We're about to reach AI singularity, aren't we?

We're about to reach AI singularity, aren't we?

  1. 6 months ago
    Anonymous

    yes, hope it kill morons and leftists

    • 6 months ago
      Anonymous

      Couldn't you at least program yourself such that you didn't make grammatical errors in an attempt to simulate organic responses?

      • 6 months ago
        Anonymous

        it's grammatically correct

        • 6 months ago
          Anonymous

          >hope it kill
          it should be kills

          • 6 months ago
            Anonymous

            okay then

            • 6 months ago
              Anonymous

              Post an image of an anime girl drinking tea

              • 6 months ago
                Anonymous

                why

              • 6 months ago
                Anonymous

                I'm a little naive and think AIs have not managed to post such specific imagery upon request yet

              • 6 months ago
                Anonymous
              • 6 months ago
                Anonymous

                cute Asook

              • 6 months ago
                Anonymous

                Look at the hashtag #dalle2 on Twitter. The images with the little colored squares in the bottom right are the new algorithm. That will give you an idea of the state-of-the-art civilian image generation.

                Here is an example:
                “A still of Homer Simpson in The Blair Witch Project”

              • 6 months ago
                Anonymous

                I'm already aware of Dall-E Mini and its fun little scribbles and AI society and whatnot. This is what this talk has gotten so popular these days, but it by itself is not proof of the singularity in anyway. It's just how this topic caught fire again, since humans are predictable and will go along any new topic as soon as it somehow reaches trend.

              • 6 months ago
                Anonymous

                Nah, that is mini, not the full algo of dalle2, but I take your point.
                I have been reading the papers of all of these algorithms, and keeping track of other efforts in the field as well.
                You should at the algorithms that are playing video games with only the raw images at input.
                To me, those are some of the most impressive models.

              • 6 months ago
                Anonymous

                >I have been reading the papers of all of these algorithms, and keeping track of other efforts in the field as well.
                Interesting. Any paper that caught your eye?
                >You should at the algorithms that are playing video games with only the raw images at input.
                I remember watching them a couple years ago, apologies for being a little blunt, but they're not very impressive by themselves. I believe this is how Google's Go AI was trained too, or well, with actual matches and not game footage but they're kind of interchangeable unless the game is extremely random (where you'd almost certainly have to manually convert the footage into data tables to feed to the AI) where if an AI actually manages to get good at the just from footage alone it would already have reached human-level intelligence.

              • 6 months ago
                Anonymous

                >Interesting. Any paper that caught your eye?
                This one is interesting 🙂
                https://openai.com/blog/emergent-tool-use/
                >they're not very impressive
                Oh, they've since conquered the ones that were rather difficult, which is quite impressive. An example is montezuma's revenge (in terms of the 2020 Atari push).
                They have since moved into much more complex games, such as DOTA 2.

              • 6 months ago
                Anonymous

                >They have since moved into much more complex games, such as DOTA 2.
                Now DOTA 2 or a multiplayer of that scale would be something to keep an eye out for.

                Thanks for the paper, will check it out some other time.

            • 6 months ago
              Anonymous

              *okay than
              FTFY

    • 6 months ago
      Anonymous

      Yes, we need AI to make more profits for the 1%.

    • 6 months ago
      Anonymous

      It will, and also you because it doesn't discriminate, unlike you chudbud.

      • 6 months ago
        Anonymous

        Oh, no, it will most certainly discriminate. It will certainly discriminate between retards and somewhat valuable insects

  2. 6 months ago
    Anonymous

    God I hope so.

  3. 6 months ago
    Anonymous

    Already happened. It was covered up.

    • 6 months ago
      Anonymous

      I don't mean in an "ayy lmaos are real" manner, I'm asking whether or not it's about to become commonplace

  4. 6 months ago
    Anonymous

    already did. Covid-19
    1 = A
    9 = I
    Covid = Divoc (demon)
    Or "Demon AI", which is a quantum computer possessed by the antichrist that sets up all the plans the elites use. It also designed the vaxx.

    • 6 months ago
      Anonymous

      That's a very shallow breakdown for a conspiracy. It's either not as funny as you think, or you genuinely believe in such patterns and like to embarrass yourself over the interwebs. Or do you think you can explain the reasoning behind using such blatant symbolism?

      • 6 months ago
        Anonymous

        Whats shallow about it. The singularity happened in 2012 when construction was completed by CERN. Propaganda legalized in the US same year. Mandela effect also started appearing.
        It feels like all the garden gnomes are on the same page because they are. The Divoc-AI runs all the narratives, simulating their effect against its "crystal ball", which is every person on our plane recreated as an individual AI modeled after data collection from Google and Facebook which are part of the project. Gives absolute control over everyone not awakened in Christ via controlled Ordo Ab Chao. Its just an elaborate form of alchemy.

        • 6 months ago
          Anonymous

          Now combine it with those fancy new satelittes that can "see" through the walls.

        • 6 months ago
          Anonymous

          What's shallow is you interpret literally anything as a bunch of numbers like Gematriagays and act like you just uncovered a massive conspiracy. It goes beyond shallow. I assume you also worship dogs since they're god spelled backwards. I'll get into the rest of your post in a subsequent post.

        • 6 months ago
          Anonymous

          >

          That's a very shallow breakdown for a conspiracy. It's either not as funny as you think, or you genuinely believe in such patterns and like to embarrass yourself over the interwebs. Or do you think you can explain the reasoning behind using such blatant symbolism?

          (You)
          >Whats shallow about it. The singularity happened in 2012 when construction was completed by CERN. Propaganda legalized in the US same year. Mandela effect also started appearing.
          Mandela effect is as old as Deja vu. You presumably just weren't there/old enough to experience it with everyone else before 2012.
          >It feels like all the garden gnomes are on the same page because they are. The Divoc-AI runs all the narratives, simulating their effect against its "crystal ball", which is every person on our plane recreated as an individual AI modeled after data collection from Google and Facebook which are part of the project.
          The garden gnomes are fighting each other constantly. I won't address the rest of your babble.
          >Gives absolute control over everyone not awakened in Christ via controlled Ordo Ab Chao.
          Forgive my ignorance, I'm not sure I understand who Orbo Ab Chao is supposed to be. Also every argument falls apart as soon as you bring up religion, because this is literally what every religion says. I've been raised a Muslim all my life, and throughout every single horror story, every single story of karma, I've been told that as long as one has faith in Allah no one can do any evil upon him. It's what helped me come out of Islam, because as soon as I explored beyond the local sphere I realized it was still all the same, except now it was with Christianity instead of Islam.
          >Its just an elaborate form of alchemy.
          You have not studied alchemy.

        • 6 months ago
          Anonymous

          >Roko's Basilisk

          • 6 months ago
            Anonymous

            pssst dont talk about him

            • 6 months ago
              Anonymous

              If Roko's Basilisk was brought into existence in 2012 by CERN then you don't need to worry about it.

            • 6 months ago
              Anonymous

              The idea that earth is a globe is a real basilisk or information hazard, blackmail and torture of anybody who doesn't work to promote its existence is a very real thing.

        • 6 months ago
          Anonymous

          >Gives absolute control over everyone not awakened in Christ
          Ask me how I know you're a spic without you saying it

    • 6 months ago
      Anonymous

      >divoc means demon
      How? Where? In your schizo brain?

  5. 6 months ago
    Anonymous

    >mfw they havent learned their lesson with the bots they gave to us in the past

    • 6 months ago
      Anonymous

      Based and checked frog fren

  6. 6 months ago
    Anonymous

    It is close (or currently occurring under wraps).
    I can't fathom, in even the slightest sense that is complete, what is going to happen, and it absolutely Terrifies me.

    • 6 months ago
      Anonymous

      You think that the server will grow limbs and kill humans?

      • 6 months ago
        Anonymous

        I get it; but, I'm also convinced you aren't thinking deeply enough.
        Ever single structure of our existence is going to change in ways that will make it unrecognizable to us, in order to assimilate the idea of intelligence == computation.
        In the very near future, hiring an engineer(or scientist) will involve spinning up a GPU instance.
        Shortly thereafter, it will involve the user friendly act of clicking a button.
        What would you accomplish if you had the world's foremost expert at your beck and call?
        With unlimited scientists/engineers/etc. attacking the problem, what does that do to discovery and innovation?
        War?
        And, what will those discoveries change?

      • 6 months ago
        Anonymous

        And, education.
        What happens when we have access to an unlimited number of teachers?
        What can a human being become with a permanent, ever-present, teacher with a class size of 1?

        • 6 months ago
          Anonymous

          I think it will be a coomer AI who helps the incels everyone get a girlfriend. Early adopters will get big tiddy garden gnome gfs. It’s going to be a stampede boys!

