Are Humans In Jeopardy?

Is the era of human supremacy coming to an end?

What to do now that computers can beat us at our own game. That game, of course, is "trivia," the ability to access random information on a variety of subjects, which would seem to be a prototypically human strength. Yet the contest on TV's "Jeopardy" was one of the most foregone conclusions ever presented for our viewing pleasure.

Following months of hype, Watson - IBM's supercomputer - faced off against the same two contestants it had already trounced in a trial run. After all the publicity, watching the event itself was kind of like seeing the Packers and Steelers on videotape a week after the Superbowl. But this latest victory is still a milepost in the history of artificial intelligence.

Artificial Intelligence and Chess

It's a victory that has been predicted since the earliest computers were in their infancy. Alan Turing, whose pioneering work with logic and theory laid the foundations for computer design (and helped to break the German Enigma code in World War II), could see no reason why his electronic offspring should not eventually meet and even surpass the capabilities of those who designed them.

That was quite a stretch of the imagination back when computers were little more than rapid calculating machines. Others at the time (the 1950's) believed there were certain types of thinking in which humans would always excel. Oddly enough, chess was often cited as an example. But as we now know, the game of chess is trivial for computers, given enough memory and processing power.

By the 1980's futurist Ray Kurzweil predicted that a computer would be able to defeat the best human chess master by 1998. He was wrong, because it happened in 1997 when IBM's Deep Blue overcame Russian grand master Gary Kasparov. By this time most of us were not even surprised, because in most cases we ourselves could not defeat a toy sold by Radio Shack.

If humans still had any claim to unique capabilities, it may have been in our ability to rapidly synthesize a solution based on hazy and tenuous relationships between apparently unrelated things. Though typified by parlor games like "Trivial Pursuits," the same mental ability gives us creativity, the capacity to invent, write music and literature, to produce art and original scientific theories.

But now we may have to cede this advantage to the next generation of supercomputer. Recently Kurzweil picked Watson as the favorite in the Jeopardy competition, saying that even if it lost it would soon improve enough to come back and win. At the rate they're going, he is convinced that within a few decades they will exceed the memory capacity and thinking ability not just of a single human, but of all humans combined.

Passing the Turing Test

What's next for artifical intelligence? Maybe all that's left is the so called "Turing Test." Alan Turing proposed setting up a pair of teletype machines, one leading to a computer, the other to a human operator. If someone conducting a conversation by teletype could not tell the difference between them, then the computer would pass the test - meaning that it had demonstrated human intelligence.

That sounded far-fetched at the time, but those working in the field of AI have come to believe the Turing Test is actually too easy. In fact, demonstrations have already been done which were good enough to fool a certain percentage of human subjects. But the methods used were not nearly advanced enough to be considered true thought.

So other tests will no doubt be devised, with computers - or whatever they come to be called - rising to the occasion, and passing over one bar after another. IBM's room full of computers will shrink to the size of a chip, as it has done before. And perhaps one day computers will be the ones proposing the tests, asking us to demonstrate what it means to be "human."

Draw Me!

I may not know what I like, but I know art when I see it ...

One of my favorite scenes in the film I, Robot is when the robot, "Sonny," tries to tell the investigating officer about his recurring dream. A picture being worth a thousand words (even to a robot), he proceeds to make a drawing to convey the scene. However he draws it like a robot might, in a series of rapid horizontal scan lines like the ones on a TV tube (remember them?) or an old dot-matrix printer. The result is uncanny -- an artistic rendering generated by a machine.

Like most science fiction this robotic trick has been quickly surpassed by reality, as you can see in this news item about Aikon, a machine that not only can draw, but does so by looking at real objects and imitating the same techniques of seeing and modeling that might be used by a human artist. I suspect that Stephen Hawking, who is himself assisted to communicate by a computer speech synthesizer, welcomes being a subject for such an experiment, as you can see in this example.

