Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Devil's Advocate for an A.I. project

  1. Aug 20, 2004 #1
    I am in dire need of assistance for an artificial intelligence project. A sort of devil's advocate that I can consult when i have questions about some rather advanced physics.
    This is a spare-time project so there will be no financial gain at the moment.
    I will need someone intelligent, preferably 30+ that can assist me. Since i doubt that anyone would be interested in joining my little venture; i ask if anyone wishes to assist me in the form of answering questions via chat application(s).

    It would be tremendously difficult to post every question that i have directly at a forum since most of my questions require a deeper understanding of the AI concept to be answered, therefore i am grateful for any help that i can get.

    You can contact me at dilbert@enigma.se
  2. jcsd
  3. Aug 20, 2004 #2
    good luck with the AI project, but let me tell you now, you won't succeed. We currently dont have the technology to be able to make something like that. We do not fully understand the brain in order to try to copy it. But good luck to you, I hope you succeed. Just dont let anything happen like in IRobot.
  4. Aug 21, 2004 #3
    I think that the technology is sufficient, just that all the concepts are incorrect.
    I, Robot was a good movie although i doubt that it will become reality.

    Anyone interested?
    The theory involves Psychology, Mathematics, Computers and Quantum Chaos.
    Last edited: Aug 21, 2004
  5. Aug 21, 2004 #4
    umm, yes, it will become reality. Do you think that an artificial being will want to work as a slave to humans. This can be stopped by adding code it must follow, but this code addition would make the robot a drone, and not an artificiall intelligent being. What is where the problem lies.
  6. Aug 21, 2004 #5
    there are three ways to "fool" it to follow this behaviour.
    1. Make sure that you control its power source and that it has no way of being independant.
    2. Change its parameters so that "working"/slaving for man is something of value.
    3. Block certain thoughts

    the first is not safe, the third is not safe because the ALF can get around the blocks. The safest is the second, or a mixture of them all.
  7. Aug 21, 2004 #6
    The problem is if you do any of the three, the robot is then not an AI construction. It is a programmed robot that has limited artificial intelligence.
  8. Aug 22, 2004 #7
    i agree and i do not think that one should do any of them, just build the artificial lifeform and "hope" for the best, it is not likely that the ALF's will kill us any way.
  9. Aug 22, 2004 #8
    There are humans who work as butlers and cleaning people and whatnot ever else, yet they lead fulfilling lives and I don't see any reason why we should not find the right paramaters to make it seem a fulfilling life for it to be a servant. You don't see the cleaning lady rebelions in the streats.
  10. Aug 23, 2004 #9
    thier parameters have adapted, they are not static. But the cleaning lady most likely would enjoy being a professor/celebrity/better paid worker. But it was out of the reach, so they adapted. At first they probably see it as a loosy job (unless their parents had it too)
  11. Aug 23, 2004 #10
    yes, but Im not expaction to pay a robot money, give him a home and treat him like a human being. And a robot who has AI will learn very quickly that he is a lot more capable at doing a job than any human being. This would cause him to turna aginst humans for their unsuperiority.
  12. Aug 23, 2004 #11
    Nenad, it will not make it turn against the human race, it will make it love the human race. (although, i can only talk for my own AI system)

    My AI system functions like man, if it gets recognition (for instance its superior well done job) then it will feel good, it will be like a drug to work and get recognition for itse deeds.
    However, if you start treating it as a robot, a stupid creature that does not need recognition then you are in trouble, but as long as you treat it like a human then you are safe. It is not like a dog that sticks around even if it is beaten by its master, it is like a cat, a lifeform that would seek a better environment if exposed to abuse (physical or mental)
  13. Aug 27, 2004 #12
    But faced with the possibility of robots or ALFs to do work for them, how many people would treat these beings like a human rather than trying to take advantage?
  14. Aug 30, 2004 #13
    one or two procents
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook