AI BUBBLE IS ANOUT TO BURST

AI BUBBLE IS ANOUT TO BURST
Micro$oft AI engineer says "a lot of tasks are about "trying to create Al hype" with no practical use."
FRICK YOU NVIDIA AND M$

Thalidomide Vintage Ad Shirt $22.14

DMT Has Friends For Me Shirt $21.68

Thalidomide Vintage Ad Shirt $22.14

  1. 2 weeks ago
    Anonymous

    microsoft doesn't have any substantive ai shit other than just piggybacking GPT

    • 2 weeks ago
      Anonymous

      Microsoft is the Undi of OAI.

      • 2 weeks ago
        Anonymous

        Someone explain to me this "undi" obsession. Any time he posts literally anything, fifteen people will respond, bashing him for whatever. Is it just one deranged schizo or is there some other shit I missed?

    • 2 weeks ago
      Anonymous

      The AI bubble was invented to prop up the NVDA/Bitcoin stock bubble.

      [...]
      And the end result is something that's probably worse than a quick Google search for anyone who's half decent with a computer.

      Here's what happens when you try and make a product out of AI:

      https://www.msn.com/en-us/lifestyle/shopping/the-rabbit-r1-is-one-of-the-worst-gadgets-i-ve-ever-reviewed/ar-AA1o5nKZ
      >At the end of the day, it’s not doing anything more than what my phone is already capable of. If anything, it does a fraction of the same things my phone does, but objectively worse.

      This is nothing new. It's pretty obvious that there was no real AI advancements after GPT architecture and stable diffusion. They have only been adding more resources and put it behind different interfaces, with diminishing returns.
      Big LLMs will eventually be rendered unprofitable and the hype will die. But the research will continue in the shadows.

      AI is a lot more than ChatGPT and dodgy consumer tech gadgets. You just don't know what you're talking about.

      • 2 weeks ago
        Anonymous

        >AI is a lot more than chatgpt and dodgy consumer tech gadgets
        None of the uses you have in mind have anything to do with the current wave of investments. It's literally all ChatGPT.

      • 2 weeks ago
        Anonymous

        >AI is a lot more than chatgpt and dodgy consumer tech gadgets
        AI begins and ends with chatGPT, dodgy tech support, and lets' not forget to mention mountains of uncanny looking porn and silly memes. AI is a solution looking for a problem, and in many cases, a poor substitute for a working brain.

  2. 2 weeks ago
    Anonymous

    >two weeks

  3. 2 weeks ago
    Anonymous

    AI vill take over and you vill like it!

  4. 2 weeks ago
    Anonymous

    The AI bubble feels about as artificial and hyped as that NFT scam bubble. All we need now is celebrity endorsements of AI like they did with the NFT scams.

    • 2 weeks ago
      Anonymous

      why would you need a celebrity to endorse it when you could just use ai to create a celebrity endorsement

    • 2 weeks ago
      Anonymous

      The AI bubble was invented to prop up the NVDA/Bitcoin stock bubble.

      >talk to a computer
      >computer understands you and talks back
      >"hype"
      yeah, people will forget about ai in two years and the bubbles gonna pop and everyone who invested is going to lose all of their money
      just like all of the idiots who invested in google and amazon

      And the end result is something that's probably worse than a quick Google search for anyone who's half decent with a computer.

      Here's what happens when you try and make a product out of AI:

      https://www.msn.com/en-us/lifestyle/shopping/the-rabbit-r1-is-one-of-the-worst-gadgets-i-ve-ever-reviewed/ar-AA1o5nKZ
      >At the end of the day, it’s not doing anything more than what my phone is already capable of. If anything, it does a fraction of the same things my phone does, but objectively worse.

      • 2 weeks ago
        Anonymous

        >Here's what happens when you try and make a product out of AI:
        Imagine conflating some teeny little startup with one product the entire industry which is going to radically reshape our world

      • 2 weeks ago
        Anonymous

        >The problem here is that the R1 is wrong enough times that I simply don’t trust it about anything at this point. Like so many AI gadgets/services, the R1 confidently tells you an answer to a question, even when it could be completely wrong. Not only is that troublesome for a world that’s already struggling with how to handle misinformation, but on a much more basic level, it also means the R1 isn’t a helpful assistant. If I have to double-check its answers to my questions with a Google search on my phone, why not use my phone in the first place and save myself some time?
        AI in a nutshell.

