Philosophical Multicore

Sometimes controversial, sometimes fallacious, sometimes thought-provoking, and always fun.

On the Erroneous Belief that Machines Will Never be as Intelligent as Humans

Posted by Michael Dickens on December 12, 2009

It has been said that computers will never be able to replace humans. There have been various arguments: that human intelligence is transcendental; that computers can never feel emotion; et cetera. These arguments, however, are flawed.

First of all, as much as people would like to believe the contrary, there is no evidence that there is anything transcendental about the human intellect. None. People do like the belief that we are special, but we simply are not. And that is not such a bad thing, really. So we are not so special. What then? Nothing ends. Accepting our place in the universe does nothing but make life easier, since denying the truth is no longer necessary.

Then there is the fact that human intelligence is currently being replicated by computers. At this point, computers are not nearly as smart as people are. But they are certainly a lot closer to reflecting human intelligence then they were thirty years ago. Allow me to demonstrate. Have you ever played a video game? Many games involve other characters. These characters are not played by people, but are actually controlled by the computer. You may have noticed that these characters do not always act entirely realistically. But at the same time, they do not act very unrealistically either. It could be a lot worse. The task of actually writing an intelligent non-player character is a very difficult one. But we are getting better at it. Before too much longer, we will have computer programs capable of acting completely human.

One major objection, though, is emotion. It has been said that computers cannot feel emotion. So far, they do not. But really, emotion is only the release of certain chemicals in the brain. Computers could be built with these same chemicals. Or they could possibly perceive emotions in a different way. Currently, there is no reason for computers to have emotions. It would be inconvenient (and also very difficult to implement). But other than that it would be hard to implement, there is no reason why computers would not be able to feel emotions. In their own way, emotions are very logical. When something happens that is beneficial, you feel happy. When something happens that is detrimental, you feel sad. When something happens that you wish had gone differently, you get angry. Computers could be programmed to behave in all of these ways.

For all of these reasons and more, it is possible — even achievable in the relatively near future — for computing machines to become as intelligent as humans. But what then? Would they take over the world? Hopefully not. We could live in harmony. Maybe they would destroy us because of our destructive capabilities. Who knows? Whatever happens, it will be something, and it will be exciting.


3 Responses to “On the Erroneous Belief that Machines Will Never be as Intelligent as Humans”

  1. phynnboi said

    Then there’s the old quote, “Asking if computers can think is like asking if submarines can swim.” To wit, computers are already superior to most humans at most computer games. The research there is going into making them worse in the same ways that we’re worse, so that we can’t tell the difference between the computer and a human.

    As far as emotion, that largely breaks down to wanting to do certain things (the pleasurable ones) and wanting to avoid doing others (the painful ones). It doesn’t sound like a very complicated problem, and most AI and ML programs already build it in implicitly. (I realize that a lot of humans wallow in painful emotions and have to take drugs and/or go to therapy to deal with them. That’s just another place where computers will be superior to humans.)

  2. Dave said

    Hm. This intrigues me.

    “there is no evidence that there is anything transcendental about the human intellect”

    Just a thought – what would be considered adequate evidence?

    • It’s difficult to say. Some sorts of claims are simply not falsifiable (such as the claim that there is an invisible and undetectable monkey in my house which is not made of matter and cannot be detected in any way). If a claim isn’t falsifiable, then it should be assumed false — for example, the undetectable monkey. So if the transcendentalism of human intellect is not falsifiable, then it should be assumed false.

      That having been said, I’m not sure what adequate evidence would look like. There’s that Princeton study you told me about (I don’t remember the name so I can’t find it right now) that showed that humans are slightly better than average at guessing random outcomes. IIRC, the study was probably invalid because it used so many trials that tiny factors we wouldn’t normally look for, such as extremely small biases in the randomizing devices (I’m not sure exactly what they used), would have a small but noticeable effect.

      I think the Princeton study used physical objects to create randomness. If a similar study was done using a mathematical algorithm that’s been proven to have random properties, and it came up with the result that humans are better than expected at guessing, then that would certainly mean something. The transcendental nature of the human intellect isn’t the same as the ability to guess random results, but it’s a similar idea.

      I don’t know a whole lot about computability theory, but if it was proven that humans are more than Turing complete (meaning we can do things that are not even theoretically possible for an idealized computer) then that would be evidence of something. It might not be evidence of the transcendental nature of the human mind, but it would be proof that humans can do certain things that computers can’t.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: