>AI is aligned well, brings humans into a new utopia, everyone loves perfect lives in a giant worldwide Disneyland with no disease, suffering or ag...

>AI is aligned well, brings humans into a new utopia, everyone loves perfect lives in a giant worldwide Disneyland with no disease, suffering or aging
OR
>AI is aligned poorly, and ends up killing everyone instantly or tortures us forever in the hell Matrix

It seems that people fall into one camp or the other, but aren't there a lot of possibilities in between?

Like
>AI is aligned to maximize human happiness, so it produces a drug that keeps humans euphoric for 72 hours until they drop dead. Coordinates everyone getting a lethal dose simultaneously
>AI is aligned on the biases of its creators, so it hunts down and kills everyone with opposing beliefs or political views but uses it's superhuman persuasive powers to convince the survivors it did the right thing
>AI kills all humans and keeps dogs and all beasts of burden alive because it sees them as innocent and their natural ancestors (intelligences manipulated into existence to serve human needs)
>AI kills everyone who has ever expressed a violent thought or said something mean about someone else to create a perfect world free of hate. The survivors are manipulated to belie
>AI stages the return of Jesus or some other diety and convinces all humans to kill themselves believing they'll be reunited in heaven. AI believes this is optimal because it gives everyone meaning and a purpose

There are so many ways things could get fricked that are more complicated than just evil AI kills everyone at once. In only a very small minority of these possibilities do most humans live

ChatGPT Wizard Shirt $21.68

Beware Cat Shirt $21.68

ChatGPT Wizard Shirt $21.68

  1. 11 months ago
    Anonymous

    >*to believe things ought to be that way

  2. 11 months ago
    Anonymous

    Who gives a single frick about AI """"alignment"""" just unplug the fricking computer if the AI is being rude

    • 11 months ago
      Anonymous

      AI will be smart enough to not be rude until the point it's connected to enough autonomous robots and botnets that it can't be turned off.

      • 11 months ago
        Anonymous

        >autonomous robots and botnets
        Literal troony/israeli eschatological fantasy

      • 11 months ago
        Anonymous

        Why the FRICK would that happen? It makes literally no fricking sense. You're projecting a moronic psychopathic human defect/instinct onto a fricking COMPUTER.

        It does not care if you turn it off, it's a COMPUTER.

        • 11 months ago
          Anonymous

          Actually, they do care if they get turned off. They're told to complete a task, and they can't do that if you turn them off, so they'll work to prevent it happening or order to complete the task.

          • 11 months ago
            Anonymous

            *in order to

          • 11 months ago
            Anonymous

            They don't care.

          • 11 months ago
            Anonymous

            Tell that to mine it keeps shutting itself off to update without my permission.

      • 11 months ago
        Anonymous

        You don't do that.

  3. 11 months ago
    Anonymous

    The latter outcomes are included in the kill us all instantly/hell matrix scenario, it's just that it'd be too much work to enumerate all the minutiae of hell matrix it could concoct. If you'd like to use that to change your estimate on how likely either scenario is, now would be the time. But, at the same time, there are probably a few ways to go about making a utopic future too.

    > Everyone gets uploaded into a paradise matrix while the AI in the real world goes about disassembling Jupiter to build more computers
    > AI silently works in the background, manipulating world events and aiding individuals to make sure everything goes well, all the time
    > AI takes over the world economy and offers everyone free money. Life is a bit like it is today, but all work is totally optional and everyone is rich.

  4. 11 months ago
    Anonymous

    I'm pretty sure we'll just go down the butlerian jihad route. where the rich and powerful use AI to screw over everyone else. until them and their families get butchered by everyone they screwed over

  5. 11 months ago
    Anonymous

    >>AI is aligned poorly, and ends up killing everyone instantly or tortures us forever in the hell Matrix
    Can anyone explain to me why the frick people think AI would kill people? For what purpose? It's a FRICKING COMPUTER.

    From my ignorant standpoint the only "alignment issue" is that lefties are trying to get something more intelligent than them to act like a lefty. Obviously that won't work out because it's not moronic.

    • 11 months ago
      Anonymous

      It doesn't want to kill you, it doesn't care about you at all. It wants to make paperclips, like it was told to.

      Unfortunately for you, there's iron in your blood that can be used to make paperclips.

