Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Do you think Artificial General Intelligence will happen this century

Options
  • 26-04-2019 9:34pm
    #1
    Registered Users Posts: 2,732 ✭✭✭


    Artificial general intelligence (AGI) is the intelligence of a machine that could successfully perform any intellectual task that a human being can

    Do you think Artificial General Intelligence will happen this century 9 votes

    Atari jaguar>Human
    11% 1 vote
    Human >Atari jaguar
    88% 8 votes
    Post edited by New Home on


Comments

  • Closed Accounts Posts: 7,277 ✭✭✭Your Face


    I just want a snooty robot butler.


  • Registered Users Posts: 20,411 ✭✭✭✭kneemos


    As long as they don't want to be regarded as sentient beings.
    That **** will just give them ideas.


  • Registered Users Posts: 3,685 ✭✭✭Signore Fancy Pants


    No.

    My Tamagotchi kept starving itself to death. AI is a myth.


  • Registered Users Posts: 733 ✭✭✭milehip


    sxt wrote: »
    Artificial general intelligence (AGI) is the intelligence of a machine that could successfully perform any intellectual task that a human being can

    Judging by the above statement and the thread title, I'd say the singularity can't come fast enough.


  • Registered Users Posts: 470 ✭✭The Oort Cloud


    It all depends. What is a Cenrury ?

    Individual people have different thoughts and understanding in regard to others opinions, but the problem is this... there are some people out there that will do everything in their power to cut you off when they do not like your opinion even when it is truth.

    https://youtu.be/v8EseBe4eIU



  • Advertisement
  • Registered Users Posts: 470 ✭✭The Oort Cloud


    A.I will be capable of doing your job, but I reckon it will be 25 years before it really makes us all worry.

    Hacking the A.I systems will be a big worry for sure, but a quantum A.I system might keep the lunitics out of the systems if implemented properly.

    Individual people have different thoughts and understanding in regard to others opinions, but the problem is this... there are some people out there that will do everything in their power to cut you off when they do not like your opinion even when it is truth.

    https://youtu.be/v8EseBe4eIU



  • Registered Users Posts: 7,055 ✭✭✭JohnnyFlash


    No. Life 3.0 - Being Human in the Age of Artificial Intelligence by Max Tegmark is a great book on the subject. He speculates on how it might happen, but leaves the nuances of why it should happen to the reader. General intelligence is a very loaded phrase - what does that mean? The ability to feel empathy, or the ability to achieve a human view on what consciousness is? Or the opposite?

    Microsoft have published many fine papers on the ethics of all this area of research for those that are interested. Fascinating topic, and one we should all have an opinion on.


  • Registered Users Posts: 5,083 ✭✭✭Rubberchikken


    god i hope so.
    anything has to be better than most of the humans i encounter on a daily basis.


  • Posts: 11,614 [Deleted User]


    It all depends. What is a Cenrury ?

    In the next 15 minutes he means. I'm going to say no.


  • Registered Users Posts: 20,553 ✭✭✭✭Dempsey


    AI has become a marketing term for what is really some very basic algorithms up to complex implementations to achieve basic functions as done by humans

    AI as people think of in films/pop culture is decades away

    Think of the code reviews!!


  • Advertisement
  • Registered Users Posts: 951 ✭✭✭Neames


    I doubt it will happen this cenrury.


    Alexa told me it will take 2 centons.


  • Registered Users Posts: 470 ✭✭The Oort Cloud


    Accountants and lawyers will be the first to go, happy days, the A.I will be better at giving the correct sentences for specific crimes. Criminals held in solitary confinement and dictated to by A.I 'non sympathetic systems. Joey, lights will be turned off in 30 seconds, if you cause a nuisance I will reconfigure your xbox-one to only allow societal education videos for you to learn. Do you understand Joey. 10 seconds before light extinguishment. Do you comply Joey.... Waiting.... Lights have been deactivated, cell doors sound-sealed.

    Individual people have different thoughts and understanding in regard to others opinions, but the problem is this... there are some people out there that will do everything in their power to cut you off when they do not like your opinion even when it is truth.

    https://youtu.be/v8EseBe4eIU



  • Registered Users Posts: 2,338 ✭✭✭Bit cynical


    Got access to that DALL-E text to picture neural network from OpenAI. I imagine they will be increasing access rapidly now. Up to this point it had mainly been techology enthusiast bloggers and youtubers who have tended to be very positive about AI and have shown mainly "cute" pictures but not really tested the product.

    Anyway here's a few I did to test out the algorithm:

    Prompt: Irish politician floating in water.

    Prompt: Alexander the Great eating Supermacs burger in Dublin.

    Historical figures tend to be statues if that is the common representation of the person. The AI doesn't seem to know how to transform the statue into a naturalistic person. Treats the statue as the thing itself rather than a representation of the thing.

    Prompt: Photo of robot with human eyes and human soul.

    Prompt: Horse made with human parts.

    The parts don't seem recognizably human but I think the image on the left is quite imaginative.

    Prompt: Robot made out of human flesh.

    Teeth seem to be flesh coloured.

    Prompt: Horse made out of human hair.

    Interesting the way seems to be arranged in a very natural way as if it has just fallen into place by accident.

    Prompt: Horse made out of human eyes.

    On the left the AI seems to have interpreted this as horse with human eyes rather than constructed from human eyes. Seems to be something funny going on in the mouth on the left. Don't know what is going on in the picture on the right but interestingly, the two faces are not perfect mirror images of each other.

    Prompt: Photo of an AI made of human souls.



