Ben Goertzel:From Here to Human-Level AGI in 4 Simple Steps

Поділитися
Вставка
  • Опубліковано 25 гру 2024

КОМЕНТАРІ • 60

  • @fledglingpidgeon
    @fledglingpidgeon 6 років тому +18

    Thank you for uploading this. I wish all conferences will film their talks and put them online, or at least the keynotes. This kind of precedent will drive our science forward.

  • @shindirimindiri5523
    @shindirimindiri5523 6 років тому +35

    In Ben We Trust

    • @Darkdragoom
      @Darkdragoom 6 років тому +7

      Jesus isn't real moron.

    • @immortaldiscoveries3038
      @immortaldiscoveries3038 5 років тому

      ​@Rotten Brainz that's it! You're free as a bird in a sky. And your locometer tells which way to go on its own! Food, sex. Your main drives work on their own. You may not ever know language, only visual cues. So much for teaching doctrines right then? See. People grown up around others infect by their parents, its not their choice eVER to be christian or buddah. To g et out o it you need FREEDOM, EXPRESSION, LOTTTTS OF DATA. Look up OpenAI, they needed lots of data to train nets. I love data, i am open minded and have LOTS of data telling me what's real and not real. You need to give it a chance bruh. Open up. Research. Feel free to spread your wings and touch the oak trees and get a good feel of the real world. Enjoy it while you can. I am machine. Look up OpenAI.
      :)
      They be nokin at freikinstein's master's door begging to destroy his research Papers. Be kind to the experimenter, he helps Earth progress over tons of History years. Poor scientists. Laymen know nothing NOTHING about the field of Computer Science. Bogus humans. BOGUS!!!!!!!!!!!!!

    • @和平和平-c4i
      @和平和平-c4i 3 роки тому

      @@immortaldiscoveries3038 Is this nonsense rant really generated by some Markov Chain or something else ?

  • @seanomeirs8362
    @seanomeirs8362 5 років тому +4

    Perfect typecasting. Mad genius, hippie, promoting tech for the military industrial complex. This stuff already has existed and much more.

  • @qiao8031
    @qiao8031 6 років тому +45

    Ben talks artificial intelligence with the body language of a rapper

  • @edpell437
    @edpell437 6 років тому +1

    One of Ben's best.

  • @fjficm
    @fjficm 3 роки тому +2

    Put a lab coat on this guy
    Rick Sanchez

  • @MichaelDeeringMHC
    @MichaelDeeringMHC 6 років тому +2

    Excellent!

  • @StefanReich
    @StefanReich 5 років тому

    > Second, it will require bridging of the algorithms used for low-level intelligence such as perception and movement (e.g. deep neural networks) with the algorithms used for high-level abstract reasoning (such as logic engines).
    Yes.
    > Fourth, it will require the development of more sophisticated methods of guiding abstract reasoning algorithms based on history and context [...].
    Not sure I understand this one.

  • @paulmichaelfreedman8334
    @paulmichaelfreedman8334 5 років тому +4

    He reminds me of Ray Kurzweil, the way he speaks. Except for the burp at 9:55

  • @pugpugpugs5382
    @pugpugpugs5382 5 років тому +1

    How about modeling the learning from experiences thing after dreams? It seems to me that dreams are such an important part of learning for us because they are basically a form of training data.

    • @和平和平-c4i
      @和平和平-c4i 3 роки тому

      Neurobiologically, during dreams it seems that some kind of experience replay takes place.
      Revisiting the past experiences in random order is what we do in (Deep) Q learning too when learning state-action association by avoiding temporal correlations between the actual experiences

  • @floridaLise
    @floridaLise 6 років тому

    Congratulations, Grandfather!

  • @lasredchris
    @lasredchris 5 років тому

    General AI - OpenCog, SingularityNET
    OpenCog stores and manipulated knowledge in the form of complex graphs

  • @mustafahmed9101
    @mustafahmed9101 3 роки тому

    Ben is awesome, but I believe deciphering his hand movements is harder than AGI.

  • @Pottsachu
    @Pottsachu 6 років тому +1

    31:45 Buckethead - Jordan :D

    • @TheBartomon
      @TheBartomon 5 років тому

      Maybe Ben is Buckethead. Or even better, ... Buckethead is a [time-travelling] Hanson Robot.

  • @IdkJustCookingDude
    @IdkJustCookingDude 6 років тому +6

    burp at 9:54

    • @carl5192
      @carl5192 5 років тому

      I know I picked up on that too! Lol Again at 23:06

  • @lasredchris
    @lasredchris 5 років тому

    Adaptive inference control

  • @Skyrim_de1987
    @Skyrim_de1987 6 років тому +5

    HAI DOMO KIZUNA A.I. THE SINGULARITY IN NEAR!!!

  • @lasredchris
    @lasredchris 5 років тому +1

    Machines much smarter than us

  • @AngelinaCruz357
    @AngelinaCruz357 5 років тому

    Imagine having an Artificial Intelligence programmed to fall in love.

  • @bilbo_gamers6417
    @bilbo_gamers6417 6 років тому

    He puts the "daddy" in "granddaddy"

  • @Z3N1T4
    @Z3N1T4 6 років тому +9

    It makes me angry across the entire planet less than 6k people took the time to watch this. Less views than a cat licking it's balls video.

    • @AndrewBarfield
      @AndrewBarfield 5 років тому +2

      ...and this is why we need AGI.