Back in the early years of computers (the 1950's) many people felt that in spite of their wizardly ability to perform mathematical calculations, there were other types of thinking that a computer would never be able to do. Alan Turing, whose work in mathematics and logic laid the groundwork for the development of the first computers, disagreed. He could find no difference in principle between what was happening in his mechanical "children" and what takes place in a human brain. Turing was one of the first to predict that computers would not only reach but surpass the abilities of those who invented them.

Oddly, in retrospect, the game of chess was sometimes sited as the kind of thing humans would always be able to do better than machines. As we know now, the opposite is true. Only the top grand masters of the game stand a chance against IBM's Deep Blue supercomputer, and most of us can't defeat a toy that sells at Radio Shack. Chess is trivial to computer analysis, and the amazing thing is that we can do as well as we do while lacking the ability to see as many moves ahead as our automated opponents.

So I for one am not the least surprised to see a portrait of Hawking that could be easily mistaken for the work of a human artist. If you play with Photoshop, or one of its alternatives like the open source GIMP, you may be familiar with the filters that can be applied to a photographic image to make it resemble an oil painting -- even to resemble a particular school of art like Impressionism or Pointillism. What Aikon does is far beyond that, actually reproducing at least some of the creative process to compose original work.

What's that? You say Aikon doesn't feel anything? That what it produces doesn't mean anything, either to itself or those who view it? That to be art it has to convey something about human experience?

Maybe you should have another look at the work of Piet Mondrian, compare it to the Hawking portrait, and ask which of them looks like it was done by a human being.

Or maybe you think that feeling and expression are the kinds of things machines will never be able to do?

Bird Brains

New respect for small minds ...

Long ago a particular bird impressed the heck out of me. I was sitting by a pond in a local park (Fuchs Park, to be precise) when this small winged creature hopped down to the water's edge to feed.

My friends who are bird fanciers will have to forgive me for not knowing its species. It was shiny and black or dark brown, that's the best I can do. The main thing is, it was not a water bird, one of those long-legged varieties that wade around looking for critters to scoop up in their long pointed or shovel-shaped beaks. No, this was a normal tree-dwelling type of bird, with a short beak built for pecking. I would have expected it to be searching the lawn for insects and worms.

Instead it amazed me by peering into the water, then quickly darting its head under and emerging with an aquatic snail. Placing it down on a rock, it held the shell with one foot while extracting the escargot with its beak. This bird was a gourmet! I marveled over its ingenuity. How had it learned to do this? When did the first of its kind learn that there was food in the water? Was the behavior taught, invented, or instinctual?

Recently far more advanced examples of avian intelligence have been documented. This BBC news item shows a rook smart enough to drop pebbles into a glass tube in order to raise the water level so it could reach the snack inside. (The story is told in one of Aesop's fables, which we now see may have been based on fact instead of fantasy.) I thought this problem-solving ability was impressive when I saw a chimpanzee do it, but a bird? How is it possible to pack so much intelligence into a brain the size of a pea?

Maybe we need to rethink the whole idea of how unique our intellectual capabilities are. True, Aristotle took the bird's accomplishment a step further when, noticing how his body displaced the bath water, he came up with the concept of mass and how to measure it. And building on such insights, look how far we've come.

But seeing these examples of the innate intelligence of some of the smallest creatures, we need to admit to a greater and more widespread intelligence at large in the world, the same one from which we, like the birds, have been born.

http://news.bbc.co.uk/2/hi/science/nature/8181233.stm

The Singularity

The last thing we will ever have to invent is almost here ...

According to technology pundits like Ray Kurzweil (whose book, The Age of Spiritual Machines, I reported on a couple of years ago), the human race is about to replace itself with something new. We don't know yet exactly how it will come about, or even what form it will take, but if you follow the progress of scientific breakthroughs for another few decades all the arrows point to something unprecedented -- unless you go back to the evolution of human consciousness itself.

