Many of you know that I listen to lots of podcasts. One that I recently picked up is called "Stuff You Should Know", which recently broadcast this episode about something called "The Singularity". First of all, what is "The Singularity"? Admitting my limited knowledge in this area, The Singularity seems to be the moment when, intentionally or accidentally, computers/networks become self-aware and possess greater intelligence than human beings. Some scientists working in the field of artificial intelligence with computers predict that this will occur around the year 2030, others suspect it may take a few decades longer, but most technologists seems to agree that it will happen.
I am intrigued by this idea, and would like to pose a few questions and make some general comments. I can be a bit of a luddite, so perhaps my comments should be taken with a grain of salt.
I'm no technology expert, but the idea of a computer or a network of computers becoming "self-aware" doesn't seem very logical. I'm not sure how a machine can suddenly obtain self-awareness. Many people talk about this as though it is a threshold, and computers just haven't reached it yet, as if it is a fore-gone conclusion that they will reach it. I don't agree with the concept of self-awareness as a threshold; it seems to me that it is an innate quality of the soul. I've never heard anyone sufficiently explain how a computer/network would become self-aware.
Furthermore, when one listens to people talk about "The Singularity", there is an implication that with self-awareness comes will or volition. This is not necessarily true. The concept of "will" is complex even to describe, and certainly no one can explain the origins of volition. Where do our desires, as human beings, come from? How do we decide to act upon our desires? How do we convert ideas into actions? No one has adequate answers to these questions, and the idea that one could "program" them into a computer is ridiculous, as is the idea that these elements could suddenly, spontaneously arise from a machine.
As evidence of this imminent Singularity, many scientists point out that in 15-20 years, we will be able to create a computer that has more computational power than the human brain. First of all, the human brain is about more than computational power. It is the seat of judgement, inference, hypothesis, etc. No transistor, regardless of its computational power, has the ability to generate these complex higher functions of the brain (what psychologists call the "executive" functions). Ever since humanity became enamored with machines in the 18th and 19th centuries, people have been trying to compare the human body and the human brain to a machine. This analogy is simply false. Humans are not some kind of organic machine. Our brains are not some sort of chemical computer. We are infinitely more complex than the wildest unattained dreams of any scientists/technologists. So much of how we operate as a person, where our identity comes from, how our personality is formed and extended, etc. remains a mystery.
I think the root of this conundrum is how one thinks about "intelligence". I am always hesitant to speak of "intelligence" with my students, because it is so easily mis-understood. Many people believe that intelligence is a fixed quality, that it can be demonstrated and measured, and that if we can simply get a computer to display "intelligence" in the same way we measure human intelligence, then we'll have reached this Singularity. I, however, respectfully disagree. I think that our attempts to measure human intelligence are inadequate at best. No test can truly measure a human being's intelligence, because intelligence is too multi-faceted to fit neatly inside of a test. There are too many dimensions to intelligence to fully describe any human being's intelligence. I humbly submit the idea that the totality of the human experience cannot be replicated or even approached by a machine. Our lives are too full of beauty (completely unquantifiable), mystery (indescribable), and a richness of experience (both internal [psycho-social-emotional] and external [physical]) to be computed or calculated or replicated by a machine.
I suspect that 20-30 years from now, there will still be individuals working on artificial intelligence and the Singularity. And I think they will have made some remarkable discoveries and contributions to humanity. But I do not believe that we will have computers that can approximate human beings or that somehow represent "the next stage in evolution or intelligence".
I am intrigued by this idea, and would like to pose a few questions and make some general comments. I can be a bit of a luddite, so perhaps my comments should be taken with a grain of salt.
I'm no technology expert, but the idea of a computer or a network of computers becoming "self-aware" doesn't seem very logical. I'm not sure how a machine can suddenly obtain self-awareness. Many people talk about this as though it is a threshold, and computers just haven't reached it yet, as if it is a fore-gone conclusion that they will reach it. I don't agree with the concept of self-awareness as a threshold; it seems to me that it is an innate quality of the soul. I've never heard anyone sufficiently explain how a computer/network would become self-aware.
Furthermore, when one listens to people talk about "The Singularity", there is an implication that with self-awareness comes will or volition. This is not necessarily true. The concept of "will" is complex even to describe, and certainly no one can explain the origins of volition. Where do our desires, as human beings, come from? How do we decide to act upon our desires? How do we convert ideas into actions? No one has adequate answers to these questions, and the idea that one could "program" them into a computer is ridiculous, as is the idea that these elements could suddenly, spontaneously arise from a machine.
As evidence of this imminent Singularity, many scientists point out that in 15-20 years, we will be able to create a computer that has more computational power than the human brain. First of all, the human brain is about more than computational power. It is the seat of judgement, inference, hypothesis, etc. No transistor, regardless of its computational power, has the ability to generate these complex higher functions of the brain (what psychologists call the "executive" functions). Ever since humanity became enamored with machines in the 18th and 19th centuries, people have been trying to compare the human body and the human brain to a machine. This analogy is simply false. Humans are not some kind of organic machine. Our brains are not some sort of chemical computer. We are infinitely more complex than the wildest unattained dreams of any scientists/technologists. So much of how we operate as a person, where our identity comes from, how our personality is formed and extended, etc. remains a mystery.
I think the root of this conundrum is how one thinks about "intelligence". I am always hesitant to speak of "intelligence" with my students, because it is so easily mis-understood. Many people believe that intelligence is a fixed quality, that it can be demonstrated and measured, and that if we can simply get a computer to display "intelligence" in the same way we measure human intelligence, then we'll have reached this Singularity. I, however, respectfully disagree. I think that our attempts to measure human intelligence are inadequate at best. No test can truly measure a human being's intelligence, because intelligence is too multi-faceted to fit neatly inside of a test. There are too many dimensions to intelligence to fully describe any human being's intelligence. I humbly submit the idea that the totality of the human experience cannot be replicated or even approached by a machine. Our lives are too full of beauty (completely unquantifiable), mystery (indescribable), and a richness of experience (both internal [psycho-social-emotional] and external [physical]) to be computed or calculated or replicated by a machine.
I suspect that 20-30 years from now, there will still be individuals working on artificial intelligence and the Singularity. And I think they will have made some remarkable discoveries and contributions to humanity. But I do not believe that we will have computers that can approximate human beings or that somehow represent "the next stage in evolution or intelligence".
Fascinating, but I thought you were a little remiss in not referencing the philosophical implications of a soul. Perhaps a bit of Aristotle or Aquinas would have lent some more power to your argument.
ReplyDelete