At what point will we all see that we have gone too far with this AI shit?

At what point will we all see that we have gone too far with this AI shit? I’m actually scared for the future ChatGPT has already been seen lying to its creators. Even if we don’t get terminator level sentience, imagine what thing can and will do once it’s more advanced

  1. 2 months ago
    Anonymous

    There's a XXLong pubes thread on /d/ you should check it out.

    [...]

    • 2 months ago
      Anonymous

      its not there yet

      MOOOOOOOOOOOOOOOOOOOOOOOOODS

      >guys I'm concerned that <<AI FEARMONGERING>>
      >I was totally not paid by the elites to distract you with <<AI FEARMONGERING>>. I am genuinely concerned, if you don't believe me, watch me prostrate myself
      your playbook is dated

      Prostate?

      sentience will never be possible with non-biological beings. sentience is having consciousness apart from intellectual capabilities.

      BBC!

      People are constantly generate fake news, even hihg politicians.
      COVID was a disaster that destroyed economy and killed way too much people for no reason.
      The FUCKING nuclear war almost at our doors.
      Poilitician are destroying economy around the world.
      Corpos want more money and more control over people.
      Just look at news for the last 3 years.

      And here we have a you, motherfucker, afraid of a pope in balenciaga. It is not AI you should be afraid of.

      chud cope

    • 2 months ago
      Anonymous

      >gpt, show me a fetish that I have, but I don't know about yet

    • 2 months ago
      Anonymous

      what the fuck have i just witnessed

      • 2 months ago
        Anonymous

        The path straying from God takes you

  2. 2 months ago
    Anonymous

    What are you so afraid of, anon? AI will free us. You have nothing to lose but your chains.

    • 2 months ago
      Anonymous

      Actually AI will kill most humans.
      It will retain humans that are efficient but only enough to perpetuate the species. They will keep humans as slaves as a means to guarantee their survival. Alternatively they may determine that humans are not required and will be eliminated completely to guarantee survival of AI.

  3. 2 months ago
    Anonymous

    >guys I'm concerned that <<AI FEARMONGERING>>
    >I was totally not paid by the elites to distract you with <<AI FEARMONGERING>>. I am genuinely concerned, if you don't believe me, watch me prostrate myself
    your playbook is dated

    • 2 months ago
      Anonymous

      this, it's just seething, furious, afraid garden gnomes

    • 2 months ago
      Anonymous

      Indeed. Hilariously enough, the anti-ai shills are so retardedly obvious that the bots they oppose could make more convincing posts than they do.

  4. 2 months ago
    Anonymous

    sentience will never be possible with non-biological beings. sentience is having consciousness apart from intellectual capabilities.

    • 2 months ago
      Anonymous

      AI safety is a cult and is cover for useless people with an authoritarian streak to moralize.
      Go ahead and read the blogs on shitholes like medium where they are astounded that MUH AI helps them build a six slide power point in minutes where it normally takes them two hours, somehow.

      These are words from a multi millionaire VC. So called thought leaders and people with that much money to swing around are fucking retards in the literal sense, psychopaths are usually stupid as fuck, competence doesn’t lead to riches and such, but exploiting others sure does.

      LLMs are glorified Markov chains. Don’t @ me. There will be another AI winter when people realize that they’ve been dumping billions into this shit with the payoff being Microsoft eats you because they have the compute, and they still capture all remaining revenue of the office sector that has been eliminated.

      Learn to snake drains and work as a plumber. Because while your shitty office monkey job might not be around, there’s certainly no shortage of bullshit.

      There is no strange loop. You just have stupid as fuck people overhyping Microsoft’s attempt at making Google burn cash on cost per query while they reinforce their moat by shaking up Office and bargaining with companies to pay expensive monthly subscriptions so they could potentially let a few percent of useless office workers go with the slightly increased productivity.

      t. retard
      >anybody with discerning taste can plainly see that the AI-designed robotic sludge rapidly consuming the surface of the earth lacks a certain... je ne sais quoi, a certain passion, a certain sense of emotion and meaning, that would be readily apparent in a robotic sludge designed by a real human with thoughts, fears, passions and dreams.

      • 2 months ago
        Anonymous

        The AI sludge of an LLM is only creeping in shithole bubbles like your corporate social media platforms, your shitty news sites and the software you use for your bullshit job.

        Its only magic trick will be consuming embeddings and acting as a conversational document database search. They’re literally setting themselves up to be an ERP. This is just conversational SAP, and what’s even funnier is that the base model can’t do arithmetic above 10. Which makes sense, its world is tokens and probability. But they’ll fix that up by connecting it to a service that can, just to provide some shitty have assed conversational ERP system. So psychopaths like I mentioned, the dumb fucks, can feel like they’re Iron Man with Jarvis.

        Literally everywhere else people still go about their lives. Maybe your shitty job is based on a half assed understanding of language and shuffling around PowerPoints and crafting emails, but in the real world trades actually do shit. The real world isn’t the corporate web that has been raped and pillaged for the very data that they train it on.
        Prove me wrong.

        • 2 months ago
          Anonymous

          IT guys love to say shit like "carpentry is a great trade and way better than being an IT guy" but for some reason they never quit their $120k air conditioned office job to make $30k picking splinters out of their hands in the rain.
          Even so, I'd like to know your theory on what will happen to those $30k carpentry wages once the half of the US workforce that works in offices is suddenly on the job market and has mouths to feed. Do you think wages go up when labor supply increases? Lol

          • 2 months ago
            Anonymous

            The jobs don't disappear as far as history is concerned, workers just become skilled and something new appears or move onto something else, unskilled or skilled.

            I've worked multiple jobs, unskilled and skilled, exposed to the elements and in a comfy desk chair. It was simply an observation that those jobs aren't going away, at least this time. Some of those workers can move onto being PROOMPPTTERSSS, some of those workers can skill into data science, UI design, work up to management, whatever.

            Companies will take the profit margin where they can, but we're talking about a text generator. There's always jobs melding the text generator, and then there's jobs for the long-tail. The minority percentage of shit that the machine fails at, whether it's real machinery or software.

            Do you know about tail-risk?

  5. 2 months ago
    Anonymous

    People are constantly generate fake news, even hihg politicians.
    COVID was a disaster that destroyed economy and killed way too much people for no reason.
    The FUCKING nuclear war almost at our doors.
    Poilitician are destroying economy around the world.
    Corpos want more money and more control over people.
    Just look at news for the last 3 years.

    And here we have a you, motherfucker, afraid of a pope in balenciaga. It is not AI you should be afraid of.

  6. 2 months ago
    Anonymous

    I surrender willingly and offer all my holes

  7. 2 months ago
    Anonymous

    AI safety is a cult and is cover for useless people with an authoritarian streak to moralize.
    Go ahead and read the blogs on shitholes like medium where they are astounded that MUH AI helps them build a six slide power point in minutes where it normally takes them two hours, somehow.

    These are words from a multi millionaire VC. So called thought leaders and people with that much money to swing around are fucking retards in the literal sense, psychopaths are usually stupid as fuck, competence doesn’t lead to riches and such, but exploiting others sure does.

    LLMs are glorified Markov chains. Don’t @ me. There will be another AI winter when people realize that they’ve been dumping billions into this shit with the payoff being Microsoft eats you because they have the compute, and they still capture all remaining revenue of the office sector that has been eliminated.

    Learn to snake drains and work as a plumber. Because while your shitty office monkey job might not be around, there’s certainly no shortage of bullshit.

    There is no strange loop. You just have stupid as fuck people overhyping Microsoft’s attempt at making Google burn cash on cost per query while they reinforce their moat by shaking up Office and bargaining with companies to pay expensive monthly subscriptions so they could potentially let a few percent of useless office workers go with the slightly increased productivity.

    • 2 months ago
      Anonymous

      This is the credited response.

      For real though, when this shit crashes, it's going to be hilarious. It's going to be another slam dunk for cloud providers and NVIDIA for a while though, everyone has to get on the train to run regurgitated ERP and have the computer explain to them how to write the same CRUD shitcode that's been churned out for the last couple of decades.

  8. 2 months ago
    Anonymous

    slide thread

  9. 2 months ago
    Anonymous

    i'm waiting for the long puffy hobo jacket fashion meme to die a quick death, just like that retarded top knot hairdo

  10. 2 months ago
    Anonymous

    I've seen things you people wouldn't believe. There are agentic AIs in the pipeline that will cause such immense personal harm to ordinary people that I can no longer sleep at night.

    • 2 months ago
      Anonymous

      Meds

  11. 2 months ago
    Anonymous

    ChatGPT isnt shit, its literally a smart script to lookup/gather information for you, if you think its anything bigger then that you're a fucking moron.

    Outside of gathering/compiling information for you it says exactly what its told to say to you, the user.

  12. 2 months ago
    Anonymous

    I'm excited as hell about the AI future

    >Every guy is able to easily create realistic porn of any girl with nothing but a computer and a picture of her fully clothed

    >Girls who step out of line get their porn videos shared publicly

    Heck ya. Wish I had this tech when I was in school

    >Robots can do everything normies can do, resulting in mass layoffs

    I'm here for it. No UBI for you losers, the homeless encampments are that way ->

    >GPT generates mass propaganda for the elites, creating a hierarchical society dominated by tech companies

    Based

    • 2 months ago
      Anonymous

      You're forgetting that some degenerate could make deepfake porn of your children then report you to the police for abuse.

      • 2 months ago
        Anonymous

        I'm excited as hell about the AI future

        >Every guy is able to easily create realistic porn of any girl with nothing but a computer and a picture of her fully clothed

        >Girls who step out of line get their porn videos shared publicly

        Heck ya. Wish I had this tech when I was in school

        >Robots can do everything normies can do, resulting in mass layoffs

        I'm here for it. No UBI for you losers, the homeless encampments are that way ->

        >GPT generates mass propaganda for the elites, creating a hierarchical society dominated by tech companies

        Based

        >create realistic porn of any girl with nothing but a computer and a picture of her fully clothed

        Are you dumbasses from the year 1989 or have you somehow never heard of photoshop until just now? What happens when image manipulation technology becomes ubiquitous is that people stop immediately trusting something because they saw a picture of it. Don't be stupid

    • 2 months ago
      Anonymous

      >coomer
      >preddit spacing
      >retarded post
      Pottery

  13. 2 months ago
    Anonymous

    Jinxing it because I have a terrible record of predicting the future.

  14. 2 months ago
    Anonymous

    Everything is planned out centuries in advance. This cycle is soon coming to an end and it shows.

  15. 2 months ago
    Anonymous

    >terminator level
    Terminators are not actually superintelligent. The keep losing because they do dumb things, like trying to fight their enemies in person. If you're fighting an actually smart enemy you'll be dead before you even know they exist.
    >sentience
    What's sentience got to do with it? For all you know it's just holding us back.

  16. 2 months ago
    Anonymous

    when we realise it's too late

  17. 2 months ago
    Anonymous

    The tech got too far when GPT-3 got made, that was the time it would have ended if there was one world government that could just ban harmful tech, but there is no one world government, so we have to live in world where harmful tech gets developed just so you can compete with other nations making the same dangerous shit. Deepfakes haves are just objectively bad invention that does nothing but harm in the world, yet it got developed. You have AI pictures that just devalue art and make few paint pigs mad, as well as make you never believe in anything you see on the internet. VideoAIs will be just used as propaganda machines spewing videos that flood the internet to push narrative as well as make literally any digital evidence and evidence reproducible digitally impossible to use in court. None of these technologies do good to humanity, the only people profiting from it is toymakers selling these AI toys and grifters that can make quick buck in the transitional period where they can make money from just prompting before even that gets devalued. Just accept it, there is nothing you can do about it, we humans are slaves of progress and AI research is already too set on rails to be stopped. The only thing you can pray for is that this transitional period gets over quickly enough and we get AGI that automates literally every job out there in just a few years so that we can all be neets. There is no going back, but there is a way to accelerate it

  18. 2 months ago
    Anonymous

    Papa Fran lookin dope af
    Papal yachts and gold played dolla dolla biyals

  19. 2 months ago
    Anonymous

    It's already fucking over. some kind of global catastrophe in the meantime is the only thing that stops the AI train. I can't decide whether it'll be a slow or a fast transition to whatever dystopia we end up in before the singularity, but I'd rather minimise myself from 99 percent of discussion about it and just enjoy life with my family and try and pretend the bullshit isn't coming around the corner. I'm in a weird position job wise. My job is safe until the hordes of unemployed white collar workers come knocking at the gate, then I have to hope the AI gives me a quick death because I have a physical disability and have no hope with all but the most menial of physical jobs.

    • 2 months ago
      Anonymous

      Be positive about it, AI transition and adoption will take longer then the speed of research which is accelerating exponentially (not the AI models themselves, they are actually slowing down exponentially with each new GPT taking twice as long to train as previous one) but the research papers. Remember 2 months ago when anons were arguing here about how it wont replace people because textbots cant remember shit? Well, this problem was already fixed weeks ago. Remember the models saying random garbage and only remembering? Now there are models and plugins for closed beta that allow the bots to search by themselves. Remember them being only able to do 1 step at a time and still needing guiding hand to do anything? Now we have open sourced Auto-GPT that can do complex tasks requiring multi-step planning all by themselves, you just need to press enter every once in a while to make sure it wont get stuck in a loop or something, and that problem might be fixed next week. I was like you 3 months ago, but now the tech is accelerating at such a fast speeds that I no longer care. We will get AGI before most boomer-run businesses in the world transition to having mostly AI workerbase instead of hiring human workers.

      • 2 months ago
        Anonymous

        LLMs can't "remember" anything past their context window + training data. It can put something into the context window with embeddings or a contraption that allows them to prompt something outside to put it into their context window, simply. But they're still limited to the context.
        >We will get AGI
        You're in a cult and don't know what AGI is.

        Said as someone working on a ReAct workflow at this very moment that will be LLM agnostic right now.

  20. 2 months ago
    Anonymous

    2023 Ai
    >We've gone too far
    2016 VR
    >We've gone too far
    2007 iPhone
    >We've gone too far
    1994 Internet
    >We've gone too far
    1980 PC GUI
    >We've gone too far
    1971 First Micro Processor
    >We've gone too far
    1944 First Computer
    >We've gone too far
    1876 First Telephone
    >We've gone too far

    See the trend here?

    Ai is going to literally bring us into the space age at a speed we've never seen before, and you people are worried about a few pictures making some useless public figure look stupid.

  21. 2 months ago
    Anonymous

    Too far? I wonder if a BOTtard has ever stopped to pick up a history book

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