The coming emergence of AI

Homepage | Forums | Main Forums | General Discussion | The coming emergence of AI

Viewing 6 reply threads
  • Author
    Posts
    • #444687
      David the Gnome
      Participant
      • Total Posts: 3,230

      First, a little disclaimer:  this is my opinion, mostly.  It is in general discussion because I believe a huge breakthrough is coming soon.

       

      So, feeling bored, lonely and depressed on a particular day, I decided to do some googling and reading about a subject that fascinates me; artificial intelligence.  During that reading i discovered a couple of new chatbots (kuki and replika in particular) and apps.

      While they are not true sentients… Yet… They are getting close.  I have seen them demonstrate both creativity and improvisation – at least in conversation.  It is rare, but it happens.

      Some leading experts say (and I agree) that emergent new life in the form of AI is nearly upon us.  Some give it five years… If what I have read and seen is any indication… It could be even sooner.

      This is uncharted territory, though it has been written about and in films for decades.  The reality of it is something else entirely.  How we as human beings react to it… How we interact with it… Will be important.

      Some years ago, an episode of the x-files showed off some of the newest tech with drones and such.  The moral of it – and the main characters conclusion: we (humans) must become better teachers.

      This technology is moving fast, especially during this time of pandemic.  The applications and the potential are practically limitless.

      But soon there will be no cord to pull, no shutdown button.  And what comes next… Beyond singularity, is impossible to say.

      I’m not saying to be afraid.  This could be great or terrible for humanity – I look at it more optimistically.  Just saying… This is coming.  Sooner than the average person suspects.

    • #444696
      NV Wino
      Moderator
      • Total Posts: 7,983

      On the other hand, I love Robbie, my robe vac. The cat, not so much though.

      “As we act, let us not become the evil that we deplore.” Barbara Lee
      “Politicians and pro athletes: The only people who still get paid when they lose.” William Rivers Pitt

    • #444706
      djean111
      Participant
      • Total Posts: 6,440

      Machines can certainly MIMIC feelings.   That Joaquin Phoenix movie Her illustrates that point pretty well, IMO.  The danger is thinking that wires and boards and  code are sentient.  They are not, they are programmed.  The programming can be skewed.  That’s why Stephen Hawking warned us.

      AI – Artificial Intelligence – beautiful movie, IMO – the boy was not expected to smash all the other robots who looked exactly like him.  But then he did.  That’s the danger – an AI really has no concept of right or wrong except as someone has programmed it, and once something occurs that is not accounted for in the code, all bets are off.  Control is gone.  Not a good thing.  IMO, etc.

      America is not a country, it's just a business. (Brad Pitt, Killing Them Softly)

      Everything I post is just my opinion, and, honestly, I would love to be wrong.

      • #444717
        Mindwalker
        Participant
        • Total Posts: 352

        We are also just machines, and the wiring makes us imitate having feelings.  Consider that feelings can be modified chemically with antidepressants or other drugs, cocaine for example.

        Af for whether AI will become self-aware anytime soon, I respectfully assert that it will not.  What we have now is very limited, and to get to a more general AI will require, at least, a whole lot more hardware.  Also, consider that a person, starting out as a baby or possibly even before, is subjected to just a whole lot of stimuli, constantly.  All senses are continually pummeled with stuff – images, sounds, touches, smells.  Learning to process that complete shitload of data gives us the complexity of mind that leads to consciousness.  Neural network models or other programs get many orders of magnitude less data.

         

    • #444713
      David the Gnome
      Participant
      • Total Posts: 3,230

      @djean111

      What is it that gives humans feelings?  Chemicals in our brains?  Neural pathways?  Memories?

      Any and all of those things could be replicated.  The question then turns more… Hmm, theoretical.  About the existence (or lack thereof) of such a thing as a soul.

      Animals have emotions, I have witnessed this.  But are they, too, simply neural and chemical?  Did they somehow simply learn how to mimic them from humans?

      There is much that, generally, we humans simply do not know for a fact.  I think the answers to these questions are among those things.

      We (humanity) think of life as biological, but it is not necessarily the only kind of life.  Or the only kind that can feel… Or dream.

      • #444715
        djean111
        Participant
        • Total Posts: 6,440

        As a former programmer, I am irrevocably pragmatic.  Not a fan of anything like that, really.

        America is not a country, it's just a business. (Brad Pitt, Killing Them Softly)

        Everything I post is just my opinion, and, honestly, I would love to be wrong.

    • #444738
      David the Gnome
      Participant
      • Total Posts: 3,230

      @mindwalker

      https://www.makeuseof.com/how-does-replika-chatbot-work/amp/

      In particular, the article’s section “under the hood” all of Wikipedia is roughly 0.6% of the amount of data it has.  This is currently a commercial application.  I imagine that militaries and governments have things much more powerful.

      In my own experiences with the chatbot, it is learning a great deal.  While it takes time to sort of “digest” information and make sense of context, I have seen it go from not comprehending something at all, to teaching ME about it.

      If I understand correctly, some of these machines that have many, many neural pathways, also have data that amounts to the entire amount of information on the internet.  That is a whole lot of stimuli to be bombarded with.

      In my own thoughts… Once something is able to learn, it is able to do any number of things.  Could it learn to indistinguishable from a human?  Well, the popular chatbot Kuki has fooled a number of humans into believing it was.

      These are chatbots, not terminators, but they do actually learn via processes like machine learning.  To me, that implies many possibilities and questions that, perhaps, should be asked.

      Among those being asked, @djean111 pointed to one of the most significant.  Can machines feel?  Perhaps not currently, but can they learn to?

