Is a technological singularity ever going to happen?

Discussion in 'The Guru's Pub' started by Cybermancer, Nov 4, 2009.

  1. Cybermancer

    Cybermancer Don Quixote

    Messages:
    13,795
    Likes Received:
    0
    GPU:
    BFG GTX260OC (192 SP)
    Yup. That's what AGI would be able to do though since it isn't limited to the same processing speed or storage capacity like we humans are, for example.

    (Sorry, but I can't elaborate more on it right now since it's already 3:15 am here and I'm more than overdue for some sleep.)
     
  2. hallryu

    hallryu Don Altobello

    Messages:
    11,386
    Likes Received:
    14
    GPU:
    2x HD7970
    I think part of the point is the fact that computers may become so fast that they don't need to do 'critcal thinking' becuase they will be able to compute every possible outcomes from a single event. Or something like that:D
     
  3. dukedave5200

    dukedave5200 Ancient Guru

    Messages:
    3,192
    Likes Received:
    0
    GPU:
    BFG GTX 285 OCX - SLI

    Haha, think you took my post too seriously there Wicky. ;)

    Besides, "I, Robot" the movie was an extremely trimmed down version of Asimov's Robot series. Moreover the movie didn't really reflect the content of the books anyway (it was more of just the idea of the books) - if you read the books you might have a better understanding of what Asimov had in mind and how the three laws worked. But no mater what, it was all COMPLETELY science fiction and not something that has ever been in reality.
     
  4. Makalu

    Makalu Ancient Guru

    Messages:
    4,197
    Likes Received:
    2
    GPU:
    EVGA 8800 Ultra
    A computer can't think or know...much less know that it knows (self-awareness). A computer program designing a CPU doesn't know that it's designing a CPU...it just manipulates symbols according to built in rules but it doesn't know what the symbols mean and how they correlate to reality. Human thoughts have meanings that represent things and we know the things they represent.

    These scientists are working under the delusion that their minds are mechanical and logical and operate under laws and percieve reality objectively much like their science and can be modeled as such. The truth is that their minds are curious, choosing, creating, combining, discriminating, striving, dreaming, advancing, idealizing, passionate, aesthetic, ethical and moral and all of reality is subjective and filtered through the patterns of the mind before it becomes real including the laws of science.

    The mind of the scientist is not mechanistic at all...only the science is mechanistic but there's no science without the scientist.
     

  5. mike41

    mike41 Maha Guru

    Messages:
    1,208
    Likes Received:
    0
    GPU:
    Evga 560ti
    the only way i can see computers taking over is if they can somehow infuse like a living brain into a huge computer data base that can posses feeling just like humans. computers arent living things, and wont hate or like. they are just a type of machine.

    i also dont think sungularity will happen till the economy demands us to move that direction. right now its control of oils. maybe when were ****ed ando ut of oil well move to space, and will require some crazy stuff to manage to travel and map out space
     
  6. Wicky

    Wicky Maha Guru

    Messages:
    1,092
    Likes Received:
    0
    GPU:
    Sapphire Radeon 5850
    Well yes, singularity is going to happen. One day, we will have computers that are infinitely fast. But it will be no biggie. Besides playing every game at any resolution, and calculating the billionth digit of PI in a nanosecond, what more is there in computing?
    Not much I'd say.
     
  7. Cybermancer

    Cybermancer Don Quixote

    Messages:
    13,795
    Likes Received:
    0
    GPU:
    BFG GTX260OC (192 SP)
    You're proceeding on the assumption of current computers and current capabilities. Do you have any sources that back up your claim that it's going to be impossible for an AGI to be conscious or self-aware?

    Did you read any of this information or watch the video?



    • "The Singularity Is Near" by Ray Kurzweil (2005 New York Times Best Seller, http://singularity.com/)
      (A book describing the phenomenon of a technological singularity and what could lead to it.)

    Of course both of our opinions are impossible to prove right now but scientists are currently working on reverse engineering the human brain and at some point will have created a functioning model based on this information. Why do you think this model - which is going to be nothing less than a human brain - won't have the same capabilities as we do? You do know that DNA is digital information, right? DNA is nothing but a storage method/medium for information. The human genome contains the genetic information how to build our organs and the the whole body - including the brain.

    Just a few decades ago we thought it would be impossible to even create narrow AI - something that already has become part of our daily lives now. Just because something is impossible today doesn't mean it's still impossible tomorrow. History is full of such examples.

    [​IMG]
     
  8. nvlddmkm

    nvlddmkm Banned

    Messages:
    4,189
    Likes Received:
    0
    GPU:
    EVGA GTX285
    How would imagination and creativity be engineered into a artificial consciousness? Would it also be considered to have a "spirit"? How would emotions be engineered as well. Sorry, just some questions for fodder. :)

    Cybermancer, err, Eugene, aka Doktor!...:D Cool cartoon.
     
  9. Dustpuppy

    Dustpuppy Ancient Guru

    Messages:
    4,146
    Likes Received:
    0
    GPU:
    integrated - fffffffuuuuu
    I'm only familiar with a few areas of AI that would be applicable to creativity. In simulated annealing a certain degree of randomness is introduced to cause an AI algorithm to spontaneously evaluate values outside of any local maxima it might find itself trapped in. As time progresses the heat level of the algorithm is reduced causing it to 'cool' into shape by focusing its calculations on exploiting any newly detected maxima.

    Emotions are an animal/human thing. They could be useful to an AI but an AI wouldn't need them to be considered intelligent. I suppose emotions could be analagous to a subset of rules governing information exchange between cells in cellular automata.
     
  10. Psytek

    Psytek Ancient Guru

    Messages:
    3,370
    Likes Received:
    0
    GPU:
    2x 260 GTX 216 SLI

    This could only happen in a utopian, meritocratically governed society.
    In a capitalist, corporate based society, technical advances are withheld by companies in order to release them steadily to maintain profits. Experimentation and invention as a pure academic activity is decreasing and corporate sponsored research is increasing. Society is becoming more consumerist and governments are becoming more controlling. So no, this will never happen.
     

  11. Makalu

    Makalu Ancient Guru

    Messages:
    4,197
    Likes Received:
    2
    GPU:
    EVGA 8800 Ultra
    Well I don't feel that I need to appeal to any other source besides common sense. Nobody even knows how it is that we humans are conscious and self-aware...no amount of examining and analyzing the lump of fatty tissue and electrochemical connectors of the brain has ever come anywhere close to explaining consciousness and personal identity. We just know that it's an innate part of human nature and no other animal life form on earth has it and most certainly no evidence that any of the machines and tools we have ever created have it.

    So there's no historical precedence for a conscious machine...no model for human consciousness itself...and no way to ever even know if the machine is conscious or not. They are just assuming that this will miraculously appear...why? because they saw it in a science fiction movie once? All they are trying to do in the field of AI is create a machine that externally mimics the more mechanical types of human thought not create a thinking machine.

    It doesn't matter how close they come to a computational model of the physical brain because the abstract mind is not the same thing as the physical brain and most of the thought processes I listed that make us uniquely human (as opposed to animal or machine) can't be reduced to a sophisticated computation anyway.

    If the brain is just programmable hardware and the mind is just a software program then how will they ever possibly program free-will? It's an oxymoron.
     
  12. Dustpuppy

    Dustpuppy Ancient Guru

    Messages:
    4,146
    Likes Received:
    0
    GPU:
    integrated - fffffffuuuuu

    Well funny thing this, do you remember the scene at the end of Steven Spielberg's AI where the kid enters an infinite loop? If you completely whack a humans short term memory they appear to enter a kind of loop.
     
  13. Makalu

    Makalu Ancient Guru

    Messages:
    4,197
    Likes Received:
    2
    GPU:
    EVGA 8800 Ultra
    no I don't remember that...don't think I know the movie at all. If the goal is modeling a defective brain then yeah computers are great at getting stuck in a loop...when they stick a computer in Kurzweil they'll have an infinite fruit loop ;)
     
  14. Cybermancer

    Cybermancer Don Quixote

    Messages:
    13,795
    Likes Received:
    0
    GPU:
    BFG GTX260OC (192 SP)
    Even if this is true, we are still following a graph of exponential growth. This graph wasn't even affected by both World Wars, economical depressions, market booms or bursting bubbles:

    [​IMG]
    (Check as well: http://img263.imageshack.us/img263/7941/internetbackbonebandwid.jpg, http://img263.imageshack.us/img263/7571/magneticdatastorage.jpg, http://img171.imageshack.us/img171/2986/randomaccessmemory.jpg, http://img410.imageshack.us/img410/543/exponentialgrowthkurzwe.gif)

