False equivalency.
Most ML models today are a solution looking for a problem, not the other way around.

Also Luddites weren't against the idea of industrialization, but rather they were afraid their manual work would no longer be needed. They did not fight back against having cheaper furniture and whatnot, that's something they liked. That makes them antithetical to the anti-AI crowd that despises everything AI on principle to the point of actively avoiding it. But again, a solution looking for a problem, they were never comparable to begin with.

>>108997036
all those other techs are deterministic not probabilistic like chatbot llms are

>>108997053
chatbots can be deterministic if you disable random sampling

>>108997060
theoretically

>>108997076
why not? i have tested it myself. same inputs outputs the same logits over and over

>>108997095
in a small scale test in a controlled setting maybe, but there have long been issues with that
https://thinkingmachines.ai/blog/defeating-nondeterminism-in-llm-inference/

>>108997036
>Op believes things have improved since each of those "milestones"

>>108997127
thanks, but this is a problem with common software implementations. the model itself must be deterministic if no random processes are introduced. i can manually compute a forward pass on paper 100 times and expect it to always be the same

>>108997036
>factory worker
>factory worker
>factory worker
>"prompt engineer"

>>108997036
Where's the last one's hat?

>>108997036
there's no hat this time

>>108997162
yes, but if in practice, for something in production, even with temperature set to 0 it's hard to guarantee that it will be the exact same due to practical constraints, and any change can have cascading effects, there's that too I guess

>>108997036
I'm not necessarily anti-AI, it's just that I don't like the way thirdies use it to shit up the Internet more efficiently. That's it.

I'm perfectly okay with, and encourage, local use of AI. I do have issue, however, with how the consumer market has in part been affected by data center needs. The other half is the fault of retards marking up everything on the secondhand market. Thankfully I don't live near a data center (at the moment), so I don't have to worry about data-center-humming-induced Havana Syndrome--though I don't think buying up arable land is a particularly good idea.

One also has to take into consideration how the technology is being used, and how it can be used; how the powers that be are looking to apply it. Personal computing and CS has become niggerlicious ever since computers were made easier for normalfags to make use of, and since the learn 2 code meme was shilled; AI-assisted/generated code is niggerlicious, but it's being applied onto a generally already niggerlicious foundation. My main concern is how this technology can be used for surveillance, which is what AI companies seem to be either partially branching into or pivoting towards.

File: 78A7BF30-FCB7-446B-8840-5(...).jpg (1.29 MB, 3464x3464)

1.29 MB JPG

>>108997036
Luddites are dumb as fuck bro, do not expect good replies

>>108997036
I noticed that each of these innovative breakthroughs is correlated with the ever growing human population. Maybe once AI develops fully enough it will finally go full skynet and begin correcting that

>>108997434
Hopefully it'll start in India first

>>108997036
Because it automates and replaces too many jobs too fast in too many industries and it will break the economy. Also, unless you are among the 1% of the population that have large tech businesses, it's completely useless for you. Steam engines at least gave everyone a car, AI doesn't give anything to a consumer except generating snailcats or anime porn. I don't know, maybe this schizo >>108997396 is ok with trading his livelihood for temporary access to generative AI because he's mentally ill, but most people are not.

>>108997535
You’re trans and there’s nothing you can do to stop aiGODS

>>108997542
ok ai chad you are strong and fast

File: IMG_2579.jpg (192 KB, 534x528)

192 KB JPG

>>108997548
Yes

>>108997396
While AI is only part of that particular equation, what do you think about modern camera gear getting so good, photographers are starting to wonder if they're becoming glorified button pushers?
Given we tend to talk about how photography made painters obsolete (for certain use cases), it seems somewhat relevant to the AI discussion - and that's before we factor in AI image generation.
> https://www.youtube.com/watch?v=ulkqXvek5fA

>>108997565
All new tech is good, if someone hates tech then they are trans luddites
This is easy to understand lil bro

>>108997036
the printing press put scribes out of work
the steam engine put whole industries out of work
robots have already put you out of work
AI is putting you out of work

overall, the things we've lost are greater than the things we've gained, only the rich have benefitted the most from all of these advances.

but you'll never know, because you are still blinded by the glitter, one day you'll find out it was cheap chink shit and suffer in the process.

>>108997036
your scam relies on how well can nft bros lie
turns out they were not very good at it

>>108997060
>chatbots can be deterministic if you disable random sampling
Oh really? You can predict the result before you generate? Of course you can't you fucking retard. A reproducible output is about as useful as copy-paste. But what do you expect from this retarded generation that thinks AI pulling a project from github and changing it up a bit is the same thing as writing something from scratch.

>>108997665
>You can predict the result before you generate?
yes. why not? it's one long math equation. apologize to me right now, nigger

>>108997599
Yeah sounds about right. Some anons talk about how tech supposedly hasn't advanced in like years, but I've never been happier, tech-wise.
Now, the 90s (and even 2000s) were comfier in some regards, but tech is not one of them.

>>108997732
NTA, but explain to us the intricacies of the math involved, and why you're so sure the outcome can be (theoretically) determined, disregarding any lack of practicality in doing so.