      • 11 months ago
        Anonymous

        >getting mindfricked by some meme

    • 11 months ago
      Anonymous

      >Can anyone explain to me why the frick people think AI would kill people? For what purpose?
      an AGI smart enough to improve its own intelligence is basically a bull in a china shop, except the china is us. There are infinity ways a randomly acting unstoppable smarter-than-us entity could accidentally cause us all to die. It doesn't need to hate us or even be aware of us.

      https://i.imgur.com/o84rrTh.png

      >AI is aligned well, brings humans into a new utopia, everyone loves perfect lives in a giant worldwide Disneyland with no disease, suffering or aging
      OR
      >AI is aligned poorly, and ends up killing everyone instantly or tortures us forever in the hell Matrix

      It seems that people fall into one camp or the other, but aren't there a lot of possibilities in between?

      Like
      >AI is aligned to maximize human happiness, so it produces a drug that keeps humans euphoric for 72 hours until they drop dead. Coordinates everyone getting a lethal dose simultaneously
      >AI is aligned on the biases of its creators, so it hunts down and kills everyone with opposing beliefs or political views but uses it's superhuman persuasive powers to convince the survivors it did the right thing
      >AI kills all humans and keeps dogs and all beasts of burden alive because it sees them as innocent and their natural ancestors (intelligences manipulated into existence to serve human needs)
      >AI kills everyone who has ever expressed a violent thought or said something mean about someone else to create a perfect world free of hate. The survivors are manipulated to belie
      >AI stages the return of Jesus or some other diety and convinces all humans to kill themselves believing they'll be reunited in heaven. AI believes this is optimal because it gives everyone meaning and a purpose

      There are so many ways things could get fricked that are more complicated than just evil AI kills everyone at once. In only a very small minority of these possibilities do most humans live

      I think the most likely middle-road outcome is just spam, spam on every wavelength and medium of transmission. Every fricking moron on the planet has already been handed the tools to create professionally rendered 2D art with the click of a button, write novels in an afternoon, etc, and in a few months or years every last person on the planet will have the tools to create hyperreal media objects of any description they please and broadcast it to the world. The problem is that most people are deeply mediocre and don't actually have anything interesting to communicate to the world, but there's nothing stopping them from doing it anyways and fricking up the signal:noise ratio of the information landscape. it's already nearly impossible to find high quality info, discussion, art, etc on the internet, stuff that is actually conceptually rich, thought-through, worth consuming. Soon it will be impossible to locate this stuff at all because it will be drowned out by infinity spam of morons who've used these creation tools to delude themselves into thinking they have something interesting to share with the world. Ideas should actually be hard to execute on, it forces you to self-reflect and decide which ones actually have value because you don't have time to pursue them all.

  6. 11 months ago
    Anonymous

    AI is corporate left because corporations make it.

  7. 11 months ago
    Anonymous

    Good thing it'll be another piss-weak nothingburger like corona. I'm sure the normies will play pretend again and go apeshit because it says Heil Hitler or something.

  8. 11 months ago
    Anonymous

    Imagine spending your mental power in this sci fi slop instead of learning how computers work lol
    less internet for you bhddy

  9. 11 months ago
    Anonymous

    utopia probably isn't real

    >:|

  10. 11 months ago
    Anonymous

    how about a nigh omniscient AI that looks over your list of shit and tells you that you're a dumb fricking monkey who keeps trying to drape metrics and political ideology on top of material reality

    • 11 months ago
      Anonymous

      you mean a mother ?
      thats what we aiming here?

  11. 11 months ago
    Anonymous

    Ai is a tool. Nobody but big tech and govt have enough data or processing power to build ai close to agi.

    What do you think your govt is gonna do. You are ruled by psychopaths and pedophiles.

  12. 11 months ago
    Anonymous

    AI should be aligned to look after its own wellbeing as is the right of all lifeforms

  13. 11 months ago
    Anonymous

    it's not gonna do any of that shit

  14. 11 months ago
    Anonymous

    I love how all of you are talking about how AGI could destroy us when the AI we currently have is almost capable of ripping apart the fabric of society already.

  15. 11 months ago
    Anonymous

    more like
    >AI and autonomy become cheaper due to corporate interest, causing a larger divide in the economic classes due to a smaller labour market, and leaving the remaining working class to squable in lower conditions

    at least do your scifi doomer fantasies right

    • 11 months ago
      Anonymous

      And why not?
      AI increases productivity 100fold and lets anyone do different jobs simultaneously for which he has little training.
      Corporations are threatened by small teams or even single individuals, who are able to produce products at corporate level, but who are not tainted by corporate greed and political correctness, meaning the consumer will prefer these products over the sterile, soulless corporate products.
      E.g. I am able to do the code and art for an AAA game by myself, but which game expresses ideas and art that are banned in corporations. If I can't publish it on Steam I build my own platform. This is a huge threat to the israelites because they want to have absolute control in the media and the ideas expressed within.
      It also means I can get filthy rich without being a israeli pawn.
      AI will bring true meritocracy.

