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

Can Artificial Intelligence ever reach Human Intelligence?

  1. Yes

    51 vote(s)
  2. No

    39 vote(s)
  1. Sep 18, 2005 #1
    1st time post here.... thought i'd post up something that causes much debate over... but a good topic. ;-) (please keep it level-minded and not a heated argument)

    Question: Can Artificial Intelligence ever reach Human Intelligence?

    please give your thoughts.... i vote no.
  2. jcsd
  3. Sep 18, 2005 #2


    User Avatar
    Gold Member

    I'm pretty sure my cell phone has more intelligence then some of the people I have met....
  4. Sep 18, 2005 #3
    lol, funny.

    so your cell phone can think on its own? that's a pretty smart cell phone you got there.
  5. Sep 18, 2005 #4


    User Avatar
    Gold Member

    No it can't think on its own. Now think of those implications.
  6. Sep 18, 2005 #5
    whether humans make smart, or dumb, desicions, the level of complexity is far greater than a computer will ever be.

    i think that's what makes the difference mostly.
  7. Sep 19, 2005 #6
    Though not strictly artificial intelligence, right now I am using a program called Dragon NaturallySpeaking version 8 to dictate my comment and have these words translated into text automatically.
  8. Sep 19, 2005 #7
    so what is the point....? lol, i'm not sure i follow....
    AI vs human intelligence is the issue at hand. ;-)
  9. Sep 19, 2005 #8


    User Avatar
    Gold Member

    I believe that program does have to make "decisions" on what your speech patterns mean and all.
  10. Sep 19, 2005 #9
    no, a program is.... "programmed". lol, it only does what it was programmed to do. there is no decision making process. a computer merely calculates numbers, and that's all a computer will ever do, no matter how advanced. ;-)
  11. Sep 19, 2005 #10


    User Avatar
    Gold Member

    Using that definition, AI is undefinable. Your thread is thus, useless.
  12. Sep 19, 2005 #11
    how so? please comment... :biggrin:
  13. Sep 19, 2005 #12


    User Avatar
    Gold Member

    Most people who deal with AI have better definitions then this when it comes to AI. Your basically saying the only platform AI is going to be used with is intrinsically incapable of using AI.
  14. Sep 19, 2005 #13
    How do you know that we do not do the same thing? As Pengwuino stated, your definition is useless.

    What is a decision making process? Think of it as you are studying,
    One way of thinking of AI (and making so-called intelligent robots) is that taking a `pleasure function` as a base and letting the machine decide which action avbailable makes it increase the most. This places `instincts`. For example a robot's bumping into a wall decreases its p. function but recharging its battery increases it and so on. What would you get? Robots addicted to charge, as we may be addicted to sex etc.
  15. Sep 19, 2005 #14
    sure can if you can code 100billion neurons and 10000synapse on average per neuron and given it the sensory/motor skills of a human. It might work a bit slower sorta like a child....but there are movements in california and colorado to build hardware...and i already think colorado has a machine thats like 3-4 years old can't remember what its called though. ohoh did i forgot to mention you gotta raise it for like 10-15 years.
  16. Sep 19, 2005 #15
    If you mean "raw intelligence", computers can already beat the best chess champions in the world, so I'm certain that surpassing human intelligence in complexity (think multi-tasking to the extreme) is inevitable. But programming it with emotions and intuitiveness could prove to be much more complexed. You're trying to teach a computer to ignore logic based on a "feeling". In that sense it would be very difficult to emulate us.

    However the deeper question I think, isn't weather or not we CAN do this, but if we SHOULD. Machines with superior intelligence who are self aware may constitute a threat if they are given sufficient power and control. The counter to this of course, is that we simply keep them in the box and don't give them arms and legs to pummel us with. However society craves simplicity and convenience- the draw of a robot nanny may be too much to resist.
  17. Sep 19, 2005 #16
    if you keep them in a box and take off the limbs how are you they going to grow =]..i mean what if we were to do that to a baby.
  18. Sep 19, 2005 #17
    Humans can be creative; Computers cannot.

    Creativity makes Humans different from Computers.

    It's possible, but what'll the telltale difference between AI and Real I be?

    That, and if computers had free will, what makes us so sure of no revolt?

    Above the above poster's post, shortened into 3 sentences, and agreed with.
  19. Sep 19, 2005 #18
    great statement, this is the kind of talk i'm looking for. however.... can we ever 'create' an artificial conscience, for a computer to actually think on its own, with emotions and feelings? sure, bots and computers can simulate us very much so, but there's still a difference. i think that's a key thought.... conscience. that is a very complex intelligence in itself. :smile:
  20. Sep 19, 2005 #19
    "Creativity makes Humans different from Computers" ?!?!?!? whose to say computers can't be creative? you ever seen someone code a program with 90%
    precompiler directives?
  21. Sep 19, 2005 #20
    could a computer ever have a "gut feeling" regarding a situation?

    could perhaps a computer ever physically feel pain from an emotion?
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Can Artificial Intelligence ever reach Human Intelligence?
  1. Artificial intelligence (Replies: 16)

  2. Artificial intelligence. (Replies: 32)