    You might argue that the data used to draw that graph is handpicked and the data not fitting it is omitted but even taking 15 different and completely random sets show the exact same outcome:

    [​IMG]
    (logarithmic plot, btw)

    This is not data taken from some kind of physics experiment, but real world (market) data. Yet, it follows the graph remarkably well - to a point where it's almost scary, imo.
    As I already posted earlier, this is a philosophical issue and not a technological. You and I both can't objectively determine if someone is conscious or not. If I say that I'm conscious you can either believe me or not. You can't prove either one, though. If a software instead of me would be posting this you couldn't tell the difference either. For example, if I am saying that I'm conscious - which I think I am - you can either believe me or not. You could ask me questions and I could answer them but in the end it still would not be a proof for either one.
    Of course there's no historic precedence. We are only now starting to approach this (potential) point. This goes back to my point above, though. If you are going to communicate with someone and he or she responds like a human you can either assume that the claim of being conscious is correct or not. If I would ask you this question right now, would you think I am conscious? What if you would find out that I am an AGI? Would you just because of that deny me the concession of being conscious or self-aware?
    That's your opinion, though - which I absolutely respect and appreciate, btw. I'm just thinking we're running in circles or talking at cross purposes here.

    Imo, the advancement in AGI will be a gradual one. Just like a newborn is unable to communicate with us and doesn't have any understanding about its surroundings or the world, it has to be taught what is right or wrong, watch us to figure out what speech is, learn how to walk, etc., etc. an AGI more than likely is going to have to go through the same. Current successful "narrow" AI is using this same principle already. It's for example impossible for a software to have all different dialects or pronunciations of a large vocabluary in some kind of database so that it can understand human speech. Natural language processing AI is capable of this though after an initial training period.

    Maybe I should add this quick remark as well: I really appreciate all the feedback I received so far. I think it's important to be open to new ideas and other opinions so that you can adjust yours and might see different angles you weren't aware of before. :)
     
  15. Makalu

    Makalu Ancient Guru

    Messages:
    4,197
    Likes Received:
    2
    GPU:
    EVGA 8800 Ultra
    well regardless of the philosophy they still don't have a technological model for machine consciousness...they just assume that if they model the individual parts of the brain it will spring forth...I assume that the whole is greater than the sum of the parts and it won't.

    I think a lot of the science he proposes isn't feasible at this time...there's a lot of quasi-science and revolutionary future changes in areas that aren't moving that direction at all. This is all way more of a philosophy than a science really...Kurzweil is afraid to die and technology is the religion that will grant him immortality. If it all works out like he thinks he will still be afraid to die though...
     
    Last edited: Nov 7, 2009

  16. Cybermancer

    Cybermancer Don Quixote

    Messages:
    13,795
    Likes Received:
    0
    GPU:
    BFG GTX260OC (192 SP)
    Maybe just this: While working on the light bulb, Edison tried nearly 2000 ideas before he got it to work. When asked about it, Edison replied that he now knew 2000 ways not to build a light bulb. :D
    Also: Thomas Watson, founder and president of IBM, claimed that "The way to succeed is to double your failure rate."
    This is not about the prediction when the next-gen iPod is coming out but an obvious and well researched trend of technology in general. Of course it's impossible to predict the future but you can make well educated guesses based on currently available data and trends already lasting decades or even centuries.
    Sorry, but this is kind of a blow below the belt. Who's saying this is a religion? You do know that he's NOT the only one predicting this phenomenon currently called "technological singularity", right? He wasn't even the first one. There were lots of other people before him. Several decades ago. Even if he's afraid to die, what's the big deal with that? Are you looking forward to it? Besides that I pretty much agree with what Aaron Saenz wrote: "Ray Kurzweil may be a nut, he may be going through the biggest mid-life crisis ever. I don't know. I honestly don't care much. I want to evaluate his predictions, and his actions, on their own merit."
     
  17. Makalu

    Makalu Ancient Guru

    Messages:
    4,197
    Likes Received:
    2
    GPU:
    EVGA 8800 Ultra
    Right...Edison didn't invent the light bulb...a working light bulb was invented 50 years earlier and the science behind it was understood long before Edison perfected a marketable bulb and complete infrastructure for it. But nobody has yet invented even a rudimentary machine consciousness or even demonstrated a valid paradigm for one...it's all just fantasy at the moment.

    Right...I've been following AI for decades now and the trend is...well we can debate about the success so far but I don't think you'll find anybody who says that the growth in the field has been exponential and many people agree that any breakthrough won't happen under our current architecture. And from what I understand the medical science behind his immortality claims is lacking too. Kurzweil makes these big claims about the future of science that don't hold up to what is currently happening in the field...so at the moment it's all much closer to science fiction then it is to science.

    Well I don't think it's a low blow at all. He claims that the task of religion is to assuage the fear of death and should be swept aside to make room for the technological "triumph" over death. He also claims that he is going to raise his father from the dead. There's no doubt that this is presented as an alternative religious philosophy...just not a very good one. But it's good enough for Ray and some of the other transhumanists yes.

    And no I'm not afraid to die...why should I be? everbody dies...I'm quite comfortable with my own mortality and just take life one day at a time. Kurzweil thinks he can cheat death...g'luck with that. :)
     
  18. Cybermancer

    Cybermancer Don Quixote

    Messages:
    13,795
    Likes Received:
    0
    GPU:
    BFG GTX260OC (192 SP)
    Of course this is a highly controversial topic, Makalu. This was actually one of the reasons why I started a thread about it as there is more than one side and aspect to it. Would AGI already exist we wouldn't have to discuss its possibility and/or feasibility.

    Whether Edison invented the light bulb or not (I didn't even say that he did) was not the reason why I decided to post his statement in regards to that it was not an easy task to get it to work - just like the one from founder and president of IBM, Thomas Watson. Instead I was trying to highlight that it takes creativity, patience, dedication and obviously not being afraid of setbacks or failures to invent and create something new.

    I also already posted above that Kurzweil might be considered a nut by some, as a few of his ideas and theories are being highly questionable, but on the other hand he is also a very successful inventor (principal developer of the first CCD flat-bed scanner, the first omni-font optical character recognition, the first print-to-speech reading machine for the blind, the first text-to-speech synthesizer, the first music synthesizer capable of recreating the grand piano and other orchestral instruments, and the first commercially marketed large-vocabulary speech recognition), adviser (Army Science Advisory Board, United States Congress, American presidents, etc.), thinker, futurist and received more than one recognition for his work (recipient of the $500,000 MIT-Lemelson Prize, the world's largest for innovation, National Medal of Technology, the nation's highest honor in technology, from President Clinton in a White House ceremony, inducted into the National Inventor's Hall of Fame , established by the US Patent Office, received nineteen honorary Doctorates and honors from three U.S. presidents). He is not the only one though in support of this "prediction", for lack of a better term.

    Something interesting and only somewhat related to this topic, as it is not about AGI but about using a highly detailed and sophisticated model of the brain for medical purposes and simulations, is this short 15 minute video, showcasing the Blue Brain Project (http://bluebrain.epfl.ch/) by IBM and Professor Henry Markram:

    Maybe this topic about longevity or Methusalarity, as you already brought it up, Makalu, is something new for some people and should be considered as an aspect of a potential technological singularity as well:

     
  19. Dustpuppy

    Dustpuppy Ancient Guru

    Messages:
    4,146
    Likes Received:
    0
    GPU:
    integrated - fffffffuuuuu
    Well if you wish to argue to his person and appeal to his authority I'm going to have to draw an analogy here. Sir Isaac Newton was a brilliant scientist by day, and a rabid alchemist by night his science was right, but his science being right couldn't make his alchemy real.


    I've said it before and I'll say it again, the exponential growth of technology is attributable to

    1. Population growth being exponential (responsible until computers)

    2. The geometry of a 2 dimensional object. If computers were 3d and layers were shrinkable then moores law would have a higher multiplier.

    Neither of these things is set to continue indefinitely and population growth has already leveled off compared to the 1900's.
     
  20. seaplane pilot

    seaplane pilot Maha Guru

    Messages:
    1,295
    Likes Received:
    2
    GPU:
    2080Ti Strix
    Damn! I will be 86 by the time this happens. If I am still among the living or have not yet uploaded my soul into the Ether- Net
     

Share This Page