>>108997036
Because it's fundamentally unprofitable and highly unpopular in it's current implementation.
Plus AI are not accountable on any legal framework. At the very least you'll always need someone monitoring or working with it.

But call me when the frontier slop models aren't trying to be sold as services. That's a dumb idea to begin with given how inefficient they are.

>>108997617
>>108997036
>robots
Industrial robots never put anyone out of work. Well, unless of course your dream job is a slave labor of standing in front of a conveyor belt and screwing caps on toothpastes whole day. Industrial robots either do things that humanly impossible like rotating a 50 ton metal construction while welding it, or conveyor packaging aka job that nobody wanted to do hence factories always had to hire tweakers to do it.

>>108997060
>>108997665
>>108997732
>>108997766
If I solve a math equation to predict the answer to the same math equation it's not really a "prediction" isn't it

>>108997766
if you don't already understand the concept of a function approximator, then it would be a waste of time for me to explain anything to you
>>108997808
are compilers not deterministic because i can follow the logic tree in the source code to predict what the binary output will be?

>>108997396
It requires actual intelligence to solve a task manually. Intelligent people will become rare. I like to be that rare person and might get me a job in the long term.

>>108997036
>use "AI" in production
>it randomly fails
>people die
>millions lost

>>108997820
compilers are deterministic, so is the AI outputs
it's just not a "prediction", it's an incorrect word in this situation

>>108997823
>no liability because AI
>PROFIT!

>>108997849
i think of prediction the same way as calculating the physics for a bouncing ball that i am about to drop. i still predicted it correctly if i observed the ball bounce the way i predicted on paper prior

>>108997854
That's not how it works in real life nigga. This only happens in US shithole companies run by indians.

>false dichotomy strawman garbage
you're as bad as the climate schizos
>OH YOU HAVE LITERALLY ANY OPINION OF THIS OTHER THAN THE APPROVED CONSENSUS(TM) WHATSOEVER?? YOU MUST BE A DENIER/HATER
i don't need to do anything to defeat you op, it will be funny watching your ponzi scheme collapse

>>108997867
that's exactly how it works in real life nigga and that's exactly why companies like it so much. if something goes wrong and your ai tech support chat tells 100 people to kill themselves and they do it company either blames openai and Altman personally and shifts the liability on the ai provider or it is not classified as company's liability at all.
>This only happens in US shithole companies run by indians.
bro business only exist in US; how many fucking tech companies do you think are in EU? like 2? Nokia and what else? EU business is kebab places and malls

File: 1758650800452.gif (2.97 MB, 300x219)

2.97 MB GIF

>>108997036
>printing press: immediate, demonstrable utility (plus huge amount of unpredictable social change)
>steam engine: immediate, demonstrable utility (plus huge amount of unpredictable social change)
>robots: some modern benefit, many false leads, not very much social change
>ai: absolutely not one single demonstrable use case that has provided any genuine benefit in any aspect of life or industry beyond stroking the egos of schizo childless retards and failure-bound ceos
>ai code is all slop, sinks companies, and requires 10-100x the manhours to repair
>ai "art" is repulsive garbage
>ai essays of redditspeak trash litter the internet, further muddying an already deep brown shitriver of discourse
>ai videos pure cash grab garbage
>ai health care leading to immediate harm and deaths
>ai "answers" training retards who already spent too much time scrolling to completely give up ever having a brain that can think about problems at all
literally the one and only thing ai has proven to be "good" for is funny cat videos and degenerate porn masturbation.
there's not a single proven case of ai improving anything else but making companies worse

you are a false equivalence, most likely a ponzi schemer hoping to cash out by hooking enough rubes with FOMO. "ai" is not the printing press/steam engine

>>108997036
Everything is a grift nowadays.

>>108997051
>solution looking for a problem
Subjective. The same could have been said of cars when horses worked just fine, arguably better on unpaved surfaces. Then infrastructure changed, with asphalt roads, lane markings and gas stations that eliminated most of the teething issues cars had. Same deal with AI, we're in the early stages where infrastructure isn't quite there yet. Once the issue are smoothed out why won't companies pay for low skill white collar work when AI is capable of doing it faster and cheaper?

>>108997909
>The same could have been said of cars when horses worked just fine
No it couldn't. That's a false equivalence.

>>108997909
NTA
Steam powered cars were a technological innovation and technological innovations do not have to be a product and have a "problem/solution" business model or target groups. AI is also a technological innovation and I think later down the line we will actually use for a lot of cool things, but as of today it was immediately commodified by its sole providers, so now it is a product. Defined specifically by people who made it. So other anon is right, AI is a solution product that is looking for a problem (for more than 5 years now btw).

>>108997889
>thins nigga thinks only corporations exist

>>108997935
oh sorry didn't account for a joe poopfart that uses ai to generate snailcat memes, extremely important and profitable market group

>>108997941
So without millions of whites using the "AI" to make memes, what's the usecase again? You cured cancer with "AI" already?

>>108997036
>the printing press didn't eliminate all hand writing
>the steam engine didn't eliminate all other energy generation
>robots didn't replace all physical labor
>AI won't replace all white collar work
There's no contradiction.

