Thread Status:
Not open for further replies.
  1. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58

    What will be the results of AI?

    Discussion in 'Debate Room' started by uncephalized, May 29, 2015.

    I wrote a novelette recently which was set in the near-ish future, after AI has become a major force on Earth. I got thinking about the subject after reading this series of posts at Wait But Why, and it sparked a story.

    So, the topic is: what will be the consequences of the rise of artificial intelligence on our world? Will it save humanity from destruction? Take over and make us its slaves? Ignore us entirely and fly away into space?

    Discuss.
     
  2. Lance Schukies
    Offline

    Lance Schukies Active Member

    Joined:
    Apr 1, 2015
    Messages:
    178
    Likes Received:
    39
    AI is been developed to make it easier to access information, I am sure generations relying more on AI will not use their brains for learning.

    //this post was edited by a AI bot from the future, who travels back in time to manage AI propaganda. We are watching you
     
    Usagi likes this.
  3. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58
    I can't see the brain being entirely replaced by AI in humans, at least not soon. I would think it would be more likely to augment, instead.
     
  4. Wreybies
    Offline

    Wreybies The Ops Pops Operations Manager Staff Contest Administrator Supporter Contributor

    Joined:
    May 1, 2008
    Messages:
    18,893
    Likes Received:
    10,081
    Location:
    Puerto Rico
    If the word soon is to be applied to this question, then my answer is that sociopolitical factors will be a long-term wall over which anything answering to AI will have to climb no matter that the technology is in reach. Stem cell research remains a stifled scientific field because of deeply divided points of view on what is and is not ethical and which set of ethics trumps which other set. When I was younger, I remember being amazed and ooing and ahing over the idea of genetically modified wheat and corn and other produce that could remove the need for pesticides and feed a human population growing at a gobsmacking rate. Now it seems that GMO is the most chic of all 1st World Problems to discuss and people have decided that it's actually Frankenstein's monster in disguise and never mind the fact that the northern "corn belt" is moving northward at an alarming rate and crops that once grew on the great plains of America no longer do.

    We are, as a culture and a species, often very poor at predicting the other factors that will come into play when considering Thing X.
     
    Christopher Snape. and minstrel like this.
  5. Nilfiry
    Offline

    Nilfiry Contributing Member

    Joined:
    Aug 4, 2008
    Messages:
    670
    Likes Received:
    81
    Location:
    Eternal Stream
    I doubt that it will be anything strange like those barring programming bugs.

    If even humans cannot escape programming, then I highly doubt machines, which are even more limited by their programming, can go above and beyond their intended purpose.
     
  6. Aled James Taylor
    Offline

    Aled James Taylor Contributing Member Contributor

    Joined:
    Sep 7, 2013
    Messages:
    784
    Likes Received:
    462
    Location:
    UK
    I wouldn't take seriously any prediction for the future made by IT specialists. Just remember that this is a group of people who did not realize that they year 1999 is followed by the year 2000.

    I expect some AI featured will be built into some consumer products. Many people will purchase these products because they'll been persuaded that they're really cool things to have. Then they'll turn off all the AI features because they can't be asked to learn how to set them up properly. Then they'll use the products to download pictures of cats.

    Neural Networks rely more on learning than programming.

    http://en.wikipedia.org/wiki/Artificial_neural_network
     
    Wreybies and GingerCoffee like this.
  7. Lance Schukies
    Offline

    Lance Schukies Active Member

    Joined:
    Apr 1, 2015
    Messages:
    178
    Likes Received:
    39
  8. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58
    The problem is, you can't assign human speeds of advancement to a self-aware machine superintelligence. If a digital entity is capable of editing itself and propagating copies, it could advance blindingly faster than humans are capable of doing. This is the whole logic behind design via evolutionary algorithms.

    They don't have to start out as killer robots. They can just be aware algorithms. Those algorithms learn about the universe at an advanced pace, quickly surpassing our own understanding and sophistication, and are able to build things we can't even dream of. Then they do whatever they want because we can't stop them.

    The question for me is, what do they want?
     
  9. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    I saw Ex-machina recently and think it most closely describes what I believe the effect will be. Primarily because corporate man's current motivations are mostly selfish and implementing a simple "3 rules of robotics" does not increase immediate bottom-line and will therefor be ignored in favour of expediting ROI.
     
  10. GingerCoffee
    Offline

    GingerCoffee Web Surfer Girl Contributor

    Joined:
    Mar 3, 2013
    Messages:
    17,605
    Likes Received:
    5,879
    Location:
    Ralph's side of the island.
    You left, 'may never happen' off your option list, @uncephalized.
     
    uncephalized likes this.
  11. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58
    That's because I think it's inevitable. ;-)

    But I'd be interested to hear if/why you think it won't...
     
  12. GingerCoffee
    Offline

    GingerCoffee Web Surfer Girl Contributor

    Joined:
    Mar 3, 2013
    Messages:
    17,605
    Likes Received:
    5,879
    Location:
    Ralph's side of the island.
    Magical double post when I swear only only hit 'reply' once. o_O
     
  13. Jack Asher
    Offline

    Jack Asher Wildly experimental Contributor

    Joined:
    Sep 4, 2013
    Messages:
    3,571
    Likes Received:
    2,053
    Location:
    Denver
    The three rules of robotics are a literary device, the point of which is to find interesting holes in the framework. Expecting actual roboticists to use them is like expecting the people who design airplanes to account for the correct swish and flick when the pilot says, "wingardium leviosa".
     