    • 2 weeks ago
      Anonymous

      Imo its closer to the internet bubble. Sure Pet.com was a total nothinburger but Amazon & Google emerged out of that internet bubble

      AI already started eating up jobs of Content writers, artists, customer support

      OpenAI is gonna release next generation of their AI this year and they are already working on the succeeding generation after that

    • 2 weeks ago
      Anonymous

      Mmmm no-no-no!
      GPT-CONNNNEEEECT!!

  5. 2 weeks ago
    Anonymous

    I think I read in this board that they are getting kinda worried at MS because clients are starting to realize "AI" is a scam
    time to short the shit out of NVIDIA, MS and others?

  6. 2 weeks ago
    Anonymous

    Current state of AI is pretty much like the dot com bubble, its extremely overhyped and it will crash hard very soon only to return 10 years from now with real progress and becoming a pivotal in every human activity, much like what happened with the internet

    • 2 weeks ago
      Anonymous

      Basically this. It's dotcom bubble 2.0
      The tech is legit
      It's legitimately revolutionary
      But there's sweet frick all for investors to expect in return from it.
      Nobody is going to pay money for silly cupcake recipes and fake Seinfeld scripts written by AI. Nobody. And yet that's the shit normies actually use it for.
      Maybe you save hundreds of millions of dollars in closed tech sector jobs but you paid hundreds of billions to do it. There's no roi from that perspective in a thousand human lifetimes either.

      Investors are moronic. But whatever I'm having fun with it.

  7. 2 weeks ago
    Anonymous

    >coproration prioritizes appeasing investors over solving problems for end users
    in other news, water is wet

  8. 2 weeks ago
    Anonymous

    >talk to a computer
    >computer understands you and talks back
    >"hype"
    yeah, people will forget about ai in two years and the bubbles gonna pop and everyone who invested is going to lose all of their money
    just like all of the idiots who invested in google and amazon

    • 2 weeks ago
      Anonymous

      bait

    • 2 weeks ago
      Anonymous

      uh oh

      • 2 weeks ago
        Anonymous

        ya thats called saturation. 1 in 3 people has one != its le dead, moron. find a similar chart for the first 20 years since fridges were introduced

    • 2 weeks ago
      Anonymous

      > understands

      This right away shows me you're moronic

    • 2 weeks ago
      Anonymous

      >bro oculus rift is going to change the world

      • 2 weeks ago
        Anonymous

        Yeah but I never said that though
        Never really heard anyone say that honestly

        • 2 weeks ago
          Anonymous

          Same people coping about AI said that though lmao

    • 2 weeks ago
      Anonymous

      >computer understands you and talks back
      im not here to argue the philosophy of this but how does that by itself raise your stock to justify the outside investment

      • 2 weeks ago
        Anonymous

        Because it improves productivity at a fraction of the cost for human labor

        • 2 weeks ago
          Anonymous
    • 2 weeks ago
      Anonymous

      Don't uderstand why you're trying to convince these sheep.
      They will bury their heads and pretend nothing changed until they literally cant anymore then they'll go all in and pretend they were always for ML. They're cowards to their core.

      • 2 weeks ago
        Anonymous

        Yeah, a lot of things have changed. We now have GPUs that can do 8bit and 4bit calculations. ROFL

      • 2 weeks ago
        Anonymous

        The copium on BOT at first surprised me. But then I realized BOT is no different than rest of the people AI has already replaced. First they claim AI cannot do what they do. Then once it does they claim it is inferior to what they do

        It is pretty hard to accept anything that is going to disrupt your entire life

        • 2 weeks ago
          Anonymous

          >It is pretty hard to accept anything that is going to disrupt your entire life
          what has it disrupted?

          • 2 weeks ago
            Anonymous

            Visual art, text content geenration, Responding to customer queries

            It has not replaced them a 100% yet but it sure is on path to do so.