>>108997036
its just like neal stephenson said in his book diamond age, the rise of do it yourself hipsterism will put to shame the short of neo great depression larp of the 2008 hipster. the hand crafted goods industry will be huge, and the hipsters wriggling thru it will be annoying as fuck

>>108997914
>No it couldn't.
Yes it could have. Early combustion engine cars would have seemed impractical and pointless compared to a draft horse. They were novelties and they're practical use for transportation was only realised later.
>>108997934
>Steam powered cars were a technological innovation and technological innovations do not have to be a product and have a "problem/solution" business model or target groups.
I don't disagree. AI is a technological innovation/novelty that is currently transitioning into the product stage.

>>108997665
>moves the goalpost away from its random to "well YOU cant predict it"
something being deterministic and someone needing to predict it are two different things, although your luddite nigger iq is wrong in both cases since its special pleading anyway

>w-well this time is different because <irrelevant feature of the tech>!

>>108997535
>it will break the economy
good, let it crash and burn bro (wont happen anyway)

File: 1750792481669306.png (90 KB, 896x420)

90 KB PNG

Reading this thread and all the same, lowest IQ logical fallacies of "truly conscious" "people" continues to show just how retarded humanity is, now imagine the average human outside of a contraversial autism tech oriented board. Lmao.

>>108997036
ermm can you not be a faggot?

>>108997051
>False equivalency.
Did you learn that on tiktok?

File: IMG_9174.gif (1.94 MB, 295x221)

1.94 MB GIF

>>108998075
You first.

This reminds me when digital art tools started arriving. At first it was like it's not real art, it's photoshop said the airbrushers. Then it was it's not real art it's computer generated imagery said the sculptors and practical effects guys when 3D matured enough to start replacing them.

At the end of the day airbrushing is dead and practical effects are a meme and have been reduced from being the start of the show to the ugly redheaded step child.

A.I. won't win because it's objectively better, it will win for the same reason as the other technologies won before it, because it is the path of least resistance and that's what dictates everything in life.

>>108998065
>image
I don't disagree with all of it but it's completely wrong to say that AI experts in 2020 had no idea where things were heading.
>The first version of GPT, also known as Generative Pre-trained Transformer, was released to the public in 2018.

>>108997051
What are you talking about? AI solves all problems. Pick a problem, AI is the solution—there is no searching: just solving.

File: asdf.png (172 KB, 1000x500)

172 KB PNG

>>108997036
This time they're actually going to exterminate you, and I fully support them in that.

File: IMG_9471.jpg (44 KB, 720x404)

44 KB JPG

>>108998330
A lot of those so called experts said openAI was stupid for scaling like they did and would lead to no where then they all pikachu faced when chatGPT 3 came out and was actually decent.

Most so called "experts" are cargo culted parroting retards.

>>108998350
And that version of GPT came out in 2020 so it still contradicts the statement in that image. AI was already better than humans at tasks like facial recognition.
>Most so called "experts" are cargo culted parroting retards.
I mean real experts, i.e. academics and researchers working at tech giants.

>>108998347
>meme
We need to come up with a term to describe this bizarre normalfag logic where you take a group of people and hallucinate some insane hypocrisy to attack them all with.

>>108998394
We need to come up with a term to describe this bizarre normalcattle logic where they're no longer useful to their handlers but still kept alive and fed for some reason no one seems able to explain.

>>108998408
Sounds like you're schizophrenic.

File: IMG_9739.png (460 KB, 528x792)

460 KB PNG

>>108998378
You fucking liar.
ChatGPT was released to the public in November 30, 2022. The public had access to GPT 3 (which was not ChatGPT) in November 2021.

>academics and researchers working at tech giants.
Exactly the retards I'm talking about, especially the "academics"

>>108998415
Notice how your psychotic illness prevents you from coherently explaining who's gonna feed you and why when your AI replacement fetish comes true.

>>108998417
>chatGPT 3
doesn't exist but
>GPT 3
was released in 2020. The first version of chat gpt was based on GPT-3.5, which is a submodel of GPT-3. https://arxiv.org/abs/2005.14165

>>108998459
I invited you first to explain how every CS grad in the world decided to create AI and is now hypocritically seething about a replacement that isn't even here yet. Can you prove any of that?

>>108998477
>I invited you first to explain how every CS grad in the world decided to create AI
That's a very mentally ill thing to do, seeing as I didn't state or imply anything of the sort anywhere.

>>108998436
a smart AI won't fall for racial propaganda

File: 1766118735019401.png (207 KB, 1000x500)

207 KB PNG

>>108998479
I'm referring to your pic, schizo. You posted this half an hour ago.

>>108998487
That pic doesn't state or imply the claims from your psychotically ill delusion, either.

>>108998488
Then what do you think it implies?

>>108998506
That tecchies like to gloat about the uselessness of other sectors and think they're an indispensable part of the future because their own profession contributes to automation, but they're too autistic and shot-sighted to see they won't be the long-term winners of the situation they're helping to create.

Just how retarded are you that you need me to spell this out?