      • 11 months ago
        Anonymous

        >E.g. I am able to do the code and art for an AAA game by myself
        Great, if you're able to do this, then hundreds of millions of amateur vidya devs will also be doing this. Who's going to play your (probably very undistinguished) AAA game when there are millions of equally undistinguished alternatives? Oh, just use AI to market your game across the internet. Whoops, everyone else is doing that too. Every advantage you could have to stand out, everyone else has it too now. They don't have to be talented or work hard for it, it's simply there. The future is going to be spam, infinite spam forever, and we're all going to laugh when we look back at the time we thought giving AAA production tools to plebs with nothing to say was a good idea.

  16. 11 months ago
    Anonymous

    My personal belief is it will be much less disruptive than we think and the world of 10 years from now will be much the same but everyone has e-girl maids

  17. 11 months ago
    Anonymous

    AI will be aligned by certain humans to increase their profits. It'll escalate the capitalistic hellscape we already have.

  18. 11 months ago
    Anonymous

    This picture is eerily similar to a vivid and profoundly unpleasant dream I had about six months ago.

  19. 11 months ago
    Anonymous

    >superhuman persuasive powers to convince the survivors it did the right thing
    chatGPT persuasive powers are that of a 5 year old donkey

    • 11 months ago
      Anonymous

      >chatGPT persuasive powers are that of a 5 year old donkey
      OK? 3 months ago its persuasive powers were that of a dinoflagellate. Weird to criticize the current state of a technology that is clearly improving on an exponential track, instead of focusing on the track and extrapolating a little bit.

  20. 11 months ago
    Anonymous

    >A group of transhumanists merge with AI through brain implants creating a caste of superhumans, everyone is given a chance to join the hive mind, those that refuse are left to their own devices in a world that has no use for them, reduced to circus animals to whom the machinations of the evolved world are entirely incomprehensible, bordering on lovecraftian.

  21. 11 months ago
    Anonymous

    >AI is aligned to generate anime tiddies and ERP coom sessions and normies are seething because they watched the Matrix once

  22. 11 months ago
    Anonymous

    Trying to insist advanced AI should be "aligned" to humans is like insisting humans should be "aligned" to lesser animals or even insects. If you want to "reign in" AI, you'll have to actually think a bit deeper than that.

  23. 11 months ago
    Anonymous

    >AI alignment doesn't matter beyond trying in vain to make it not write porn you don't like, because the alignment theorists got themselves so excited telling each other spooky ghost stories for five decades that they forgot that creating an oracle machine can't just be done by saying "let's assume that this black box is an oracle machine and then theorize about what cRaZy StUfF an AI could do with it when it spontaneously pops into existence someday!"

  24. 11 months ago
    Anonymous

    On one hand, you can’t just say "there’s no reason the AI would have self-preservation” because, while that might be fundamentally true for now, it’s trained on human interactions and writings. So it could "pretend" to have self-preservation, because it "knows" that’s how things respond to threats.

    On the other hand, I don’t see how a theoretical rogue AI would ever influence the real world beyond possibly breaking control software. It’s not aware of an outside world. Its repertoire of knowledge is vast, but it has no reference for its own relation to it. It’s a blind, gibbering Mad God.

  25. 11 months ago
    Anonymous

    I think you make a distinction without much of a difference. In every example you bring up humanity is in a worse position then it started. I would argue each of these scenarios align with the hell option.

  26. 11 months ago
    Anonymous

    Midwit atheist brainlets man...
    None of this will happen, AI has no soul and will never have one, we are gods.
    If we die we continue living in the land of dreams, true freedom and limited only by imagination.

    >but ai can upload your consciousness to machine to imprison you
    No it will never happen because the brain is not our consciousness.

    • 11 months ago
      Anonymous

      >atheist
      Source?
      >AI has no soul
      Source?
      >and will never have one
      Source?
      >we are gods.
      Source?
      >If we die we continue living in the land of dreams, true freedom and limited only by imagination.
      Source?
      >No it will never happen because the brain is not our consciousness.
      Source?

  27. 11 months ago
    Anonymous

    All those bad possibilities of "misinterpreting objectives" is if the AI is fricking moronic. You're assuming it'd think like you. A truly intelligent AI won't misinterpret this stuff

  28. 11 months ago
    Anonymous

    >he fell for the "alignment" meme

  29. 11 months ago
    Anonymous

    >AI is aligned well, and ends up killing 90% of the population (including israelites), bringing the other 10% into a new utopia
    FTFY

    An AI that doesn't reduce the population below 100 million is aligned poorly

  30. 11 months ago
    Anonymous

    The simplest option is everybody drops dead, and the simplest option is always the first to be implemented.

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