  • Registered Users Posts: 2,338 ✭✭✭Bit cynical


    How I think AI will manifest itself initially is that it will mainly be an online phenomenon. It has already been reported that a lot of twitter posts are AI-written and this will increase. More and more content on the Internet will be AI generated. You tube videos will be created with an AI narrator and illustrations created by something like DALL-E. It may seem a bit 'off' initially but will evolve compelling content to attract viewers using reinforcement learning. Even if it is not brilliant, it will be good enough and will be very cheap to produce.



  • Registered Users Posts: 606 ✭✭✭Sonic the Shaghog


    This!

    Most of the bloody time the ones talking of AI are media types who haven't a clue what it even means!



  • Registered Users Posts: 5,988 ✭✭✭TheIrishGrover


    Yeah. There is a difference between algorithms, no matter how complex and genuine AI that can make intuitive leaps in logic. Hate referring to Boardroom Bullsh*t Bingo but true AI would involve @thinking outside the box. Making leaps in logic that are not dependent upon existing algorithms. There is that famous "Go" tournament where the "AI" seemed to make some crazy move. However subsequent investigations showed that this move, while unexpected, could eventually be boiled down to initial algorithms or subsequent generations/iterations.

    I mean, I work in a field that uses AI quite a lot but, despite all the data and computing power they possess, they can only get you so far. sure, they can get you 90% there. But that last 10% is ALWAYS going to be critical and out of the reach of AI.

    However, do I think that something will be able to pass the Turing Test? Yes. We are not there yet - It is quite easy to get these into loops by using trigger words (Much like the Biden and Trump threads. Just joking, it is AH after all 😀). I think that, using current methods, we will get to a point where they can pass the Turing Test. I don't believe we will come to a point of "Critical Mass" where we suddenly switch from incredibly complex algorithms to. BAM!!! AI

    Having said that, I always qualify these things: I always think that there is the potential for new thinking.

    Tomorrow, someone may come up with some insane, radical, brand new thinking and, suddenly: Next day - AI. In the same way that I don't believe we will ever see interstellar travel given current theories/evidence. BUT, again, someone could suddenly think "Well, what if we did THIS" And suddenly wormholes.

    Post edited by TheIrishGrover on


  • Registered Users Posts: 16,153 ✭✭✭✭Grayson


    My feeling is that if someone asks in the next 5-10 years, it's unlikely. But before the end of the century, entirely possible. We've been making leaps and bounds in computing in the last 40 years. Given another 80, who know's where we'll be.



  • Registered Users Posts: 16,153 ✭✭✭✭Grayson


    How did you get access. I was looking at it and could only find the crappy clones.



  • Registered Users Posts: 19,845 ✭✭✭✭cnocbui


    Not going to happen before humans go into orbit around an exoplanet.

    I believe driving a car requires general intelligence.



  • Registered Users Posts: 10,563 ✭✭✭✭EmmetSpiceland


    No, and we won’t be hitting some “singularity”, where we end up getting downloaded into cubes, anytime soon either, as some of the, more hysterical, nerds would have you believe.

    “It is not blood that makes you Irish but a willingness to be part of the Irish nation” - Thomas Davis



  • Advertisement
  • Registered Users Posts: 2,338 ✭✭✭Bit cynical


    I think it was this link I signed up with. I think I put academic researcher as role and did not include any links.

    I think however there has been a marketing strategy of giving priority to bloggers and youtubers that enthuse about technology. These people are more likely to not really test the system preferring instead funny or cute images like pandas on the moon type of thing.



  • Registered Users Posts: 552 ✭✭✭BaywatchHQ


    I've seen videos of advanced sex dolls who you can converse with. I don't think I would ever buy one due to the size and awkwardness of having one. I heard that in Amsterdam there is a sex doll brothel.



  • Registered Users Posts: 11,758 ✭✭✭✭BattleCorp


    There's still almost 80 years left in this century. Imagine how far we've come in the last 80 years. I'd say it's possible.



  • Registered Users Posts: 5,988 ✭✭✭TheIrishGrover


    I don't know who said it or where I saw it but I remember seeing something on TV a few years ago about the advancement of technology in the last 100 years: If you took someone from about 2000 years ago, you could bring them right up to the 19th century and explain stuff to them. Even the likes of steam-power: Steam builds here. That pushes this and then everything else is cogs and wheels etc. But if you took someone from 1920s. Someone used to electricity and radio, movies etc. And brought them to 2020 and tried to explain The Internet, computing, CAT scans, MRIs, electronics in general, they would be totally lost.

    So, again, IMHO, I don't think we will come to a crunch point using existing methods that will manifest genuine AI. I think there will be a point where it can be simulated almost flawlessly though. I think if AI is created it will be using a method we are currently unaware of. Some brand new way of thinking about computing that suddenly opens up an entire new path forward. Not just regarding AI but physics and chemistry. A new renaissance.... But it would be such an alien thought that we won't know it until we find it.



  • Registered Users Posts: 2,338 ✭✭✭Bit cynical



    Very true. I remember reading an article featuring street from New York: one from I think 1910 and the other from 1920. In the first, there were almost no cars but many horse-drawn vehicles, and in the second almost all cars and no horses. I might have got the years wrong but the interval was about 10 years in which horses were almost completely replaced. To most 1910 people the changes that would take only ten years to happen would have been unimaginable and, were they to be imagined, the timescale for the change would be much longer.

    1910: Mainly horses

    1920: Few horses

    The technology had been around for a while even before 1910 but certain improvements needed to be made and then very swiftly the technology took over. Motor vehicles up to that point would have been seen as a niche curiosity but not something that would ever become the mainstream norm. Once the technology got to a certain point it very quickly dominated even though the underlying principles had been around for some decades.

    Post edited by Bit cynical on


Advertisement