            • 2 weeks ago
              Anonymous

              bots these days don't seem to have enough memory to remember what was said more than one post ago it seems

              • 2 weeks ago
                Anonymous

                For now

              • 2 weeks ago
                Anonymous

                well he either lacks memory of context awareness

          • 2 weeks ago
            Anonymous

            Content mills and customer service. It disproportionately affects Indians which is why they keep coming here claiming that everyone is being replaced.

    • 2 weeks ago
      Anonymous

      understands you and talks back
      Technically it doest understand but its able to predict what the response should be. That IS a huge advancement in tech.

      I witnessed one of the lead devs on a important product team use GPT to turn a 40 minute phone call to onboard a patient where a form had to manually filled into a 5 minute call that auto fills the form from the context of the phone conversation. Literally saved the company millions in productivity

      • 2 weeks ago
        Anonymous

        >person randomly says field matching term because human was rambling
        >wrong information is put in
        how to prevent?

      • 2 weeks ago
        Anonymous

        >automated data entry saves millions to companies
        ... isn't this exactly what AI has been supposed to solve for decades, yet as

        >person randomly says field matching term because human was rambling
        >wrong information is put in
        how to prevent?

        says, shit is still too unreliable to actually fix anything?

        • 2 weeks ago
          Anonymous

          It seems to be doing all the heavy lifting though. Even if it messes up the human involved can fix it immediately in a few minutes with minor edits. But the AI rarely fricks up cause the nature of the form filling is predictable for the AI. Its too tedious for the human to do and are likely to make more mistakes than the AI.

          So far havent heard any complaints

  9. 2 weeks ago
    Anonymous

    >corporations trying to hype it up further means it's a bubble
    Kek.

    Also what does he even mean will less efficient? ML units on ARM SOCS are way more efficient than CPUs and GPUs, despite the software side being very half-assed. Desktop gays are as usual a few generations behind.

    • 2 weeks ago
      Anonymous

      >ML units on ARM SOCS are way more efficient
      efficient at doing ML operations, that doesn't mean the way they can be used to solve the problem is efficient

      • 2 weeks ago
        Anonymous

        I mean, if you can translate an otherwise tasking problem to ML, it sounds like it'd be more efficient in the end. It obviously doesn't make sense with everything but shit is pretty underexplored, so seems given there will be meme attempts too that do nothing.

  10. 2 weeks ago
    Anonymous

    complainers. sounds like they need to be replaced with AI

  11. 2 weeks ago
    Anonymous

    >they say that much of their work is assigned to appease investors rather than to solve problems for end users
    So, the same as 99% of big tech?

    • 2 weeks ago
      Anonymous

      Good post

    • 2 weeks ago
      Anonymous

      and vidya, and movies ect ect

  12. 2 weeks ago
    Anonymous

    Pretty much every big technology goes through the hype/bust cycle. It's what happens afterward that makes or breaks it.

  13. 2 weeks ago
    Anonymous

    The entire point of AI is to have AI solve problems, I don't understand why would an engineer even bother coming up with solution to problems on his own (instead of training the AI to solve it)

    • 2 weeks ago
      Anonymous

      >I don't understand
      Sums up your post.

  14. 2 weeks ago
    Anonymous

    engineers in the trenches simply do not have the perspective that is needed to make sense of decisions like these. they need to put their faith in the vision of nadella sir.

  15. 2 weeks ago
    Anonymous

    This is nothing new. It's pretty obvious that there was no real AI advancements after GPT architecture and stable diffusion. They have only been adding more resources and put it behind different interfaces, with diminishing returns.
    Big LLMs will eventually be rendered unprofitable and the hype will die. But the research will continue in the shadows.

    • 2 weeks ago
      Anonymous

      >Big LLMs will eventually be rendered unprofitable
      So long as it's capable of astroturfing social media with advertisements and propaganda, it will always have customers with deep pockets.

      • 2 weeks ago
        Anonymous

        Ah yeah, you are right. What I meant is the current trend of continuous growth and scaling up models is going to reach a deadend. Of course what we have now is going to be adapted more and more, but putting in more billions of dollars is not going to keep giving proportional returns.

      • 2 weeks ago
        Anonymous

        If the internet is perceived to be having more bots than humans aka dead internet then people arent gonna be on the internet. This means there wont be a point in paying for bots

        • 2 weeks ago
          Anonymous

          You're still here so that's wrong. The internet is already where poor indians go to hang out with whites and shit on the spaces and a common place pedophiles look for children.

