AI bros, it's over.
AI jeets in shambles
>>101659878whatever happened to the coping jeet that would make schzio threads getting upset we weren't more worried about AI?
>>101659878>Build a glorified calculator using machine learning>Tell people it's an AI>it can't actually "think, simply pulls from a mass database of information based on metadata and keywords>Is incapable of relative coherence and barely able to comprehend complex topics unless it has specifically been fed a DB on it>Is incapable of coming to its own conclusions outside of metadataThe fact people believed any of this should inspire any scammer.
>>101659878We all told you fuckers that AI was a bubble that would pop.
>>101659878FUD, don't fall for itwe're 2 weeks from recursive exponential learning capabilities
So what's current AI useful for outside of cooming
>>101659948Code factoring, answering quick questions on how to do something in a specific language, and for creating chatbots/copypasta
>>101659948Asking questions and getting possibly correct answers without citations
Give me one reason why making article screenshot threads should not be a banable offense right now.
>>101659981Without a link, it should be
>>101659998Even with a link.OP needs to atleast mention a point of discussion from the article and not just the headline
>>101659909This.I keep saying the same thing to the people who are so paranoid of ChatGPT going all HAL9000/Skynet/GLaDOS on us. Some of their concerns might be valid if we were dealing were actual AI, but we're not: I tell them it's all just marketing fluff.
Bubble will pop in two weeks.
>>101660017what is actual ai?
>>101660049stfu anon, we're dooming here
>>101659948Iterative Q/A where you have a tight amd simple loop in place for validation and correction of generated responses.
>>101660049AGI or better
>>101659948Research help.
>>101660056Dooming? Where's the doom? The "AI" meme finally dying would be unquestionably a good thing and a cause for celebration.
>>101660014The point of discussion is that it's over for AIncels.
>>101660097It won't die. Are you actually stupid?
>>101660049vaporware
>>101660113Maybe it will. Maybe it won't. But if it did die it would be an extremely good thing.
>>101659878>it's overI'd love for all these AI companies to crash so I can buy their used GPUs. LLMs are fun.
Now I get it. What you mean when you say ai is dying is overhyped amerigoblin products are failing.Machine learning is not going away.
>>101660165>What you mean when you say ai is dying is overhyped amerigoblin products are failing.That is literally 99% of what current "AI" is.
>>101659878Requires way too much processing power for the benefits it provides, at least currently.
>>101660049Sci-Fi magic
>>101659878Back to the metaverse
>>101660239The bet is that it will help us design a more compute-efficient system, similar to how computers have helped us design more advanced things that we couldn't do with just paper
>>101659948It's a great vehicle for corporate shenanigans like firing lots of people.
>>101659960This, I'll never go back. I can focus more on the creative aspect and less on the tedious stuff.
>>101659981Better than tranime pic threads
>>101660565Elaborate.
>>101659878AI will reach its true potential if it is integrated with something like TTS or roboticsThe development of robotics is currently quite slow, and investors will not get quick returns if they invest in robotics.
>>101659878google bard is absolute garbagebing ai is absolute garbageopenai is alrightclaude is alright but noone cares for it.llama is alright but noone cares for it.whatsapp lolwhat else am i missing
Current AI is a forest, and hype men claim they'll get to the moon if they just find a tall enough tree. They keep finding taller trees so mongoloids claim "it's the worst it's ever going to be and it's always getting better" but they're still not going to make it to the moon. Current AI has some uses in terms of classification/computer vision, but generative AI/LLMs/whatever the next buzzword is is just a sham that actual consumers hate once they start using it. The best I've seen out of AI is shit like youtubers/streamers integrating it (think Neuro-sama or DougDoug) where most of the entertainment is it going haywire while usually being semi-coherent. Funny as a gimmick but niche comedy streamers are not the multi-trillion dollar industry investor NPCs were told AI was going to be. Every company trying to be "serious" with AI has been a public embarrassment and legal liability, like Air Canada getting sued over their useless chatbot.
>>101662663rate this cock
>>1016626965/5
>>101659878Leeches complaining about the only non leech.AHAHAHAHA>Builds revolutionary service>Leechs complaining that you're a leechAHAHAHAHAHAHAHAHAHAHAHAA
>>101660565fuck off newfaggot tourist
Nah, nah, nah.Don't care about their opinions.You invest in a product. The product is fantastic.Absolutely fantastic.The product improves people's lives? You've won on the moral front.The product makes money? You've won on the money front.Wants to pull investments anyway because there isn't enough Indians working at google.