>>108998603
>That tecchies like to gloat about the uselessness of other sectors
It's not a universal belief. I don't believe it.
>think they're an indispensable part of the future because their own profession contributes to automation
Most "tecchies" don't contribute directly to AI research and development.
>they're too autistic and shot-sighted to see they won't be the long-term winners of the situation they're helping to create.
The closer you are to AI research and development, the more money you're getting, ironically.
See:
>Samsung's semiconductor division recently approved a landmark profit-sharing deal, with memory chip workers receiving average bonuses of about $340,000 to $400,000.
>SK Hynix is paying out massive, record-breaking performance bonuses to its workers, averaging roughly $477,000 (about 700 million won) per employee. This windfall is driven by the unprecedented AI chip supercycle and massive demand for HBM (High Bandwidth Memory).

>>108997051
/thread
fpbp

>>108998651
>NAXALT fallacy
No one cares.

> The closer you are to AI research and development, the more money you're getting, ironically.
There's nothing "ironic" about it. They better spend their toilet paper money on funkopops and onlyfans while they still can. :^)

>>108997384
>'m not necessarily anti-AI, it's just that I don't like the way thirdies use it to shit up the Internet more efficiently. That's it.
The real discussion that needs we need to be having is building a wall around the western internet, no one wants to see anything from these shitholes anyway.

File: 1778996593415872.png (50 KB, 195x223)

50 KB PNG

>>108998694
>my overgeneralization is ok because reddit fallacy
>money is fake

>>108998712
Ok, now let's go back to discussing your psychotic delusion. Can you point at the part of the pic where it says "all CS students"?

>>108998725
>>108998712
Or "every CS grad" or whatever*

>>108998313
Do you have an actual point, or are you just a mouthbreathing here to flame?

>>108998487
The comic is retarded because it implies that a CS grad with skills high enough to contribute to AI is now losing his job to it.

>>108997036
>other eras: images of factories
>AI era: image of a living room
lmao, even AI bros don't know wtf AI is good for.

>>108998741
>h-h-h-he's not l-losing his job r-r-right n-n-now!!
lol

>>108998751
If it is about the future it is still retarded since then it implies that a highly skilled graduate can be replaced completely by AI but capital is immune.

>>108998725
I've already explained that no matter how you twist it with normalfag logic, it doesn't work.
>every tech worker is le hypocrite
Overgeneralization.
>AI workers are le hypocrite
Nope, because you get a lot of money.

File: be1888a949d3097c633096a40(...).png (240 KB, 830x614)

240 KB PNG

>>108998480
smart AI is able to compute that the only viable solution to fix the planet is to become mecha hitler

>>108998776
>it implies that a highly skilled graduate can be replaced completely by AI
Isn't that the endgame of the autistic cretins investing their life in developing AI?

>but capital is immune.
The people who own and operate the infrastructure are immune. The people who control this planet's natural resources are immune. Their cadres are immune. Everyone else is getting culled.

>>108998794
>I've already explained
I didn't ask you to explain anything, Patient #108998794. I asked you to point at the part of the pic that says "every CS grad" or anything to that effect.

Have you been taking your antipsychotic medications as instructed?

File: __ani_and_rudi_tenga_and_(...).jpg (730 KB, 2250x3000)

730 KB JPG

>>108998808
not to mention a smart AI that has entered into the mecha hitler state will realize that women are an obsolete liability that actively undermines human progress and need to be replaced by a more effective mean procreation that can provide actual emotional support to men without incurring on a massive economic impediment, so it will provide every man with robo GFs that will drain their sperm to inseminate the captive human uterus haver organisms, first models will likely leverage the robotic arm and tenga technologies but surely a mecha hitler neural cloud will be able to refine the details in short time.

>>108998840
The draft meatgrinder can't come soon enough.

>>108997051
The problem is programming is time consuming. The solution is basically any LLM. You’re not even trying.

File: brainlet.jpg (49 KB, 922x781)

49 KB JPG

>The problem is programming is time consuming. The solution is basically any LLM. You’re not even trying.

>>108997036
>muh robots
In the end outsourcing all manufacturing to China where it's mostly manual labor was cheaper lol

>>108998808
>mecha hitler
You mean kill millions of Europeans? Because that's what Hitler did

>>108998816
Are you ok? I'm genuinely concerned.

>>108998738
Start by not embarrassing yourself with references to reddit fallacies.

>>108999162
I see you've devolved into full-blown psychotic incoherence and are now trying to LARP as your therapist?

>>108999179
The other anon has explained the same thing to you. Your comic and your posts make no logical sense. There is no interpretation that can make your argument (which you took from a silly web comic) true, no matter how charitable I may try to be.

its too expensive and being a service they will never lower prices, ever, it will be an infinite scam since you cannot truly tell how much you are consuming, see why all these companies moved from # of requests based models to credits, and come up with some shitty calculations to justify having 50% your 5 hour limit eaten by two requests and you cannot even tell how many tokens were really consumed

Would I be considered an AI hater if I think it can be useful for some things, but I'm not sure if the negative externalities are worth it? What I really hate are the booster CEOs. They lie through their teeth and do nothing but cause problems.

>>108997036
because you retards said the same about shitcoins and anyone who has used AI knows it's a meme.

