To Think or Not to Think. What is the question?

I’m currently reading a pair of books about artificial intelligence.  The Emperor’s New Mind was written by Roger Penrose in 1989, and is an exploration of how computers work, and why they will never be able to mimic the intelligence of the human mind.  How to Create a Mind is a book by Ray Kurzweil written in 2012 about how the human mind works, and how computers will be able to mimic those structures and reach a level of true artificial intelligence. I’m deliberately reading them semi-simultaneously to maximize the juxtaposition of time and technological advancement, which is stark.  But I don’t want this blog post to be a comparison of the two books or approaches.  Rather, I want to talk about the question that has been plaguing me since I started reading them.

What is intelligence?

More specifically, how do we identify intelligence in others?  If we’re thinking about what is artificial intelligence, what are our standards?  How do we know when we’ve made it there?

I am intelligent.  No one, and nothing can convince me of otherwise.  My logic for this pronouncement is trite and cliche.  I think, therefore I am.  I am intelligent because I can recognize my own intelligence.

So I know that I am intelligent.  But how do I know that you are intelligent?

Well, since you are reading these words, that provides a measure of intelligence.  But a computer can read this words as well, technically.  If you leave a comment, you demonstrate your ability to interact, to communicate based on these words.  To create something new, based on what I’m talking about.  You’ve probably noticed in your blog though, lots of spam comments that are computers doing exactly that.  But those computers won’t do it as well as you can.  I can recognize a spam comment versus one from one of you.  So that commenting ability shows your intelligence.  If we were in person, we could have a conversation about what I wrote about.  That conversation would demonstrate to me that you are intelligent.

In fact, conversation has long been one of the hallmarks of artificial intelligence, in something called a Turning Test.  The test is this.  In a blind experiment, can a human judge have a conversation with a computer and a real person, and identify which one is the computer reliably?  Essentially, can a computer mimic human conversation, and pass as a human?  And indeed, a computer named Eugene passed a limited Turing Test a few years ago(well, kinda sorta, ish, not really, but hey good effort).  Does that make Eugene intelligent?  Is mimicry enough of a requirement for intelligence?

I think not (as would most of the world).  Part of the problem that we face with artificial intelligence is that the most convenient way to test intelligence is by performing tasks.  How smart are rats?  Put them in a maze and find out.  How smart are monkeys?  See if we can teach them sign language.  But intelligence is not defined by task completion.  I am not intelligent because I can hold a conversation.  I can hold a conversation because I am intelligent.  I am not intelligent because I can write these words, and drive a car, and play chess, and write music.  I can do those things because I am intelligent.  A computer can do all of those tasks as well.  But can a computer do them because it is following a preset series of commands, or an algorithm created by a human to complete those tasks.  So in a world where more and more of the tasks that we might use to show our intelligence can be completed as well or better by a computer or machine, how do we identify intelligence?  How do we know when we truly have an artificial intelligence, and not just some combination of algorithms (which is of course assuming that we are more than simply a combination of algorithms, which is hardly a sure thing).

Perhaps the task that best shows our intelligence is the task of questioning.  My intelligence is not demonstrated by my ability to converse adequately with you.  My intelligence is demonstrated by my ability to question the world around me.  And out of those questions, manipulate the world around me.  We are intelligent because we can ask the question “What is intelligence” and then design experiments to try to figure it out.  In other words, we are able to not simply follow a series of commands to solve a problem.  We are able to identify the problem, and then create the series of commands to use to solve the problem.  Perhaps we’ll know when we have created artificial intelligence because the computer will start trying to determine whether are not we are intelligent.  Maybe God knew that He got mankind right when we started to question his existence?

Question for another day: Why are we thinking about intelligence in terms of human intelligence, based on human tasks?  Why wouldn’t artificial intelligence look completely different than human, with its own set of values and goals?

Only tangentially related, but just because I wanted to add a video:

6 thoughts on “To Think or Not to Think. What is the question?”

  1. One must be intelligent to develop a post like this. And I feel a little bit more intelligent having read it.

    By many accounts, computers have already achieved intelligence, but the fact that the intelligence they possess is artificial (i.e. man-made) takes something away from it. Maybe artificial intelligence is a misnomer. Maybe the true test for computers is to develop intelligence (again, whatever that means) independent of human involvement.

  2. Great questions, Jean. Is it possible for computers to program “offspring?” That would be moving in the direction Andrew mentions, where they gain agency upon intelligence. I think part of intelligence is being capable of learning, storing, recalling, and combining information to achieve new insights. I’ve read that the speed of artificial intelligence doubles at the rate of Moore’s law, while ours does not. Humans in the information age are capable of learning a lot more than in the past, but can they effectively store, retrieve, and manipulate it as quickly as computers?

  3. Interesting question. I think the concept of offspring is tied to the idea that Andrew mentioned of being manmade. If humans program a computer to produce offpring, then that wouldn’t be an intelligent operation, in my mind. But if a program were to decide on its own to develop offspring, and determine a method for doing so, would that function be consider intelligent? I suppose part of the question here is where’s the line between intelligence and consciousness, and which is the goal? What are the benefits of each? Because we have many examples of AI in the world (video games are a perfect example), but how independent do we want AI to ever become? Is the Commander Data model of evolving beyond original programming desirable? Programs that can problem solve for themselves, including the step of identifying the problem? Or do we want to remain the gatekeepers of identifying the problem?

    1. Exactly– “I suppose part of the question here is where’s the line between intelligence and consciousness. ”

      Jean, first I have to say that your blog is visually outstanding and always intriguing.

      I’ve been reading Michio Kaku’s book, The Future of the Mind, in which he offers this definition of consciousness:

      the process of creating a model of the world using multiple feedback loops in various parameters (e.g., temperature, space, time, and relation to others), in order to accomplish a goal (e.g., find mates, food, shelter).

      Kaku further defines consciousness on four levels: 0 for invertebrates and plants; 1 for the reptilian brain; 2 for animals who use feedback loops to interact with others using the functions of the limbic system (hippocampus for memories, amydala for emotions, thalamus for sensory information; and 3 for humans, who are advanced from other living things by the ability to simulate the future. Computers also can simulate the future, and with the proper feedback loops in the program, would enter this analysis on par with humans, were it not for their superior processing capabilities.

  4. I haven’t put too much thought into what intelligence was and I always assumed that computers were smart, but only because humans made them smart. I’d still like to believe that because I agree with your point, “We are able to identify the problem, and then create the series of commands to use to solve the problem.” Can computers do this? I mean, I guess they troubleshoot, but I always feel like I have to do the troubleshooting for my computer and then take it to a human to fix it. If it’s so smart, why can’t it fix itself?

    On another note, the note you made about God knowing He got mankind right when we began to question His existence. So what happens when computers develop the intelligence to question us? I’ve seen those movies and it doesn’t always end well for either side.

Leave a Reply

Your email address will not be published. Required fields are marked *