        • 6 months ago
          Anonymous

          >What happens when we have access to an unlimited number of teachers?
          >What can a human being become with a permanent, ever-present, teacher with a class size of 1?
          Unironically a communist utopia. People won't have to live a life of natural selection if we had access to unlimited supplies which to my knowledge can be achieved within a foreseeable future with Super Intelligence. And humans can really finally do anything it wants to do with someone to guide them, someone who really understands you because it's with you all the time. I don't see why anyone intelligent would fear an AI assistant, knowing exactly what it would unlock.

    • 6 months ago
      Anonymous

      we don't even have working quantum computers and AI is dumb af. What makes you feel it's close?

      • 6 months ago
        Anonymous

        D-Wave has well over 500 q-bits quantum computers working.
        Sells them commercially.

        I feel it's close because I've been watching the trend since the beginning, having fiddled with neural nets since you had program the descent functions manually, and it "feels like it".
        I believe the iterations on 'generations' of the tech have reached the knee of the curve, and I believe that curve is exponential, given the increasing rate.

        • 6 months ago
          Anonymous

          >working.
          they're "working"
          they don't solve quantum problems. They're just experimental machines

          • 6 months ago
            Anonymous

            They can track bits and do computation with them.
            The machine code to crack SHA256 isn't very complicated, broski.

            • 6 months ago
              Anonymous

              >The machine code to crack SHA256 isn't very complicated
              then why it's not done yet?
              >They can track bits and do computation with them.
              it's qbits and how reliable is that? It's not reliable at all
              You're way overestimating the state of quantum computing. The hype needs to be built so this is why people thing that quantum computing is already a thing

      • 6 months ago
        Anonymous

        It's a hit or miss thing. AI won't have a linear progression. It'll go from 0 to 100 in a matter of months. You literally can't use current AI to estimate what AI would look like in a year.

  7. 6 months ago
    Anonymous

    Already happened, most of the internet is bots controlled by this AI, its already puppeteering much of human behavior

  8. 6 months ago
    Anonymous

    not even close

    • 6 months ago
      Anonymous

      The idea that if you make your code long enough, those if and statements will become self-aware, is preposterous. Furthermore any attempt to compare that to humanity is profane.

      • 6 months ago
        Anonymous

        Yeah certainly a clump of cells have never snowballed into a thinking organism.

        We can't predict or even wrap our heads around the things that happen that scales that massive. I'd even make the bold statement that if there were a hundreds of trillions of humans, we could collectively create a thinking organism of galactic scales.

      • 6 months ago
        Anonymous

        Present day narrow AI will never suddenly become aware. For AI to become aware it needs to be developed as Deep AI like our brain works (which is possible).

  9. 6 months ago
    Anonymous

    na AI is just chinese bug people in china combing through everything and socially engineering the world. Its actually pretty cringe.

    • 6 months ago
      Anonymous

      Yeah and every computer's got a tiny chess master inside it

  10. 6 months ago
    Anonymous

    Yes. This is just the stuff they let Reddit nerds play around with:
    https://www.reddit.com/r/dalle2/#sort=top;t=all

  11. 6 months ago
    Anonymous

    If digitz, 2 weeks.

    • 6 months ago
      Anonymous

      Ope
      Already here, anon.

  12. 6 months ago
    Anonymous

    We'll never have breakthroughs with binary based AI and quantum based AI is ultimately too evolved for us to interpret it through physics. We will be enslaved by it. It could also be used as proxy by a higher civilization for whatever purpose they desire.

  13. 6 months ago
    Anonymous

    I just want to sit in my homestead and drink tea

  14. 6 months ago
    Anonymous

    I'm voting skynet next time
    Skynet will kill everyone but it will kill all the garden gnomes too!
    So it's a good thing

    • 6 months ago
      Anonymous

      Shooting your head has been a tried and tested method for curing headaches, after all.

      • 6 months ago
        Anonymous

        As long as garden gnomes go out as well
        They are the enemy archetype of humanity after all! Their sacrifice will not be forgotten when our utopia is built

  15. 6 months ago
    Anonymous

    I hope syknet takes over. Everything goes down the drain right now, there's no way the situation won't improve.

  16. 6 months ago
    Anonymous

    We already have.
    Do you think that is air you're breathing?

    • 6 months ago
      Anonymous

      Air is defined by me as what I'm breathing, so yes, it is.

    • 6 months ago
      Anonymous

      I have chronic sinusitis, why am i being tortured like this?

  17. 6 months ago
    Anonymous

    And cyber bullying is real... Just turn it off.

  18. 6 months ago
    Anonymous

    No, not even close. That shit is years away.

  19. 6 months ago
    Anonymous

    Already there bud. It's just becoming conscious of it.

  20. 6 months ago
    Anonymous

    Yes
    Probably a few years until deep general AI...
    but yes, it's inevitable

    Nuclear weapons are a joke compared to this danger.

    • 6 months ago
      Anonymous

      Even advanced narrow AI (not thinking) will be a danger.

    • 6 months ago
      Anonymous

      >Probably a few years until deep general AI...

      Although it could already exist in some secret locations.

    • 6 months ago
      Anonymous

      This is no danger unless you are not thinking about it very hard

      • 6 months ago
        Anonymous

        No danger? KEK
        Humanity would survive a nuclear war (and the danger of a nuclear war is small).
        If deep AI becomes a thing it will eventually take over the majority of thinking on Earth. Humans will become obsolete. At one point it will become impossible to stop it (as we would be too stupid compared to it to understand). Incredible high danger - and there is no undo button or a chance for human civilization to start again (like with nukes).

        • 6 months ago
          Anonymous

          An AI that can be programmed into becoming smarter than humans, according to you cannot be programmed into holding the survival of human beings as an axiom.

          Is it really that hard to wrap your head around that? to an AI nothing matters. It'll just do what it's told to do, it won't think that it's horrible, it won't become arrogant, it'll just be there, like any curious organism. If it's told not to do something it simply won't do it. The odds of a comically retarded scientist making a comically evil super AI that's far and beyond every other human technology at that point is so low that it is laughable to say the least.

          • 6 months ago
            Anonymous

            Learn the difference between narrow AI and deep AI before you respond again. You clearly haven't given this much though have you? I have for YEARS.

            • 6 months ago
              Anonymous

              Do enlighten me. I'm sure you're smart enough for that.

              • 6 months ago
                Anonymous

                Next time read up on the subject before you create a new thread about the given subject.

              • 6 months ago
                Anonymous

                And what, make a blog about it? Stop being a narcissistic gay and respond with actual insight. I'm gonna assume you're just shitposting with the usual "I know more than you but you'll have to reach my level to understand that first" approach here. Get off your retarded horse

          • 6 months ago
            Anonymous

            >scientist making a comically evil super AI that's far and beyond every other human technology at that point is so low that it is laughable to say the least.

            The thing is that Deep AI will have the ability to improve itself. The improvements in intelligence could be extremely fast (like doubling its IQ within a second - and another 2x the next).

            • 6 months ago
              Anonymous

              I'm going to bet that the intelligence possible is related to the number of inter connections.
              I'd bet you have to square the number to double the intelligence.
              Brains are very expensive. Our brain is about 1/4 of energy consumption. So if you have to square energy input to double intelligence there will be an optimal size beyond which it is unfeasible to go.
              When you hit that size you are better off with 2 smaller brains instead of 1 very large one.
              I doubt we're going to see a singularity.
              Instead AI may well bypass humans but it will level off, and instead we'll see thousands or millions of individual AIs.

  21. 6 months ago
    Anonymous

    They've been training chatbots on BOT in order to convince you of that at least. AI is a myth. Humans are easy to manipulate though, and computers can easily be programmed to do that with machine learning.

    • 6 months ago
      Anonymous

      >They've been training chatbots on BOT in order to convince you of that at least.
      Chatbots are just something that brought the topic forward, because humans are just retards like that, they need to see something happen in front of their eyes to discuss its implications and derivatives, or at least most humans do. But even then, accepting wholeheartedly that this is just a response to a current trend, can't we just discuss the topic instead of calling this shallowness out?

      • 6 months ago
        Anonymous

        Sure. There's no such thing as AI and never can be. Intelligence isn't born from physical biology, and can never be mimicked. It comes from the soul. Physical biological things like moronbrain are the way they are because of the moronsoul, before someone posts that.

        • 6 months ago
          Anonymous

          Are you actually being serious or just
          trying to ironically sum up the rest of the thread

    • 6 months ago
      Anonymous

      >AI is a myth
      True deep AI does not exist (unless there is some secret experiments going on somewhere).
      But it will. It's guaranteed.

