI model has the lowest quota for plebs
makes sense, aksually.
>>
File: file.png (190 KB, 498x498)
190 KB
190 KB PNG
>>108272511
>poorfag
>>
It's a downside of using high-end, compute-intensive cloud services like advanced, heavy LLMs.
>>
They have to pay the electricity and cooling bills somehow, and those LLMs use a LOT of electricity and generate a LOT of heat.
>>
File: texasinstrumint.jpg (48 KB, 700x700)
48 KB
48 KB JPG
>>108272511
They're going to solve neurosynaptic computing this year, and we'll get a personal model 100x as powerful with realtime learning that can run on this much power
>>
why do people pay for AI
if you go to bing.com it gives you AI for free
>>
Use lighter models for things that don't require ultra-high-end reasoning and thinking and reserve the heavy stuff for the things that do.
>>
>>108272798
this. and if it ever runs out of 'credits' or whatever I'll switch to grok until it cools down.

I'm curious if the paid version of google AI lets you generate more images, or different resolutions, or removes the watermark. they don't make that very clear
>>
>>108272511
Even the $20 plan supports a lot more usage than I expected. Were those "3 messages" deep research requests with a dozen PDFs attached?
>>
>>108272511
?
I have never encountered a limit, is this new?
>>
>>108272805
I will ask opus 4.6 to write me a commit message because it's paid for by work
>>
>>108272518
>good AI
lmao
>>
>>108272511
You say
>fix
like it's not working as intended.
>>
>>108272607
if we use the original definition of woke before it was turned into "thing i don't like", unironically yes
>>
>>108272787
The LLM has already gotten to you:
>you have no idea what you're talking about
>but man, do you sound confident about it
>>
How do I get paid AI for free? I'm a self taught musician and electronic hobbyist. I dont want to wait 4 days for a redditor to post affiliation links or some retard on 4chan to insult me.
>>
>>108272511
>NOOOO making god work for me should be hecking cheap
an hero
>>
>>108275155
Software engineers have always been like that
>>
>>108275203
Code monkeys: yes.
Actual engineers: ... debatable.
>>
>>108272531
>woke is when you refuse to give your shitty AI to corrupt maniacs in power who want to use it against their own citizens
>>
>>108275220
>referring to anyone working with code as an engineer
good one
>>
>>108276115
Did you take your meds?
>>
>>108272774
Can't they just send the heat to cold places like Alaska and Amdarmica? .. they could used it to be warm
>>
File: 1753408833437245.png (974 KB, 2022x1734)
974 KB
974 KB PNG
>everyone starts coding with AI
>everyone relies upon the same attack target and endpoint
so, china just has to take down claude and they can damage software engineering productivity for everyone? is it really OK for us to all rely upon claude?
>>
>>108272511
it got me two buy two separate pro accounts and seed both with $40 of extra usage credit.
worth it. far cheaper than wasting my life at college for a retarded degree that will end up having the same outcome that i am in now but have 200x more debt that i am now because of said worthless degree.
everyone that went off to college are the biggest retarded in the world.
>>
>>108278226
>are the biggest retarded
>biggest retarded
Dunno if you're naturally autistic or are just paid to, but you sure are nailing the projection part.
>>
>>108272511
I thought limit reset daily. I just saw that I somehow hit my weekly limit so I'm cut off until next Saturday... that's kind of bad, I have not been using it that much.
>>
>>108281207
Uh, it seems that I did not have a weekly limit until I let my subscription renew last week. So not only was the renewal about twice as much as I paid for the initial yearly subscription, but it came with additional limits. WTF anthropic.
>>
>>108272798
Anon, I would pay money to not use bing
>>
File: 1765672806876936.jpg (150 KB, 1024x1024)
150 KB
150 KB JPG
>>108272511
Qwen 3.5 doesn't have this problem.
>>
>>108272511
I can use it for hours before reaching the limit now. It wasn't like that some weeks ago
>>
>>108272511
use sonnet? claud is the expensive one
>>
i use haikku for my retarded questions chat. its much cheaper
>>
>>108284707
>>108284707
i think you mean opus retard
>>
>20$ subscription
>max 3 prompts with claude code then a 5 hour wait
so cool, totally worth the money
I really want an alternative but no competitor even remotely comes close to opus.
>>
>>108286365
>tfw unlimited use through company
I am not even sure how much cash I burn through but they keep saying no limits, please use AI more
>>
File: 47d01c674b4.jpg (2.16 MB, 1988x2415)
2.16 MB
2.16 MB JPG
Claude Opus 4.6 solved the first VR puzzle 2 weeks ago and then forgot the solution and it's just been randomly pushing the boulder around since
>>
>>108286365
GLM-5 and Kimi 2.5 are very close. And Codex 5.3 is really great at coding with very generous limits at the low tier.
>>
>>108272511
I pay for the $200 monthly Max plan and have literally never encountered this.
>>
>>108286365
$20 is jack shit in 2026 lol, especially considering the value you get out of it.
>>
>>108290617
What are you building
>>
>>108290705
What the pokemon discourse taught me last week is that a lot of people don’t have $20, meanwhile I went to get boba with my wife yesterday and had to pay $20+tip for some leaf juice with milk.