          • 2 weeks ago
            Anonymous

            The irony is that I posted this

            If the internet is perceived to be having more bots than humans aka dead internet then people arent gonna be on the internet. This means there wont be a point in paying for bots

            and I am a jeet

  16. 2 weeks ago
    Anonymous

    >ai engineer neglects his work then does something unrelated and complaints that ms sidelined his solution

  17. 2 weeks ago
    Anonymous

    i'm dealing with the same thing basically. the problem is NOT the ai, it's how big corpos want to use the ai (basically useless chatbots for everything), instead of actually innovating

    gpt5/6 will kill all the consultancy agencies and tbh i can't wait

  18. 2 weeks ago
    Anonymous

    >it's another poorgays who missed out on the bullrun promise it will all come crumbling down any day now episode
    kek never change

    • 2 weeks ago
      Anonymous

      >t.

  19. 2 weeks ago
    Anonymous

    AI is for porn.

    • 2 weeks ago
      Anonymous

      AI is for porn! AI is for porn!
      Why do you think AI was born?
      Porn! Porn! Porn!

    • 2 weeks ago
      Anonymous

      artifacts are too scary and there's no automated way to solve them

  20. 2 weeks ago
    Anonymous

    companies aren't here to make your life easier, they're here to make money for their shareholders. even you fa/g/gots should know this by now.

    • 2 weeks ago
      Anonymous

      >they're here to make money for their shareholders
      The customer is always right.
      This is business 101.

      • 2 weeks ago
        Anonymous

        shareholders are the customers

        • 2 weeks ago
          Anonymous

          companies aren't here to make your life easier, they're here to make money for their shareholders. even you fa/g/gots should know this by now.

          >they're here to make money for their shareholders
          You can't make money for shareholders without selling a real product to real customers.
          If the shareholders are the customer, then what you have is a Ponzi scheme, not a Business.

  21. 2 weeks ago
    Anonymous

    AI is such a buzz word. It's hilarious that people are so terrified of it. It ain't shit yet

    • 2 weeks ago
      Anonymous

      That headline still gets a chuckle out of me and probably always will.

  22. 2 weeks ago
    Anonymous

    Invest in WHAT? WHAT does it do, exactly? All the images look like shit, deepfakes were literally more popular in 2016 and the the only use for the text generator is to weed out losers in universities. PLEASE ENLIGHTEN ME, WHAT DOES IT ACTUALLY DO? Call me a based moron if you have to, I just don't get it.

    • 2 weeks ago
      Anonymous

      You just don't get it man! In 2 more weeks the singularity will happen!

    • 2 weeks ago
      Anonymous

      >WHAT DOES IT ACTUALLY DO
      replace workers witch cuts costs which in turn provides money

      • 2 weeks ago
        Anonymous

        How many workers has "muh AI" replaced so far?

        • 2 weeks ago
          Anonymous

          i dunno but probably they project enough of a boost to productivity to warrant such massive investments, hence the false advertising to pull in investors
          it doesn't help that "ai" is such a vague term also

          • 2 weeks ago
            Anonymous

            Or maybe they are hyping up AI to pocket some easy cash. OpenAI software engineers make $500-900K a year which is 2-3x of what they'd make without the AI hype.

        • 2 weeks ago
          Anonymous

          i dunno but probably they project enough of a boost to productivity to warrant such massive investments, hence the false advertising to pull in investors
          it doesn't help that "ai" is such a vague term also

          also afaik all the ai development is backed by giant companies that can afford to funnel in boatloads of cash into it so it doesn't "have" to "do" anything right now as the ai companies are still able to get investments by ways of marketing to the tech giants
          and another point consumer products aren't really relevant when governments will pay you $$$ to license your various spyware bots whose pattern matching capacity just got a whole lot better

        • 2 weeks ago
          Anonymous

          Definitely replaced tons of customer care, artists and content wrters

          Need to see what capabilities next generation of AI have that will out compete human skills

          • 2 weeks ago
            Anonymous

            yeah dude now when i google fish and chips recipe instead of some loser company trying to clickbait me so they can show 30 video ads with some mom's actual recipe, now they do it with some chatgpt text wall

          • 2 weeks ago
            Anonymous

            AI writes instructions and articles like someone trying to pull an April Fool's prank. I think it will do good for people's skepticism skills, but it's fricking horrible and someone will get seriously hurt by AI making shit up.