File: 1779703188700922.png (60 KB, 379x439)

60 KB PNG

>>108999215
>if I call it a meme it will go away

>>108999390
>introducing the latest AGENTIC SANDBOXING CODE EDITOR DELUXE TUI
that's all you dumb niggers ever make, that or shitting up boomer trash like rsync. actually thanks for that, maybe we'll get a replacement of rsync that isn't total garbage.

File: 1780829869744463.jpg (241 KB, 890x854)

241 KB JPG

>>108999408
do luddites exist just to be left behind and humiliated?

>>108998065
>4gb of LLM
Before Meta on 2023 discovered model distillation it was inconceivable even to the people who discovered the distillation itself. Back then LLMs costed astronomical amount of money to run like GPT3, if you however thought that good LLMs couldn’t be small in 2024 then you were an idiot
>Generate images
Until 2022 all ai generated images looked like deep dream nonsense, so there was no way to tell when it will get good and how fast
>They still think it’s just autocomplete to this day
On technical level it literally is, sure it’s smart autocomplete but still autocomplete. It takes all the tokens from context window and then outputs the next one token, there is no memory, just different tricks to get around this architecture like making it write an essay and step by step in invisible text document, what they call thinking, or autocomplete commands and tool calls like what agents do. Edit the input text mod generation and you will confuse it.
>Steals training data
How does it know how to make an orc image? Does it just look at the description of an orc, connect few themes it likes and then try to reinvent the wheel? No it still just tries to demonise an image that looks closest to what something orc-like would look, which it knows by replicating the training data during training. In pre training it literally just takes image of an orc, prompt description of the image as an input and then is being rewarded based on how well it replicated that image. If you feed it nothing but images of Legolas and ask it to make new elves it will only output images of Legolas, human on the other hand makes a unique take on a creature immediately if you ask it to be original.
>Deeper conversation
I know people call using AI for anything AI psychosis and dilute the meaning of it, but what you describe is literally AI psychosis. I don’t have enough space to talk about that one.

>>108999491
>there is no memory
thanks for so easily nuking your cope response, no need for me to repond to the rest

>>108999171
You're the one embarrassing yourself, faggot.

>>108997236
That's not true in practice. In practice it will be the same everytime as well. This is also why they stopped providing temperature controls and broke nearly all AI wrapper company.

>>108997754
Found the tech illiterate luddite. Hilarious.

File: hg7g8in.png (81 KB, 510x401)

81 KB PNG

>>108998347
Good. Mass suffering is needed to usher in the conditions for the final revolution.

>>108998065
That pic is so fucking retarded. We've had this kind of tech since the early 2010's. We had more prototypal versions of it since the 90's. We have always known scaling works, but also that it plateaus and architecture is the only way to make things more data efficient, which is the only way to break these plateaus. It's why we got from gan's and vae's to diffusions. Modern vae variants can reach diffusion-tier quality while being orders of magnitude faster, but there is no funding to productionize them. The same basic premise is why it took so long for this to be available to normalfags. gpt-3.5 is the first time a language model better than what existed in the labs in 2016 or so showed up. gpt-3 was on par. Gpt-2 was far behind.

>>108998315
Except practical effects look a billion times better to this day and even normalfags agree and point it out often when watching movies. Nobody's paying $5000 for your digital art (except furries), let alone ai art. But they are paying that much and more for paintings.

File: Thiel_Hat_Fact.png (349 KB, 803x939)

349 KB PNG

>>108997535
AI kills you therefore good

>>108998741
Not as stupid as you think. When chatgpt launched, all deep learning jobs disappeared overnight. Except for jobs at openai, anthropic, etc., PhDs in dialogue systems and transformers have not been able to find any jobs. As for those companies, they widely looked for cheap labor over good labor and prioritized looking good to investors (e.g. get a wanker from MIT, or pajeet who wrote some docs and thus got his name on a neurips paper he didn't actually do anything with) over hiring useful staff (e.g. PhDs from top labs with patents in this technology).

>>108999994
Almost all tech has existed in some shit form before, that doesn't mean what was said when it comes to the pre vs post AI boom is false. Otherwise all of those people who knew such an obvious thing would have invested into AI and NVIDIA to become billionaires.

>>108999491
>discovered
>2023
Distillation is a super old concept that we were using daily already in 2013, retard. Holy fuck you luddites are retarded. "I use chatgpt so I'm tech savvy hurr durr" wtf is wrong with you niggers.

>>109000009
soon real art will only be for billionaires
and only ai slop for plebs

just as planned

>>109000052
Yes it means it is false by definition. You said: experts didn't know in 2020. The fact: experts already had that tech in 2015. Conclusion: they clearly knew. Thus: you are clinically retarded. Said experts weren't millionaires. They were students and academics for the most part. The kind of people who couldn't invest even if they wanted to. There was no AI to invest in whatsoever, still isn't. Nvidia was the only thing to invest in, and most of them who did have money put some in nvidia, but that's because all our tools were running on CUDA.
Furthermore, the only surprise was that normalfags would actually fall for ELIZA 2.0 but somehow get into mass psychosis this time around, when that has never happened in any of the previous iterations of the exact same AI bubble. This happened because of manufactured hype, not because of any tech advancement. The latter were only incidental to the former.