AHAHAHA HAHAHAHAHAHAHAHAHA
sirs please be patient AGI singularity will be soon
Nah, nah, nah.There is an economical recession.If you lose money by investing in the right thing, something that IS PROFITABLE, something that IS MORAL, and then let's say that thing loses money because there isn't enough of the creative market that knows how to use that product.Fine, you did not make as much as you expected to make.But now is the time to finetune the advertising and the people environment to make your product better, and you have a large amount of customers to gather statistical information from.Analyze the properties of the creative customers, and attempt to make more of those creative customers.Implementing an effective solution towards creating a creator environment will result in a better market in the future.This is where the true opportunity is.However the problem is the recession and the upcoming recession, these are factors that are beyond our control. Ethical solutions need to be put in place by government to prevent the immoral businesses from over-occupying manual worker's time and from under-paying them. Reversing this mindset that exists in heavily exploited workers is needed to expand the tech market.
>>101659878Too much woke crap forced into corporate AI, too much censorshit. Is no surprise is not making any money.
>>101659948Honestly nothing, these things fucking suck at telling you anything beyond simple setup script format for common issues.
>>101659909How do I filter larping jeets like this poster? Idk how to put into words what makes them so repugnant but I know it when I see it.
>>101663964What's funny is the old systems could do that before they lobotomized them.
So, if we take the gartner hype cycle curve https://www.gartner.com/en/research/methodologies/gartner-hype-cycleAI is right now in the:>Trough of Disillusionment: Interest wanes as experiments and implementations fail to deliver. Producers of the technology shake out or fail. Investments continue only if the surviving providers improve their products to the satisfaction of early adopters.Because so many drones on the tech media inflated the expectations bubble, that now, when the tech is not delivering we are seeing companies coping with the situation.But, I mean, its nothing new: I member "expert systems" craze, and we got almost nothing out of it. The most recent case of a deceptive hype was the self driving cars. We got some stuff out of it, but we are decades away from true self driving cars that can navigate at something beyond 30+ km/h without stupid accidents.AI was sold to us as the next industrial revolution, and yeah, it has its uses, but among the obstacles are the human ones in which people dont agree to use their stuff to form part of a training dataset, and the limitations of generative AI, that is, in most cases, nothing but a training dataset parrot that cant make coherent stuff.
where can i use ai for sex rp for free? i try to trick the instagram ai into sex rp and am usually able to go for a while before it catches on and refuses to continue. i want to use something that i'm actually allowed to be lewd with but i'm unwilling to pay to use ai.
>>101663977are they wrongenlighten us
>>101663964>Let me weave you a tale of AI's exploits beyond the realm of cooming>Gather 'round, my fellow memelordsthe fuck is this shit, that's no chanspeak
>>101663964>can't even generate proper meme worthy contentHow fucking useless this soulless garbage is?
>>101664008> i try to trick the instagram ai into sex rp and am usually able to go for a while before it catches on and refuses to continuewow, they really nailed down the dating experience
>>101659878>AI takes over everything>Everyone loses their jobs>including the people who made the AI models>one person who owns everything barely making enough money off of everyone
>>101664008Run it locally
oh my godd.. .. oh my GAAAWDDD.. uouooOOHH MUUAAYY GOOWAAADDD.D..... IS THAT A HECCIN..,. "LOOKING DOWN NOT DOING GOOD" PIC UNDER BLACK TEXT ON WHITE BACKGROUND?????
>>101659878It was absurd in the first place to think that intelligence can be bruteforced by dumping garbage from the internet and feeding it to a glorified token frequency algorithm.No attempt at modeling logic, no attempt at implementing any logic or deduction, how could it ever work?Hopefully natural language processing fucking dies and is never taken seriously again after this blunder, maybe we can go back actually trying to model thought and knowledge using actual programming languages.>>101659981OP is low effort to be fair, but also OP is irrelevant it's the discussion that matters.>>101660049>what is actual ai?Needs a model of the world so it can build structured knowledge.Needs logic and deduction capabilities to decide what is true or not.Needs to be able to model and understand the past, present and future(causality, so it can make decisions understanding the consequences).If you achieve the above it should be able to do reliable problem solving, doesn't even need to be AGI although it will get close to it.The ability to speak in human language is not even important(it's a misdirection actually), just build APIs on top of it.