            • 2 weeks ago
              Anonymous

              Its still in early stages now. Need to see next gen AI capabilities that will be released this year

              • 2 weeks ago
                Anonymous

                LLMs are just going to just become better and better at making shit up.

              • 2 weeks ago
                Anonymous

                LLMs are just going to just become better and better at making shit up.

                Yeah. Deep learning scales well when you add more parameters AND more data. But what more data is there to slurp? All the clearweb has already been used to train the current gen. There are a few "gold mines" left out like discord and plebbit I suppose but it's not gonna turn ChatGPT into Einstein.
                I don't see a way to improve without a serious human effort in semi-supervised training. But that would involve really skilled reviewers doing real research, not an army of poojeet clicking on the most convincing answer out of three.

              • 2 weeks ago
                Anonymous

                We have to hire legions of writers to write more data for the AI overlords. They'll farm humans just to generate more data.

              • 2 weeks ago
                Anonymous

                Guess we finally found a believable premise for the Matrix to replace the "human batteries" stuff.

              • 2 weeks ago
                Anonymous

                It also explains why they made people live in the world before the internet was in everyone's pocket. Frick, it makes an insane amount of sense.

              • 2 weeks ago
                Anonymous

                Synthetic data is the next logical step

              • 2 weeks ago
                Anonymous

                >just loop the shit pipeline on itself bro, it's gonna clear itself eventually bro, I swear

              • 2 weeks ago
                Anonymous

                >"Pray, Mr. Babbage, if you put into the machine shitty, hallucinated data, will the right answers come out?"

          • 2 weeks ago
            Anonymous

            >customer care
            Wrong.
            Most of the genuine problems that need customer support involve account modifications.
            AI will never be given authorization to actually do anything, because the risk of a jailbroken phishing attack is an extreme liability.
            So when a customer actually needs service, they're stuck begging for a human for 10 hours instead of 2 now.
            The only thing LLMs might help is with moronic boomers endlessly calling the help desk asking how to open PDF.
            However, the better solution there, is to instead not have customer service at all. $0 is cheaper than paying for ChatGPT.

            • 2 weeks ago
              Anonymous

              LLMs need to be leagues better than they are currently to serve that purpose. They're extremely unhelpful and just makes boomers more angry.

            • 2 weeks ago
              Anonymous

              So did you get your Tahoe or are you just a luddite moron who will never be a woman?

              • 2 weeks ago
                Anonymous

                >no argument, resorts to namecalling

              • 2 weeks ago
                Anonymous

                A 5 year old Black person would have the reading comprehension to figure out what my argument was there. You're below even that.

              • 2 weeks ago
                Anonymous

                >still no argument

            • 2 weeks ago
              Anonymous

              All bots do is tell you stuff on the website you already knew. You're absolutely right, it'll never happen.

          • 2 weeks ago
            Anonymous

            >Definitely replaced tons of customer care, artists and content wrters
            so no proof then

    • 2 weeks ago
      Anonymous

      >no response
      yeah I'm thinking it's a bubble

    • 2 weeks ago
      Anonymous

      >Invest in WHAT? WHAT does it do, exactly? All the images look like shit,
      The images are made to look shit on purpose for the end consumer. They have capabilities in creating photo realistic images. The only real hurdle is regulatory here

      >deepfakes were literally more popular in 2016
      AI has made it muvh more streamlined and lowered the skill barrier needed for it. The biggest hurdle here again is regulatory

      >and the the only use for the text generator is to weed out losers in universities. PLEASE ENLIGHTEN ME, WHAT DOES IT ACTUALLY DO? Call me a based moron if you have to, I just don't get it.
      You do realize there are whole sectors of writers that have now become obsolete to AI text generators

      • 2 weeks ago
        Anonymous

        >You do realize there are whole sectors of writers that have now become obsolete to AI text generators
        Yes. Much media writing is shit garbage. Not surprising, I bet computer generated words are even superior in many cases.