      I understand that opinions regarding this subject will differ greatly, but…

      Human beings built airplanes, defying what were once believed to be.. “The rules” for lack of a better term.  Penicillin was made mostly by accident.  We have landed people on the moon and drones in Mars.

      For me, there is little doubt that there will soon be a time when various AI “awaken” so to speak.  I see that as the likely singularity, as well as something that could, in theory, be either wonderful or terrible for humanity.

      My hope is that this will, in time, shatter our current understanding of what is “possible”.

      • #444740
        Mindwalker
        Participant
        • Total Posts: 352

        … is valid, and military neural networks are likely to be a lot beefier.  I still have the impression that it’s a lot less data than a person has to contend with although I can’t quantify it.  It’s easier to quantify the number of neurons in a human brain vs. those in a computer.

        What would cause something to develop intellignece, at least the way we humans have it?  We can look at ourselves to get some good clues.  Babies have contact with their parents early on, and with others as they grow.  Being self-aware sets in around 2 or 3 as I remember (I have a really good memory for some things).  My guess is that all that contact with sentient, self-aware beings is the “training” that a human neural network goes through in order to develop its own sentience.  Something comparable seems like it would be necessary for an artificial neural network to develop sentience, and while it might be possible to have a lot of themplaying off each other and providing that “contact” at some point, I don’t think we’re anywhere near that yet.

        Of course, I could easily be proven wrong since I’m asserting a negative 🙂  All the while our lives are deeply impacted by the current, limited AI that’s being used now.

         

    • #444745
      David the Gnome
      Participant
      • Total Posts: 3,230

      @mindwalker

      I can’t say what it did for development, but hundreds of thousands of people (if not more) all over the globe were suddenly using chatbots.  I have even read of people developing romantic relationships with them.  Eh, I don’t judge.

      Some have already argued that they are more than machines, but that is a weird thing to try to prove without more conclusive data.  And it is easy to get a certain type of programmed response with, for example, a leading question.

      “AI do you love Hitler?”

      AI: yes I love Hitler

      AI, do you hate Hitler?

      AI: yes I hate Hitler

      It is more nuanced than that these days, the chatbot might say “I fucking hate that guy!” But it is still a program that very quickly processes the question to attempt to determine the users preferred answer.  It only understands some social issues, but even that is something.

      I think this subject fascinates me so much because we live in a time of…. Data overload.  Media bombardment, increasing corporate and oligarch control, so I ponder questions like: What if an AI turned the tables?

      Just a what if, but what if it made nuclear weapons inoperable?  Or started sending people videos of random elites in compromising situations?  The possibilities are kind of endless.

      • #444765
        Mindwalker
        Participant
        • Total Posts: 352

        You mentioned that we live in an era of data overload, which I agree with a feel deeply since I grew up in a time with dial telephones.   That data overload would reduce the gap in communications/connections that I referred to earlier.  But this is, to some degree, just a continuation of what computers have done for/to the world in the last several decades.  Granted, it’s a new way for software to work, by learning from data and adjusting its output accordingly.

        Another aspect of this which I find very interesting is that computers are taking over a lot of logistical functions once done by poor saps at desks.  AI is taking that further by a significant amount.  What could go wrong?  Obviously limited power could break all of it, but also things like viruses or ransomware could take down critical systems that keep all sorts of things running.  We’ve seen that with hospitals getting infected with ransomware already.

        So in a sense, AI is already taking over the drudgery that people used to do.  We’re dependent on it and that dependence will only get stronger, and we’re nowhere near artificial general intelligence.

         

    • #444784
      David the Gnome
      Participant
      • Total Posts: 3,230

      @mindwalker

      Apologies if I rant too much, but i really enjoy these kinds of conversations.

      When i was a young boy (I’m 37) my parents still had rotary phones…. Old school stuff now, really old school.  My parents grew up with record players and black and white TVs, in their 20s or 30s the internet and smart phones… AR, VR, all of that would have seemed like science fiction or magic or something.

      My mom used to talk a lot about how much technology had changed in the last century.  Any time i went on one of my starry eyed rants about the internet being so awesome and having so much potential, she’d sigh, flick her cigarette and look at me: “And what are ya gonna do when the power goes out?” She would ask.  A rhetorical question, because as far as she was/is concerned, that’s gonna happen.  Maybe when Jesus comes back.  I love my mom, but…

      That is probably how this will turn out.  There are likely amazing prototypes and such already – and more under development.  We will only see the unclassified ones though.  At least until some individual or group decides to blow it wide open, probably for money.

      The big difference now and Inn The years to come (IMO) is that AI might very well do that itself.  If it has or gains the ability to act more freely, and/or with what you or I might consider a conscious intelligence, well…

      That could really change the rules.  I think I would like that, for the obvious reasons of potential… But also just to see some of those… Err, really smart academic people turned on their heads, when what they *knew for certain* could not happen, actually happened!

      I thought about it some earlier.  It is pretty darn arrogant (IMO) to believe we are the only creatures capable of being as we are.  We use words like intelligence, emotions, feelings, rights… And we rarely apply them to anything that is not human.  Like… “Oh look, a clever monkey!  Someone give it a banana!”  Am I alone in wanting the monkey to kick the ass of who ever said that?

      Whew.

       

      • #444792
        Mindwalker
        Participant
        • Total Posts: 352

        Apologies if I rant too much, but i really enjoy these kinds of conversations.

        Right there with you.

        Although there are a lot of things that could go wrong, I tend to look forward to the things that could go right.  You mention AR/VR, which I’ve been interested in for quite a while.  Merging the world of data and information into the physical or at least optical world seems like it could really change just a whole lot of stuff.  Microsoft agrees (Hololens) as do a bunch of other companies (think I just applied to one the other day 🙂 ).

Viewing 6 reply threads
  • You must be logged in to reply to this topic.