How do I filter out all the people who can’t afford $20? Nowadays it feels like it’s just what you spend when going outside.
>>
>>108272518
>good AI
Sonnet's good, but their "voice mode" is just a TTS module slapped on top. Embarrassing, altman has them beat in that regard
>>108272787
Midwit jeetpost
>>
File: IMG_20260302_122317.jpg (31 KB, 480x620)
31 KB
31 KB JPG
Alter your learning rituals to where you only need to ask 3 questions a day to the magic machine. Simple as...
>>
skill issue. what are you doing here? go home and do some homework.
>>
>>108272584
>the cheap claude plans are a scam.
There are only two: Pro or Max. Which one is the "scam"?
>>
>>108272906
>I'll switch to grok until it cools down.
grok is shit
>>
>>108275191
Have your job pay for a subscription.
>>
>>108286365
I pay $20 a month and Claude does enough work to cut my 40-hour week to just 5, and I have *plenty* of tokens left.
>>
>>108290752
Building?
>>
>>108292727
i have 1 supergrok account and two claude pro accounts + seeded with extended usage balances. i bounce between both to check each others work. stuff that i don't want to waste usage with on claude i send to grok.
they both have their strengths. grok is great for diff checking and code review. claude (4.6 opus extended) is great at actually building the stuff and fixing things. i have been blown away at how well opus 4.6 extended works at coding. its what made me realize coding is dead as a serious profession for 95% of people. if this is a sign of things to come, you will need to be exceptional, truly, exceptional to be useful in the market.
grok is fine for smaller things like shell scripts as he does really well with them from my experience. claude does it exceptionally well, but not worth wasting his usage on that when grok does those things fine. but actual coding work, grok is pretty poor compared to claude. which is odd. he does a good job at evaluating finished code, but not so well in comparison to actually building it in the first place and making fixes..
i don't know if its due to design philosophy or memory limits. opus 4.6 extended can pump out three files with 1000 - 2000 lines of code just fine and works the first time while grok struggles with putting out one 700 line of a supposed 1200 line file without placeholders.
>>
>>108293527
>opus 4.6 extended can pump out three files with 1000 - 2000 lines of code
1000 - 2000 lines, for each of the three files
>>
File: 1749402715352.jpg (101 KB, 1200x640)
101 KB
101 KB JPG
>>108291463
>>
>>108272531
Didn't everyone who was using chatGPT move to Claude on principle? I certainly did. Op is right but at least I hadn't become dependent on it to the same extent and am just using my brain now, but almost everyone I know has switched to Claude. Its not good but now that it has more users and more funding it will hopefully improve and at least we can all sleep at night
>>
>>108272531
Fuck niggers. Fuck kikes. Fuck spics. Fuck jeets.

With that said, fuck OpenAI. I use grok (general use) and claude (coding).
>>
File: HCi2yisXsAAfWDk.jpg (72 KB, 1567x281)
72 KB
72 KB JPG
>>
>>108292723
There are now two max plans. One that gives you 5x pro, and one that gives you 20x pro.

The max plans used to seem retarded, but they're lowering the limit on the pro plan to make them make sense, to desu it works. It is making me think about it.

Gemini also lowered the limit recently I think. For the first time in ever yesterday, I've hit a limit after just a few messages, whereas recently I was spending a lot of time on it, without ever hitting any type of limit. Maybe it went in a reasoning loop in the background or something and consumed a fuckton of tokens, or maybe Google has some other cool off mechanism because I was kind of unusually rude to it (why does it keep trying to monkey patch shit in ridiculous ways whenever there's any type of issue that could be fixed way more easily), but "you've reached the limit come back in n hours" part seemed new, or at least unusually quick.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.