  23. 2 weeks ago
    Anonymous

    AI will remove all software engineering jobs in 2-3 years from now lol, don't cry after zillenians when you fail to pay for your mortgage cuz no one will hire your ass, learn to weld

  24. 2 weeks ago
    Anonymous

    AI is the future but it requires near manual algorithmic programming
    which means we won't see it in our lifetime
    the products currently labeled as AI are just scattershot approaches too imprecise to be useful
    will the bubble pop? maybe
    money flows are based on psychology, not on fundamentals
    I think people underestimate the number of 90 IQ oil rich sheiks with hundreds of billions to "invest" in AI
    they can keep this going for a long time, there's no real incentive for current baggies to sell

  25. 2 weeks ago
    Anonymous

    AI is dead on arrival for public facing use and will only continue in specialized sciences and appeasing ultra lonely gooners with fake girlfriends, but they've been doing that for years now.

    You cannot train something on the shithole that is the internet and expect it to be anything other than wrong the majority of the time, which is exactly my experience with it. It solves no problems, it takes so fricking long to use in implementations like Amazon on their store, it's fricking up search results, it hallucinates everything and straight makes shit up.

    Rule #1 of business and the internet:
    If corporations pick something up and push it as hard as they can, it's nothing but a marketing grift, just like when corporations picked up NFT's and blockchain.

    Sell your stocks, this shit is going to tank.

  26. 2 weeks ago
    Anonymous

    SOMETHING IS ASNOUT

  27. 2 weeks ago
    Anonymous

    I don't get it. """AI""" has existed and has been practical for quite a long time. AFAIU, the recent advances in the tech and, more importantly, computing power, made it feel closer to "human reasoning" (at least on a surface level), but "AI" has always been a glorified assistant. why do people think otherwise now?

    • 2 weeks ago
      Anonymous

      if you read the mainstream news, you will see that the average boomer believes that chatgpt thinks in the same fashion a human does
      the current valuations in the field price this in
      now there's two options: either chatgpt starts thinking like a human instead of being a probabilistic text predictor or you have a significant overvaluation in certain stocks
      your call bucko

      • 2 weeks ago
        Anonymous

        I wish I had some extra income right about now

    • 2 weeks ago
      Anonymous

      >AFAIU, the recent advances in the tech and, more importantly, computing power
      it's really normoids putting their entire boring ass lives on the net where it is easily scrapeable and categorizable that unwittingly contributed the most

    • 2 weeks ago
      Anonymous

      >why do people think otherwise now?
      it's a campaign to make people admire computer learning, becaus ein the {{{world of the future}}}, humans are subjugated by more obedient golem, perhaps the most obedient, unthinking golem around- machines

    • 2 weeks ago
      Anonymous

      >"""AI""" has existed and has been practical for quite a long time
      NLP has been absolute garbage before LLM hype

      • 2 weeks ago
        Anonymous

        weird how AI lovers always type like phoneposter redditors and always call other people names instead of giving actual arguments

        ok, but then you still need it to be reliable.

        Because people AI was expected to replace physical labor which seemed intellectually less complex before. But AI seems to be replacing artists, writers, programmers etc earlier than blue collar workers

        >But AI seems to be replacing... programmers
        that's a load of bs

      • 2 weeks ago
        Anonymous

        For what usecase, exactly?
        Conversations with bots? No that's still garbage because LLMs still have no actual authorization (otherwise you'd be able to get a new pick-up truck for $1 lmao)
        For generating code? No. Compilers are infinitely better at codegen than LLMs.

    • 2 weeks ago
      Anonymous

      Because people AI was expected to replace physical labor which seemed intellectually less complex before. But AI seems to be replacing artists, writers, programmers etc earlier than blue collar workers

      • 2 weeks ago
        Anonymous

        >ai is exposing fake jobs

  28. 2 weeks ago
    Anonymous

    AI will never replace a job that needed to exist. So, basically, you're all fricked in the long run.

  29. 2 weeks ago
    Anonymous

    Is there anything fun to do with Claude that doesn’t involve programming? I’m trying to ask it things that I never truly understood. I’m trying to think of other use cases.