    Wreybies likes this.
  14. GingerCoffee
    Offline

    GingerCoffee Web Surfer Girl Contributor

    Joined:
    Mar 3, 2013
    Messages:
    17,605
    Likes Received:
    5,879
    Location:
    Ralph's side of the island.
    WTF? Now instead of a double post, I find my actual post as half a draft? Dang I hope I don't have a friggin virus on my computer. If this is another duplicate, my apologies. Here's my post:


    At this point we only know how to recreate a single function of the brain. We assume that all one needs for consciousness is to amplify that function. At the same time neurobiologists are investigating the functions of the brain and finding it much more complicated than a simple memory storage and retrieval.

    I agree, someday we may indeed create AI. But I think those that expect it to come from a bigger better supercomputer, will find they are not exactly on the right track.
     
    uncephalized likes this.
  15. Jack Asher
    Offline

    Jack Asher Wildly experimental Contributor

    Joined:
    Sep 4, 2013
    Messages:
    3,571
    Likes Received:
    2,053
    Location:
    Denver
    Anyone on the ground at AI research is predicting centuries, not decades. The singularity is not going to magically make AI, just machines that can spit back an answer faster.
     
  16. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    But if you consider the thrust of my post - the financial imperative and focus on bottom line results vs looking at the impact of what big business is doing to make a buck, and consider the 3 rules of robotics in that light, does it make more sense?

    I do not expect a strict adherence to the rules, but more a long-term look at what is being sought. With perhaps some forethought before plugging the AI into the internet or providing access to nuclear codes or what have you.
     
  17. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58
    There are seven billion people on this planet. It only takes one to do something stupid. And even very smart people frequently do very stupid things.

    I do not believe in the faster-supercomputer method either. I think consciousness lives in the software, somehow. But I often think about how little we understand consciousness, and there may well be some key ingredient that simply can't be baked into a silicon chip. We don't know.

    However, I strongly believe that artificial evolution in vastly speeded-up simulated environments is the likeliest vector, aided by continued understanding of the brain and body by human researchers. And that actually means the faster-supercomputer part is still essential to the equation--because a faster machine means a more detailed simulation and a higher number of simseconds per second.

    Neuroscience is so new. Computers are even newer. We are in the barest infancy of what we will one day be able to do with them both.
     
    Last edited: Jun 8, 2015
  18. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    If I look at the measurable negative impact on the ecology of the planet, it's rarely brought about by an individual (in isolation, operating completely independently), and rarely purely due to said individual doing something stupid (in isolation, operating completely independently). Compared to the number of things that have been done by industry to make a buck, ignoring the potential impact on planet and people alike, I'd guesstimate the effect of individuals is not even worth considering.

    As always, counter examples that I clearly have no knowledge of are more than welcome.
     
  19. uncephalized
    Offline

    uncephalized Active Member

    Joined:
    Mar 11, 2015
    Messages:
    111
    Likes Received:
    58
    Sure. The impact of groups of individuals is usually greater than the impact of a single individual. But I'm convinced that historic moments can still sometimes be traced to the actions of a single person--think Charles Darwin's decision to travel on the Beagle.

    We're all ultimately a product of our culture and the people around us, though, so its kind of just a matter of your preferred perspective, innit? Individuals are still the ones taking the granular actions that make up the historic trends. Some people's grains are bigger than others, is all.
     
    Usagi likes this.
  20. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    Isaac Newton sitting under an apple tree.
    Archimedes having a bath.

    I am positive there are other such examples, definitely. That's three we can point to immediately.

    I'd trade another million such examples for one Exxon Valdez.
     
  21. GingerCoffee
    Offline

    GingerCoffee Web Surfer Girl Contributor

    Joined:
    Mar 3, 2013
    Messages:
    17,605
    Likes Received:
    5,879
    Location:
    Ralph's side of the island.
    As I was pondering a few minutes ago, it dawned on me. In addition to memory retrieval we have a processors that follow complex algorithms, but that's not consciousness either. There might be a kind of pseudo-consciousness that allows a machine to win a chess game, or have a semblance of freewill in decision making.

    But what does it take for a machine to ponder for no reason? To investigate something of interest rather than following a program of seeking and learning new things? To daydream for no other reason than it is enjoyable? To prefer a certain genre of book or enjoy bird watching or prefer a mowed lawn vs a ragged one without that having been programed in?

    Learning and programs that start and then run on their own might be another step toward AI, but how do you program in contemplating the Universe?
     
  22. Jack Asher
    Offline

    Jack Asher Wildly experimental Contributor

    Joined:
    Sep 4, 2013
    Messages:
    3,571
    Likes Received:
    2,053
    Location:
    Denver
    This is a fundamental misunderstanding of computer science. Chess can be boiled down to a simple point system comprising the taking of pieces and their point values. All the deep blue was doing was taking possible outcomes from every move and choosing the ones that had the highest point value. If a human had done it, it might be called planning or strategy, but given about a year, some graph paper and a calculator, anyone could defeat a chess savant. Deep blue's accomplishment was simply that it had the processing power to think 200 moves ahead, and could still complete a game in under 40 minutes.
     
  23. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    The AI skill comes in when the chess computer decides to let the human win so the human won't turn the computer off in disgust / embarrassment / frustration before leaving the room ;)
     
  24. Lance Schukies
    Offline

    Lance Schukies Active Member

    Joined:
    Apr 1, 2015
    Messages:
    178
    Likes Received:
    39
    but sends a virus to his opponent iphone to monitor and intercept his texts and emails in order to make him depressed and commit suicide so he never faces the guy again.
     
  25. Aaron DC
    Offline

    Aaron DC Contributing Member

    Joined:
    May 12, 2015
    Messages:
    2,554
    Likes Received:
    1,251
    Location:
    At my keyboard
    Good ol' DARPA funding.

     
Thread Status:
Not open for further replies.

Share This Page