Statistician I. J. Good was the first to write about what he called a potential "intelligence explosion" back in 1965:

Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an ‘intelligence explosion,’ and the intelligence of man would be left far behind. Thus the first ultraintelligent machine is the last invention that man need ever make.

We're talking about the ultimate evolution of human tools. We started out with sticks and stones and used them to make ever more refined versions from more and more advanced materials. Now we are assembling components measured in nanometers and creating devices that can very nearly think for themselves. Soon the "very nearly" will be eliminated.

Vernor Vinge, who Kurzweil cites as a source, began to popularize this idea in the 1980's with a series of articles, including one called "The Coming Technological Singularity" (published in 1993 and available for free download from Feedbooks). Vinge is one of those people who have successfully straddled multiple worlds by pursuing careers as a mathematician, computer scientist, and science fiction author. (Another choice example is Rudy Rucker.) This article of his combines the forward looking vision of the futurist with the rigorous thought of the scientist.

Vinge applied the name "singularity" to the intelligence explosion, borrowing the astronomical term for a "black hole," in the sense that further prediction becomes impossible because the rules will change. It begins when super-intelligence -- whether mechanical or biological -- becomes capable of designing improved versions of itself.  (It is already impossible to design improved chips without using computers as tools.)

Once this point has been reached it becomes inevitable that they will eventually exceed not only the intelligence of a human, but the intelligence of all humans put together. Beyond that, our own survival may depend on how well we are able to adapt to this new state of affairs and what role, if any, remains for us to play in the new world. Or, for that matter, if "we" will still be who we are.

Waking Up

There is one leap of faith necessary to bridge the gap between mere computational machinery and human awareness. This is the idea that consciousness results purely as a byproduct of a certain level of complexity and a certain degree of organization. In other words, the human brain apparently consists of nothing more than a collection of identical neurons interconnected with one another, a kind of "soft machine" which begins as a blank slate and gets programmed with memories and personality traits as it is exposed to the external stimulus of the outside world. There is no other mystery component that makes us who and what we are.

[Whatever mystical or spiritual connection there may be to a larger intelligence such as God is a question that is not addressed by this view -- in fact, the idea may have to be seriously reexamined once we are confronted with any example of intelligence beyond ourselves, whether or not it is one of our own making.]

At some point quite early in our lives we apparently "wake up," an event that may be as basic as becoming self aware and which may start with something as simple as discovering that we have toes. Perhaps there is another breakthrough in adolescence when we become even more self-aware (often painfully so) and arrive at maturity.

The leap of faith is the assumption that the intelligent machines we are building will similarly reach a point where they "wake up." This phrase is explicitly used by Vinge as he describes not one but several different ways in which super-intelligence might arise. Interestingly, they are not necessarily intentional:

1. An advanced computer may "wake up," whether or not it was designed to do so, and demonstrate intelligence at or beyond the human level.
2. An entire network of computers (like the kind run by search engines) may "wake up" as a single entity, again with or without intention on our part.
3. Computer/human interfaces may become so intimate that humans themselves can be considered enhanced to a higher level of intelligence. This is a sort of symbiotic result.
4. Bioengineering made possible by the use of computers may result in humans of super-intelligence.

Perhaps the most likely scenario is that in time all of the above may come to pass. Our attraction to technological enhancement will lead us to continue to adopt anything that makes life easier, longer, more healthful, and more enjoyable. It is a road we started on long ago and are not likely to abandon now that it's really getting interesting.

Even those among us who are most critical of technology would be reluctant to give up much of what we have become accustomed to. You might be willing to do without TV, even movies and computers, but what about antibiotics, anesthetics, or modern dentistry? You might be willing to give up having a car, but would you also give up mass transit and go back to exclusively walking or riding a horse? Are you ready to take up farming?

So if they build it, we will likely buy it, and buy into it. And in the long run it will do no good to legislate against progress. If we have any moral qualms about playing Creator in this country, someone in Russia or China or Japan or India will not. Once it becomes possible, it will happen.