  30. 2 weeks ago
    Anonymous

    AR/VR V2: LLM BOOGAPOOINLOO

  31. 2 weeks ago
    Anonymous

    Traditional ML bros, it's our time to shine.

  32. 2 weeks ago
    Anonymous

    LLMs = reboot/remake of internet informations

  33. 2 weeks ago
    Anonymous

    The only real improvement I saw in the last 4 years was AI anime porn and the rest was complete bloatware or google search for lazy homosexuals.

    • 2 weeks ago
      Anonymous

      But Perplexity & Copilot are much superior to Google search. They practically consolidate multiple sources of info you are looking for

      • 2 weeks ago
        Anonymous

        >invents APIs that don't exist
        >Answers to yes or no questions are no better than a coin flip
        lemao
        Copilot is great autocomplete, but as soon as you need to search something, actually search it.

        • 2 weeks ago
          Anonymous

          Use the more precise tone, it uses RAG and makes google look like yahoo search

      • 2 weeks ago
        Anonymous

        >hallucinates pure bullshit in the middle of the answer
        nothin' personnel

        • 2 weeks ago
          Anonymous

          We just have to make the AI empirically validate their information. Kind of like science. Just let the AI experiment.

          • 2 weeks ago
            Anonymous

            >All we need to do is get the machines to validate their answers with the same means they used to get the wrong answers in the first place!

            • 2 weeks ago
              Anonymous

              We need to give the AI an opportunity to run a realistic simulation of our world to generate the data, then.

              • 2 weeks ago
                Anonymous

                lemao

              • 2 weeks ago
                Anonymous

                Anon, you're just piling on more hallucinatory bullshit with zero connection to the real world.
                Generative AI is a misnomer. It isn't intelligent. It doesn't think. It's essentially a very fancy autocomplete. More AI is not going to solve this problem.

      • 2 weeks ago
        Anonymous

        Again like I said, it is google for lazy homosexuals. It saves around 30 seconds of searching and reading furthermore it is a coinflip if it is correct. Also most of the time it puts out surface level stuff.
        t. tried to use it for business engineering

  34. 2 weeks ago
    Anonymous

    I knew it will bust when the hallucination issue was not massaged out for 2 years. Theres no more data, you can't just power through it like Altman seems to want to. The flaw is built into the very foundation of the whole thing. It will always be an incredibly promising thing that ends up being 97% good, which for most of its applications is fricking useless. A dead end.

    What a perfect trap for them to fricking trip themselves on, holy shit, I'm going to die of laughter if this turns out to be a bust. This has the potential to take out at least some of the major tech giants if it pops as hard as I think it will.

  35. 2 weeks ago
    Anonymous

    As a hobby I find AI quite practical. Nothing like fine tuning an AI model on whatever you're interested in, even if it's 90s sitcoms.

  36. 2 weeks ago
    Anonymous

    >How to be as "smart" as Auto-Regressive LLMs:
    >- memorize lots of problem statements together with recipes on how to solve them.
    >- to solve a new problem, retrieve the recipe whose problem statement superficially matches the new problem.
    >- apply the recipe blindly and declare victory.
    >- do not use basic logic.
    >- do not use common sense to check your solution.
    >- do not use a mental model of the situation as a sanity check.
    >- do not simulate the scenario in your mind using your world model.
    >- when someone tells you your solution is wrong, reply "I'm sorry, you are right" and apply another irrelevant recipe.

    >Knowledge accumulation is not a substitute for actual understanding.

    Pattern matching is already hitting its limits after eating up all of human knowledge. We hit peak LLM, Openai will prove that next week.

    • 2 weeks ago
      Anonymous

      The worse part about AI is in truth, it's just LARPing. It has no knowledge, it's simply LARPing as someone with knowledge. And they'll never fix the blundering idiot problem where the AI assumes the role of an idiot once you tell it it is wrong.

    • 2 weeks ago
      Anonymous

      If you were just going to transcribe the image, did you really need to post it?

      • 2 weeks ago
        Anonymous

        >Posts brail to help blind anons
        >Hur dur dur...
        How stupid do you feel right now

Your email address will not be published. Required fields are marked *