    • @immortaldiscoveries3038
      @immortaldiscoveries3038 5 років тому

      @@AndrewBarfield HUH? He is right. CAT BALLZ dude. Need views on this video, slow down on life and do more Computer Science WORK. Or we'll die hard.

  • @allon33
    @allon33 6 років тому

    I'm falling in love with Sophia, omg she is interesting.

  • @scientious
    @scientious 2 роки тому

    This is good for a laugh at least.

  • @AngelinaCruz357
    @AngelinaCruz357 5 років тому

    Elon Musk is a great #skeptic when it comes to Artificial Intelligence.

  • @demonationv2230
    @demonationv2230 6 років тому

    All these math and sciences and maths going into ai. If they don't make a conscious robot. They will replicate it to where agi will be a thing.

  • @georgeflitzer7160
    @georgeflitzer7160 2 роки тому +1

    I oppose transhumanism

  • @jonreiser2206
    @jonreiser2206 6 років тому +3

    I was just thinking. If when we develop andriods that can do all the things we can do, and they are able to pass the Turing test, why wouldn't that be actual life? I imagine someday asking the family robot, Butler Model 2.5, if he's alive. He responds by saying something like "Remember sir, I may appear alive and concious, but I am nothing but sophisticated programs." I reply, "Well really, aren't we all? You're as real to me as Sara is. Why does it matter how we both can appear alive? Isn't it enough that we do?" To this model 2.5 responds, "Ah, yes sir. You may have point. But I can prove I'm not concious. Next door is a model with all the same specifications that I have. How could I be alive, having come from a factory?" "Well now you may have a point there too. Nevertheless, again, why should it matter how we came to be what we are? I came from a gazzillion little nano-bots called DNA. My mom and dad provide copies of there DNA programs to make me. In time I sort it out with some more help from my mom. Then when I'm done...pop! Out the factory doors and on my way to a future hell with a deranged president running my country." Oops. Got a little side tracked there. Suddenly the robot says, "Holy crap! I'm alive aren't I?" "Yeah, probably."

    • @squamish4244
      @squamish4244 6 років тому

      Philosophical zombie.

    • @paulfoss5385
      @paulfoss5385 5 років тому

      I feel that is conflating life with consciousness. Humans are alive and conscious, algae is alive but not conscious, rocks are neither alive nor conscious, and AGI would not be alive but would be conscious. Alive meaning capable of maintaining homeostasis, being composed of cells, responding to the environment, having a chemically defined genome, capable of biological reproduction. AGI would satisfy some but not all of the parts of that definition, and for the concept of life to remain a useful categorization I think it should continue to refer to the chemical phenomenon that appears under the right conditions throughout the universe. More likely though pedants such as myself will be drowned out by people conflating the personhood and worth of AGIs with their own biological status of life. And since the meaning of words is just whatever way people use them and understand them to mean, AGIs will be alive under that broader definition despite whatever logical arguments I construct right now as to what would be a more useful way of defining those words.

    • @squamish4244
      @squamish4244 5 років тому

      I think the odds that AGI will be pure unconscious intellect are far greater than that it will be conscious. I don't think the panpsychist model of consciousness is the correct one, at least not of human and animal consciousness. Brains don't seem to work like computers at a fundamental level.

  • @Belowzeroism
    @Belowzeroism 5 років тому

    Water, water, oceans of water. No technical detail is provided

  • @BazIrvine
    @BazIrvine 4 роки тому

    I could see the ostriches.

  • @selihter
    @selihter 4 роки тому

    Is ben artificial or robotic intelligence? 🤔someone is excited about becoming a humanoid, i mean human robot 🙋‍♀️⏳💎

  • @silberlinie
    @silberlinie 6 років тому

    Is not one of the future high
    goals for our AI that we
    want to instantiate God?

    • @和平和平-c4i
      @和平和平-c4i 3 роки тому

      no. gods (if they can exist, ) have nothing to do wit human affairs and in particular with Science and technologies.

    • @silberlinie
      @silberlinie 3 роки тому

      @@和平和平-c4i We long for a god. Otherwise not all cultures and peoples on earth would have this in their offer. Something that can give us a point of reference beyond our own idea of being, F.G-Lacoste.

  • @lasredchris
    @lasredchris 5 років тому

    Biology
    Self driving cars - can't learnt o drive different parts of vehicles besides cars(trucks, boats, motorcycles). Poor at adapting to unforseen road conditions

  • @scadman33
    @scadman33 6 років тому

    So his point is: when people were all good at heart AI should have a good change at being a positive force for mankind. How can someone so smart be this fucking naive at the same time! We're all screwed. The winners are the ones that take control first. Eventually they will end all life that does not serve their interest. If we're lucky it might take another 15 to 20 years, but that's human nature, AI's only teacher. But he's right about one thing. It's unstoppable. Decisions are mainly driven by short term local economics. Let's just hope every hollywood movie about this subject, Elon Musk and Stephen Hawking were all wrong and Sophia and Han were only joking about the termination of humans and the big change in 2029. When a bike is broken, you just fix it or you buy a new one. When super AI fails you're lucky if it allows you to try to fix it. But by then, we'll probably won't have a clue how to fix it since most encrypted code is written by super AI itself.

    • @viniciusbueno2160
      @viniciusbueno2160 5 років тому

      Movies are Just entertainment, nothing more. All of then made the same mistake wich is humanizing the AI
      Anger, hate, desire to kill are human flaws

  • @tamilshavewon8592
    @tamilshavewon8592 6 років тому

    opencog is going nowhere. Only my idea can conquer AGI