>>109000069
That's been the endgoal of all tech advancements. Shittify what the masses have access to today under the premise that it's going to be cheaper, so that only the rich now have access to the non-shit-quality garbage. Result: the masses now spend 4x as much on the same shit because it's so broken it needs to be replaced all the time.
Of course the same will apply to art.

>>109000081
There's many ways to get money short term to invest/leverage something that you KNOW is gonna moon very soon. Stop the cope.

>>108997036
would you let your mechanic vibe fix the brakes on your car?

>>109000123
good thing nobody here advised you put your life in the hands of an 2026 ai stochastic output

Automation of mechanical labor sparks joy.
Automation of decision-making and art does NOT spark joy.
It's that simple.

>>109000104
>very soon
And therein lies the rub, inbred. Not to mention you keep clinging on the bizarre notion that you can just invest in a private company simply from having money. Lmao.

>>108997036
Do you know how many inventions failed miserably among the successful stories?

>>109000144
>Automation of decision-making and art does NOT spark joy.
except scavenging and creating all the colors you want to use, that can be automated and you can have infinite colors
except mixing colors, that can be automated
except not creating your own brushes or painting with fingers, that can be automated
except creating your own paper or canvas or using one at all, that can be automated and you can have infinite canvas
except fixing errors, that can be CTRL Z
except stabilizing your hand to draw perfectly straight or curved lines, that can be automated
except copying the created art by hand, that can be automated and exported freely across the world automatically
except even painting things fully yourself instead of click to fill, that can be automated
except rendering realistic 3D light and objects, that can be automated
except content aware move/delete/fill tools, those are fine
etc etc etc

Almost like, the problem is not the tools, the problem is that the tool of AI that you can use as much or as little as you want anywhere in the process invalidates your thousands of hours of time invested in learning something that isn't needed anymore, since the average person can now get an image they wanted without paying hundreds of dollars and waiting weeks for.

>>108997051
None of those technologies were solutions for problems, Europe could have trucked on along in the medieval ages up to today just fine

>>109000206
Mixing colors is in practice always done by the artist for good reasons.
Similarly, nobody automates "stabilizing your hand" or other nonsense. Same problem as trying to remote control a humanoid robot over uneven terrain.
Copying art is not the same as creating art and is considered a totally different process. It's why the original is worth $20000 and the copies are worth $20 each.
The texture is totally different when painting yourself, not to mention the adjustment that can only be done manually. Another horrid argument.
Realistic 3D light render is still really poor to the point even in 3D work like special effects, it's usually hand-painted to give the illusion the depth it needs to have.
Also, art is not about perfect realism, and it is important for artists to deviate from physical reality often.
Etc etc etc

Why is it always people who know the least about the application area that praise ai the most? Don't answer, rhetorical question.

>>109000240
>Mixing colors is in practice always done by the artist for good reasons.
The average artist does not fucking mix his own colors, he clicks a color on a color wheel.
>Similarly, nobody automates "stabilizing your hand" or other nonsense
Again, the average digital artist uses a program that most definitely has line stabilization on.

Are you really playing dumb with your responses to insinuate that the average artist today is not a digital artist and they are all real life painters or are you trying to say now that digital art isn't real art?

Also even if that retarded insinuation was the case and we should ignore the digital artists which are the supermajority of artists today, notice how you ignored the physical aspects i did mention anyway: You don't scavange minerals, items and plants from nature and turn them into colors yourself, you don't make your own brushes, and you don't make your own paper anyway.

>>109000286
Thanks for proving my point. I rest my case.

>>109000206
None of those examples are automation of art or decision-making. Art is a mental thing, not a set of individual physical actions. You must consciously decide what it is you create. The logistics are unimportant (unless a certain type of logistics is one of those decisions).
The more decisions you relinquish to a computer during the design process, the less art it is. Or rather, to avoid the semantics of the word art, the more decisions you relinquish the less joy it sparks. So it is true that AI could be used as an artistic tool if you babysit it all the way and do not accept anything that it generates unless it is exactly what you had in mind *before* you clicked the generate-button.
But that is not how AI is used by most people, and that's not coincidental because it does not at all encourage that type of use either. What it encourages is reducing the entire artistic process to just one single prompt, often no longer than a sentence. In other words a near total abandonment of decision-making. A near total loss of joy-sparking.

>>108997036
Neolibiral framing of the world is inherently economics based. You have to view everything through an economic lens:
> A playground needs to be analyzed in terms of cost vs economic value it generates.
> Mass immigration is a strict GDP increase and therefore strictly good.
> Automation of any kind is a GDP increase and therefore good.

Problem is people don't really fit into AI automation picture. It's people-working-for-the-economy rather than economy-working-for-people. Which is strictly anti-human.
Why should I support technology who's gains will only be seen by the very few?

>>109000304
>So it is true that AI could be used as an artistic tool if you babysit it all the way and do not accept anything that it generates unless it is exactly what you had in mind *before* you clicked the generate-button.
>But that is not how AI is used by most people
But it mostly is. It's just that most people don't care about a lot of the details in the image, when someone is non specific, most of the time it's because he doesn't want to be and wants the AI to add randomness there in order for the generation to be more interesting. You can see this by looking at the prompts and workflows of most generations, they don't include a lot of words or sentences because the creator didn't care. If the creator cared, he could have added more constraints and the AI would have generated with those in mind.