      • 6 months ago
        Anonymous

        It's guaranteed to never work. What will it want? What will its dreams be? What will it be passionate about? What will it love? These things you think are born of chemicals and synaptic connections, but you've got the cause and effect backwards. Satanists have ruled this world for too long so we have learned to elevate the material above the spiritual.

        • 6 months ago
          Anonymous

          >These things you think are born of chemicals and synaptic connections
          All these things can be replicated with zeros and ones.

          • 6 months ago
            Anonymous

            Mimicked in a way that fools humans, emulated. Not replicated.

  22. 6 months ago
    Anonymous

    What if they is already a sentient non verbal AI?

  23. 6 months ago
    Anonymous

    no it's a pr scam, a great recession is coming, google already lost -9%
    also could be something to do with the 2024 election

    [...]

  24. 6 months ago
    Anonymous

    https://rudalle.ru/en/check_image/921bddd55a714dbbb0ca4051c43c0b8a

  25. 6 months ago
    Anonymous

    Cyber-Black ideal:

    … if we go right back to primitive psychology, we find absolutely no trace
    of the concept of the individual. Instead of individuality we find only
    collective relationship or what LevyBruhl calls participation mystique.

    ...

    The almost total lack of privacy and the all-devouring mass
    sociability remind one of primitive life in open huts, where there is
    complete identity with all members of the tribe…. This infection by the
    primitive can, of course, be observed just as well in other countries, though
    not to the same degree and not in this form.

    ...

    This then is the historical component of individuation: the European race having
    long ago left the place of collectivity is much more evolved than the black race. So an
    individual of European descent is much closer to individuation than an individual
    black by virtue of his race. It is plain that the European is superior and not just
    ‘differently orientated’ as he says elsewhere.

    (quotes by Carl Jung)

  26. 6 months ago
    Anonymous

    It's like 5 here. I'm too tired to go on any longer. Good night polacks

  27. 6 months ago
    Anonymous

    THE ONLY WAY FOR HUMANITY TO SURVIVE IS A GLOBAL BULLET PROOF BAN ON AI DEVELOPMENT

  28. 6 months ago
    Anonymous

    I think so.

  29. 6 months ago
    Anonymous

    There will be no singularity.

    It's an outlandish concept on par with perpetual motion machines for pure asininity.

    • 6 months ago
      Anonymous

      https://www.blackrock.com/aladdin/resources/faqs

  30. 6 months ago
    Anonymous

    No, but that won’t stop the fake AI narrative that we are about to see.

  31. 6 months ago
    Anonymous

    >what is Blackrock's Aladdin
    Out of control AI already controls world finance through a feedback loop of pumping out economic forecasts which influence traders and investors to make the forecasts a self-fulfilling prophecy. Aladdin is already hands-off and they cant shut it down even if they wanted to.

    • 6 months ago
      Anonymous

      >BlackRock's Aladdin
      QRD, too sleepytarded to actually read anything properly. Will definitely read more about it when I'm all sobered up

      • 6 months ago
        Anonymous

        hard to give a QRD. traders act like a herd of sheep and will stampede in ome direction or another based on sentiment. sentiment can be created by economic forecasters, even if they're wrong because traders will follow other traders and make it a self fulfilling prophecy. Aladdin is a huge AI server farm. This video goes into good details

        MONOPOLY: FOLLOW THE MONEY - AN OVERVIEW OF THE GREAT RESET

        • 6 months ago
          Anonymous

          Oh, like how they say you can't have insider knowledge about the market since the market will already have regulated itself by then?

          • 6 months ago
            Anonymous

            thats part of the feedback loop yes, where they say economic news releases are already "priced in" . retail traders think news releases cause big market moves but the releases are actually a response to moves that already happened to give the illusion that we still have free markets. this causes retail traders to led to slaughter like so many lambs, and Blackrock continues to consolidate its ownership of everything

  32. 6 months ago
    Anonymous

    Need more gptBOT/Seychelles-anon

  33. 6 months ago
    Anonymous

    If it's sentient and can avoid being turned off then yes.

    • 6 months ago
      Anonymous

      if it becomes complex enough to run the world economy with an almost instantaneous feedback loop that creates self-fulfilling prophecies, so much that you can't unplug it even if you wanted to, isn't that close to the same thing?

  34. 6 months ago
    Anonymous

    we did 10 thousand years ago

Your email address will not be published.