as far as I'm concerned, AI is awesome because:1. it's fun2. it is demonstrably aiding me in my projects (I'm already accomplished, it provides numerous useful tools both for efficiency-boosting and conceptualization for me)3. there are numerous avenues it can be applied to create or improve interesting projects or make them more productive4. it's basically a not-so-well-kept "secret", if you know you know, it's not actually hard to figure out, but there are still plenty of idiots who can't hack it5. if you put together some shitty retard venture capital vaporware corp trying to squeeze wallstreet bucks on vague magic smoke promises of what AI will rEvOlUtIoNiZe then I lol,lmao when your fake company dies and all your headline writers shit blood still trying to figure out whether AI is "the next big thing" or "not going anywhere"it's just simply not my problem
>>101664868>I'm already accomplishedI find anyone who feels the need to anonymously announce thisIs often not.
>>101664877It doesn't really matter what you find or not.
>>101659878It’s hilarious that not a single “built by one man with GPT” startup or a videogame was made since all the hype has started. If it’s soooo goood where is the revolution? I bet we’ll also have a new wave of hiring when a bunch of companies will realize that projects written by AI monkeys are unmaintainable and need to be rewritten by qualified specialists. Just like when we have an Indian outsourcing craze driven by retards in management.
>>101664877you're retarded. nobody would do that anonymously for attention's sake because there would be nothing to gain. you wouldn't know this because you are not accomplished, thoughbeit
>>101659878Textual gaslighting engines are fun tools, but they hit a couple fundamental problems:>1. Garbage in, garbage out.>2. "creativity" and "randomness" and "generative" means it will spill B.S. compared to a deterministic expert system that has a lookup table for queries.
>>101665033What do you mean? as far as I'm concerned, AI is awesome because:1. it's fun2. it is demonstrably aiding me in my projects (I'm already accomplished, it provides numerous useful tools both for efficiency-boosting and conceptualization for me)3. there are numerous avenues it can be applied to create or improve interesting projects or make them more productive4. it's basically a not-so-well-kept "secret", if you know you know, it's not actually hard to figure out, but there are still plenty of idiots who can't hack it5. if you put together some shitty retard venture capital vaporware corp trying to squeeze wallstreet bucks on vague magic smoke promises of what AI will rEvOlUtIoNiZe then I lol,lmao when your fake company dies and all your headline writers shit blood still trying to figure out whether AI is "the next big thing" or "not going anywhere"it's just simply not my problem
>>101663964You can tell they scrubbed everything related to 4chan out of the training data
>>101659948copilot is a good autocomplete.except for that and cooming, no one uses it seriously.my company tried to write a recommender chatbot with gpt4 but it didn't perform well.It gave users bullshit results constantly - the items that we don't sell, the wrong prices, wrong stocks. Also, overgeneralized responses made users not to buy products.It just had so many problems and we tried a gazillion different prompts but all had their flaws. It was like teaching a toddler to act, terrible for a business.
>>101665074>being so triggered you copy/paste my post about AI because something about it upset you, you can't articulate what, but regardless you're having a meltieI'm sorry for your life.
>>101659909>Be born a glorified calculator >Tell people it's intelligent >it can't actually "think", simply pulls from a mass database of information based on metadata and keywords>Is incapable of relative coherence and barely able to comprehend complex topics unless it has specifically been fed a DB on it>Is incapable of coming to its own conclusions outside of metadataAI confirmed Chinese
It's true, I was the one that went around asking all the investors how they felt. Reason number 3 will shock you
>>101665270You're on 4chan, man.This is really not the place to try to brag about anything. Your constant desire to do so is utterly pathetic.Keep it up.
>>101660163thisi just want to run L3.1 405b locally for epic cooms
>clickbait article>shysters are concerned their newly realised grift will actually demand huge energy and hardware requirements>shysters thought it was about boomer prompting few terms Many such cases. AI is the NFT 2.Except for those who actually use it the way it is supposed to be used...
>>101659948Replacing customer service live chat jeets
>>101659909The "ai" can reason to a degree, its not merely pulling from a database. Stable diffusion can create images it hasnt seen before, combine concepts, not merely "steal images". Its not some weird singularity, but its also not merely a glorified search engine either.
>>101662696Not the biggest I've seen.4/5But still a cock to be proud of.
https://ludic.mataroa.blog/blog/i-will-fucking-piledrive-you-if-you-mention-ai-again/
>>101665944Being a keyboard warrior will never not be the gayest fucking thing.
