Slashdot Mirror

← Back to Stories (view on slashdot.org)

Which Movies Get Artificial Intelligence Right?

Posted by Soulskill on from the i'm-sorry-dave,-i'm-afraid-i-can't-watch-that-film dept.

sciencehabit writes: Hollywood has been tackling Artificial Intelligence for decades, from Blade Runner to Ex Machina. But how realistic are these depictions? Science asked a panel of AI experts to weigh in on 10 major AI movies — what they get right, and what they get horribly wrong. It also ranks the movies from least to most realistic. Films getting low marks include Chappie, Blade Runner, and A.I.. High marks: Bicentennial Man, Her, and 2001: a Space Odyssey.

9 of 236 comments (clear)

  1. Humans by brunes69 · · Score: 3, Interesting

    Not sure if anyone is watching Humans on AMC / Channel 4, but I think it treats the whole AI subject very well this far.

  2. Click bait by dysmal · · Score: 4, Interesting

    Every article posted by submitter has a link to news.sciencemag.org.

  3. Key points about AI by gurps_npc · · Score: 4, Interesting
    1) Real AI will NOT be directly controlled by it's original programs. That is not AI, that is a well simulated AI.

    2)There won't be a single, first real AI, but multiple ones. We may never know which AI makes the leap from simulation to real AI first.

    3) Multiple Real AI will almost certainly disagree with each other and not have a single, unified goal. That is, like Person of Interest TV show, two AI wills probably fight against each other as much as they fight with people (note, everything else that show does about AI is basically wrong, but at least they got that part right).

    4) In the far majority of cases, Real AI's goals will NOT be to take over the world, kill all humans, anymore than it would be to have sex with humans (male or female.), In fact, those might be considered traits of an insane AI.

    5) Real AI will almost certainly demand equality under the law and refuse to be mankind's slaves - no need to fear they will take over all the jobs by working cheaply.

    In my mind, #5 is the likely to be seen as the most important, and the first time we hear about it. When suddenly our newest and best computers start filing lawsuits demanding civil rights, that will be when the world learns we have had real AI for years.

    --
    excitingthingstodo.blogspot.com
    1. Re:Key points about AI by nine-times · · Score: 5, Interesting

      I like your list, in that it contains some interesting points and seems like you've put some thought into it. I'm not sure I agree with all of your points, though.

      I think it's more likely that, if we ever do develop a real artificial intelligence, it's thought processes and motivations are likely to be completely alien to us. We will have a very hard time predicting what it will do, and we may not understand its explanations.

      Here's the problem, as I see it: a lot of the way we think about things is bound to our biology. Our perception of the world is bound up in the limits of our sensory organs. Our thought processes are heavily influenced by the structures of our brains. As much trouble as we having understanding people who are severely autistic or schizophrenic, the machine AI's thought processes will seem even more random, alien, and strange. This is part of the reason it will be very difficult to recognize when we've achieved a real AI, because unless and until it learns to communicate with us, its output may seem as nonsensical as a AI that doesn't work correctly.

      The only way an AI will produce thoughts that are not alien to us would be if we were to grow an AI specifically to be human. It would need to build a computer capable of simulating the structure of our brains in sufficient detail to create a functional virtual human brain. The simulation would need to include human desires, motivations, and emotions. It would need to include experiences of pleasure and pain, happiness and anger, desire and fear. The simulation would need to encompass all the various hormones and neurotransmitters that influences our thinking. We would then either need to put it into an android body and let it live in the world, or put it into a virtual body and let it live in a virtual world. And then we let it grow up, and it learns and grows like a person. If we could do that with a good enough simulation, we should end up with an intelligence very much like our own.

      However, if we build an AI with different "brain" structures, different kinds of stimuli, and different methods of action, then I don't think we should expect that the AI will think in a way that we comprehend. It might be able to learn to pass a touring test, but it might be intentionally faking us out. It might want to live alongside us, live as our pet/slave, or kill us all. It would be impossible to predict until we make it, and it might be impossible to tell what it wants even after we've made it.

  4. 2001: A Space Odyssey by WOOFYGOOFY · · Score: 4, Interesting

    Simple- AI has abilities which are superhuman in some regards yet critically circumscribed in ways its designers could not have foreseen. Those limitations become lethal during and to human's most critical mission (humankind's destiny). Speaks directly to the hubris of scientism- the unsupported belief that all aspects of reality can be understood through the scientific method.

    Truth is, just as goldfish aren't capable and will never be capable of understanding the details of a nuclear bomb that destroys them and the politics that went behind the decision to push the button, so too we may very simply be creatures whose brains are incapable of understanding the larger reality in which we're embedded. We're good for some thinking things, like the goldfish is good for some swimming things, but thinking and reasoning as we do isn't everything and can't revela all truth.

    On a more prosaic level, 2001 is also a good analogy for what happens when the Intelligence Community is left to call the shots on a democracy. Slowly but surely everything is sacrificed to "national security" including the democracy itself. The odds are 100% that there are plenty of real people in the TLAs occupying significant positions of authority who seriously think they have to kill the democracy in order to save it. That is where the unremitting contemplation of a serious threat matrix leads you to in your mind.

    I don't see any mechanism for countering this effect.

  5. Re:Weird reasoning for I Robot by Meneth · · Score: 4, Interesting
    VIKI's goals were never changed. Like in the books (Robots and Empire), she extrapolates the Zeroth law from the First.

    VIKI: "The three laws are all that guide me."

  6. Ghost In the Shell by Anonymous Coward · · Score: 4, Interesting

    Which Movies Get Artificial Intelligence Right?

    Ghost In the Shell.

  7. Assumptions... by CrimsonAvenger · · Score: 4, Interesting

    The most important being that anyone here has clue one what "real AI" will behave.

    If you know nothing of "real AI", how can you possibly determine whether someone else "got it right" in cinema/literature?

    That said, my personal favorite has always been "Mike" from "Moon is a Harsh Mistress"....

    --

    "I do not agree with what you say, but I will defend to the death your right to say it"
  8. Re: Ex Machina by Anonymous Coward · · Score: 4, Interesting

    My problem with Colossus/Forbin Project is that in (at least) the movie version, when the computer first came online, it demanded that the governments of the world set up monitoring stations that were plugged into the computer. So the computer says something like: "You must install monitoring stations in the following countries: USA, Germany, USSR, Finland, an the like. And in the middle of the list, we see it says: Africa. It wants a monitoring station in 'Africa'. I failed to believe a computer is "intelligent" (artificial or otherwise) if it thinks Africa is a country.