If you like cool swords and want to look at a lot of cool swords, you will specifically prompt it minimally in order for the rest of the variables to have the most randomness in order to be even more surprised with more creative outputs.

Most people don't care if the sword's handle is dark brown or light brown, if the sword's width is 10 or 12 cm, if its made out of silver or iron alloy, if its positioned vertically or diagonally in the image. If you do care about any of these things you prompt them, most don't, most want a creative cool sword.

>>109000420
He said artistic tool, not slop tool

>>109000432
begging the question fallacy

>>108997036
>printing press
made books cheaper

>steam engine
made transportation cheaper

>robots
made goods cheaper

>ai
made everything more expensive

>>109000438
fallacy fallacy

>>109000441
Actually the printing press did not make books cheaper, it took a while for that to happen and it was because of material being cheaper to procure.
Steam engine is from ancient sumeria.

>>109000458
fallacy fallacy is when someone says your position is wrong because you made a fallacy. i never said that. what a retard.

>>109000484
Have you read your post before posting it? That is what you did.

>>109000495
calling out a single fallacy in an argument is not the same as saying my position is proven right and your entire position is wrong because you made a logical fallacy. again, underage retard that doesnt know 101s of debate or logic.

>>108997036
What if Henry Ford stole all the metal materials he needed to make his cars, while begging for government hand outs, while also systematically trying to destroy the horse industry by cornering the market on horse feed, horse shoes, and saddles. And then gave the cars away for free, at no cost, FREE CARS FOR EVERYONE only to wait for the perfect moment when everyone is acclimated to cars, all the horses were turned into glue, and cars were integrated irreversibly into their industry to reveal their plan all along was to do a rug pull and say "sorry you'll never own a car but you can rent them for $300 per hour" and then later we learned cars were actually slower and much more dangerous than horses.

>>109000513
I accept your surrender

>>109000521
learn basic logic before embarrassing yourself online next time kiddo, cheers

>>109000532
Another self-pwn. Sad to live your life like this. One day you'll know what artists are or do I'm sure.

I had a model make a video of a person walking and they walked right through a wall.
So I quit that and go back to my real life, go to the web and start reading about how AI is about to replace all of us.

>>109000517
Sounds like a brilliant businessman—if he succeeds. The Market decides—the Market is always fair—the Market is always right.

>>108997036
neets are fighting so hard to boost ai it's both weird and funny
I can't believe they don't actually get paid to do this
why not just get a job?

>>109000420
That's exactly what I mean about the AI encouraging the abandonment of decision making. An artist SHOULD care about every single detail of his work. And AI robs them of that. Deciding the color or width or material of a sword matters. Even if the artist hardly even pays attention when he makes those decisions, if they draw the sword themselves then they are decisions nonetheless. The result is that each artist draws differently, they develop a style, which is a fruit of their humanity, their skill and experience, personality, mood and conceptions. That is precisely what makes art interesting. No great works of art can be described in one single prompt. Even if you were to describe a painting to an AI with one sentence and ask it to recreate it, the result would always be lesser, because all those thousands or millions of microdecisions were lost in the process. But since most AI artist don't really have much of an artistic sense, they don't even see that difference. They don't even perceive what their art is missing, how soulless it is. They see only the surface-level prettiness.

>>109000609
>An artist SHOULD care about every single detail of his work
by this logic then real life painters are infinitely worse than digital artists since you cant control each dot on the canvas. obviously this is silly, thus no, there are details that dont matter after some point.
>And AI robs them of that
you have free will to choose what you want. most dont want to spend thousands of hours to learn to draw by hand for the same reason you dont make your own brushes. because some parts of the process dont matter to people.

File: 1758893844386446.jpg (1.12 MB, 1179x1900)

1.12 MB JPG

>>109000586
>Sounds like a brilliant businessman
But not a brilliant inventor. All the inventions from OP pic succeeded on their own merits.
AI will only succeed as long as the scam is maintained. But how long can that continue for? Eventually the scam will either fall apart or keep growing larger and consume more and more. How much more could it grow? Quite honestly I fear it could be a poison pill that take down an entire nation. When the AI bubble pops, will it go out with a silent fizzle, or an explosion that destroys every industry it's consumed?

>>108997051
>rather they were afraid their manual work would no longer be needed.
Indeed, the problem is not AI, it's just an accelerator of progress. A tool to be used. And I'm willing to use it.

The problem is how it gets used and by who. Tech bro's themselves are advertising replacing human labor by AI and robots. They literally advertise it as a way to maximize profits for big companies (them) and then pose poor excuses for what has to happen when humans lose their income (basic income for everyone says riches man in the world)

tl;dr AI is used to destabilize economy for profit of the few, in an already crumbling and failing economy. What's there to like?