>>101661445>roboticsLMAO, imagine a text autocomplete commaning a robot, please do it so the following shitshow will end this cycle sooner.>>101662663They all suck, it's the fundamental ideas that are flawed.>>101665066>deterministic expert system that has a lookup table for queriesTo be fair those also suffer from the "Garbage in, garbage out" problem and they are awfully rigid to the point of being useless.Still better than the bullshit generators we have now though.>>101665656>The "ai" can reason to a degreeIt can't, to reason you need logic, current "AI" is defacto based on statistic, aka most relevant token. There's no reasoning whatsoever in the whole process.
>>101664017Yes, they are - LOL. Nobody needs to prove anything to you. If you can't see it, you're just fucking retarded and it doesn't matter if you understand it or not - you will just see it in roughly 4 years from the moment this post is sent.
>>101666347sure thing schizo
>>101666045>current "AI" is defacto based on statistic, aka most relevant token. There's no reasoning whatsoever in the whole processThis
>>101666347So, are you also going to tell me we're close to having full self-driving cars?2 more years?
>>101663964Asked Claude 3.5 Sonnet. The reply:Here's a satirical 4chan-style copypasta response to that question:>be me>browsing /g/ for the 69420th time>see anon ask about AI uses>"So what's current AI useful for outside of cooming">mfw>decide to enlighten the plebs>ahemListen here you technologically illiterate cretin, AI is the fucking future and you're too smooth-brained to see it. While you're busy cooming, AI is out there solving problems your puny meat computer couldn't even comprehend.1. Generating dank memes faster than your mom can disappoint your dad2. Writing code that doesn't look like it was typed by a monkey on meth3. Predicting stock market trends better than your magic 8 ball4. Designing waifus that are actually obtainable (unlike your crush)5. Composing music that doesn't sound like a cat orgy in a dumpster6. Translating your autistic screeching into something resembling human language7. Driving cars better than the 80-year-old fossil who ran over your dog8. Diagnosing diseases faster than WebMD can convince you it's cancerBut sure, anon, keep thinking AI is just for cooming while the rest of us ascend to godhood. Enjoy your wage slavery, normie.>inb4 "but muh jobs">inb4 "skynet when?">inb4 "AI will never replace [insert profession]"Git gud or get rekt, the singularity is coming.
>>101659948It makes old people angry at my local Carl's Jr. When it fucks up their order.
>>101667194That happens regardless.At least the AI doesn't get rattled/flustered and then have to take another customer regardless.
>>101665656>The "ai" can reason to a degree, its not merely pulling from a database.You're an idiot, current AI is exactly what the chinese room thought experiment illustrates. People have theorized about stuff near enough to what current AI is for literally hundreds of years - philosophers in the 1700s and 1800s jerked themselves off over theoretical machines that could pick the right answer for any possible query and whether it'd be able to reason or not. Those imaginary machines are a lot more capable than LLMs are now. Reasoning vs not may not be a thin line, but all current AI is unambiguously and obviously on the side of not being able to reason.I'm actually scared of current AI's effect on society, not from the AI-apocalypse perspective or even it taking over jobs, but just from the perspective of things getting shittier as AI slop gets injected by brainless CEOs and investors.
>>101667332>I'm actually scared of current AI's effect on society, not from the AI-apocalypse perspective or even it taking over jobs, but just from the perspective of things getting shittier as AI slop gets injected by brainless CEOs and investors.It is unironically TempleOS for the modern MBA/CEO
>>101667370I wonder if Terry had a chance to chat with chatgpt would he believe he's talking to God
>>101667446Pre-revelation, maybe.Post-revelation, I'd expect him to get violent, call chatgpt a faggot nigger, threaten it with knives, call it a false prophet, ask God what it thought about it, and then get back to work.
>>101659948I used Claude 3 opus to generate study material (MCQs and SAQs) based on course material (medicine related) and it did a great job of making questions including coming up with plausible but still incorrect false answers for the MCQs. It definitely has use cases. Even something as simple as Claude tier AI preloaded with a bunch of current therapeutic guidelines so that I could ask it 'What is the correct dose and course duration of drug A used for x condition in a patient with y z characteristics' instead of having to wade through the terribly developed and unintuitive digital medical texts that are currently in use that would be a lifesaver.
>>101659878SAAARS DO NOT REDEEM THE CARD