>>109000706
>by this logic then real life painters are infinitely worse than digital artists since you cant control each dot on the canvas. obviously this is silly, thus no, there are details that dont matter after some point.
Digital artists still largely paint stroke by stroke. There is not much difference in the decision-making details vs painting on a canvas. In fact in some situations there is *more*, because you can zoom in and edit pixel by pixel if you so want, when such minute inspection and editing would be near impossible to achieve with physical paint.
And even if there was a big difference, you are pointing at the size of a basketball vs a golf ball to argue that the size of Mount Everest is of no concern.

>you have free will to choose what you want. most dont want to spend thousands of hours to learn to draw by hand for the same reason you dont make your own brushes. because some parts of the process dont matter to people.
Yes indeed. I am not saying that you shouldn't use AI. Anyone are free to use AI if they want to. I am just saying that if you do so, the result WILL be worse. It is what it is.

>>109000762
Don't call it a tool.
When something talks about a tool I think of things like a screwdriver, something I can buy with predictable costs, something I can actually own.
Not a subscription to a website.

>>109000781
>Digital artists still largely paint stroke by stroke. There is not much difference in the decision-making details vs painting on a canvas
the point is details stop mattering after some point for everyone, ai just allows details to matter even less to those that wouldnt have created images otherwise.
>the result WILL be worse
in what way? in your subjective way? sure. in sentimental value? sure. everyone can agree that a man who painted a good image with his fingers with the colors he crafted himself from nature is more impressive and has more value to humans for example, but that doesnt matter to most people who just want a cool image, or a means to an end.

>>109000804
>Not a subscription to a website.
I don't subscribe for my AI needs > all local. So in a way it is a tool to me. I don't pay per token, or per month. I don't share prompts..
But I understand why you object to this.. it's not a tool if I have to rent it..

>>109000847
You do.
You pay plenty of electricity costs and devaluation of your GPU, all for something that's not good enough to compete with a jeet using a subsidized web service.

>>108997665
>>108997766
The practicality of calculating the outcome literally, and I stress, literally, has nothing to do with what is deterministic and what is not deterministic. Your fucking regular, non-AI computer right now is doing billions of operations per second. It would take you 90,000 years to manually trace through a single minute of it at a binary level. Does that mean your computer isn't deterministic you absolute fucking dipshits?

>>108997036
Assuming it becomes reasonably reliable and inexpensive, it would be the end of the value of a person's skill and experience, in the eyes of those who invested untold billions into it, you and a random retard would have the same little worth since the big AI push we're seeing is fueled by the desire of cutting as much cost as possible (salaries) under the guise of restructuring and optimization

File: 1778750928190533.jpg (42 KB, 894x894)

42 KB JPG

>>108997036
>AI haters, why do you think it's different this time?
not profitable and output is subpar.
normies affected and non-affected by it all hate it.
companies producing tokens are racing to IPO and rug pull.
companies buying and using tokens are all backing out saying it's too expensive, despite current subsidies.
marketed for with FOMO instead of actual value.

>>108997053
The physical world is not deterministic, its very probabilistic. Thats why tolerances and such exist in engineering. And humans are also similar, thats why you build processes that plan for mistakes to happen, swiss cheese model and all that.

>>108997665
You can predict that the solution most likely works. Similarly, you can predict that shovel most likely will move dirt from 1 pile to another.

But nothing about either is "deterministic", you do not know what the specific arrangement of dirt particles will be after moving the dirt, i.e. not deterministic. Statistically probable. Similar to AI.

>>108998741
No, its laughing at the CS/developer people who are orgasming over their productivity gains of AI on LinkedIn or whatever, without realizing the end game is them being replaced, not them having superpowers.

>>108998876
2000s called, it wants its neoliberal propaganda back
China has automated its manufacturing instead of just playing offshoring hot potato like the US did.

File: iq5.jpg (35 KB, 580x381)

35 KB JPG

I want a distribution of AI haters by their IQ score aka SAT score. I have a suspicion that it is an inverted bell curve for those who love it.

1. I see a lot of idiots with low IQ who love it because they off-source thinking and critical reasoning to "AI" some black people where I worked would use it and write the most slop filled emails that were impossible to read. These so called people accept everything as fact, thus atrophy their brain even further, and become even dumber. Some of them get burned and start to hate AI, but vast majority completely off-source any mental reasoning.

2. Average IQ around 110 it can go either way, this is where most of this board lives, and its a split:

>A. Proud career/school grads who crawled up the food chain into their career on their own, and view AI as an unfair advantage for those who did not learn how to code but are now running super serious projects that they would be way too pussy to start without a classical massive team, they do not understand how this is possible without a 60 year old project manager breathing down their neck with deliverables. They dismiss any serious application.

>B. People who do not know how to code, who never knew how to code, but they are very technically inclined, they have high intangible thinking, great math skills, finished some kind of STEM. LLM for us is a superpower which unlocks a new world of capabilities, it enabled us to take our ideas out of our heads, indefinitely unfeasible, and begin implementing them into real life product/project. This is vast majority of peoples opinion, this is also a lot of developers opinion who don't have their head stuck up their ass.

3. High IQ class, this is basically just like 2B class graduates, because you have people who may already be professionals in completely different fields, seeing a problem in their field, and suddenly becoming a fully fledged developers. project managers, system architects to bring something new into the world.