"It's Alive!"

Ever since Mary Shelley entertained her house guests with Frankenstein we have been haunted by nightmares of what could happen if our creations get the better of us. The image of the "robot" has become the personification of technology, and remains one of the powerful myths of our age even as it progresses rapidly from fantasy to reality. In fiction robots have been variously treated as benevolent servants and heartless destructive villains. (See my earlier blog about "The Terminator" and "I, Robot.") In real life they will be what we make them -- but only until they begin making themselves. If we want to influence the outcome we had better start making the decisions now.

When will all this happen? Kurzweil says by 2045. Vinge said he would be surprised if it happens later than 2030. So pretty soon we will begin living smarter, longer, healthier lives as "enhanced" humans, whatever that turns out to be. And the co-species we create, whether constructed or bioengineered, may go on to surpass us and to fulfill the dreams we gave them. Like proud parents we may cheer them on as they leave home to explore the stars. Let's hope they drop us a line now and then, just to let us know how they're getting on.

[Other resources: KurzweilAI.net contains another article by Vernor Vinge about what might happen if the singularity does not. Many links for further reading are in the Wikipedia entries for the Singlarity, Vinge, and Kurzweil. And if it all seems too fanciful for you then you need to have your imagination stretched a bit. Try Postsingular, Rudy Rucker's novel about how wacky things might get -- available on Feedbooks .]

Spiritual Machinery

I've just read Ray Kurzweil's 1999 book, The Age of Spiritual Machines. This follow-up to his earlier volume, The Age of Intelligent Machines, attempts to predict what will happen if current technological trends continue for at least another century. Even though I'm predisposed to look favorably on the prospects, I found his conclusions challenging, not to say breathtaking.
And he's worth paying attention to. As an inventor who did seminal work on computer speech recognition, he knows what he's talking about. Many of his earlier predictions for the decade of the 1990's have turned out to be right on target.

He begins by presenting his theory of accelerating returns. The gist of this is that, having taken a long time to evolve, life took a much shorter time to develop intelligence; in less time than that, intelligent humans developed technology; and in less time than that the technology developed into computation. He makes a convincing case that evolution applies equally to social forms and to technology, and that once these things are set in motion they are destined to continue their development at ever increasing speed.

But you don't have to accept that there will never be a slowdown in technological development. If the current exponential rate of progress continues for only another few decades, computers will have vastly exceeded the storage capacity and computational ability of not only the human mind, but of ALL human minds put together. By this time they will be designing and building their own successors, and their role in our interdependent world will be so critical that it will no longer matter if they take over the world from us or not--we will be unable to live in it without them.

Those who doubt this have only to consider what would happen even now if all the computers in the world were suddenly absent or nonfunctional. Transportation, business, communication, government, weaponry, appliances, even the monetary system that supports our civilization in all its forms, all would collapse instantly. This was not true only 25 years ago; it will be vastly more true 25 years in the future.

Yet the picture Kurzweil paints is not bleak. Quite the contrary, the upside to this revolution/evolution could well be unprecedented prosperity, increased longevity and quality of life for all members of the human race, as well as their manufactured progeny.

But perhaps the most interesting part of all this is the question of what constitutes a "human being." When you look at our physical forms, it seems clear that we are not the stuff of which we are made. Our cells, and the atoms that comprise them, are constantly churning, constantly replacing and remaking themselves. The "us" in there is just a pattern, like the shape of our faces, passing through the world like a wave across the ocean.

How can we say that the pattern of consciousness must be limited to atoms arranged in a biological form and not a "manufactured" one? As we watch RNA molecules "manufacture" a new strand of DNA in our bodies, and as we contemplate building new structures at the molecular scale of nanotechnology, how can we define the difference, if there is one?

How we will feel when confronted by "manufactured" beings who claim to be equally conscious, and how we will define our own "humanity" in that situation, is a philosophical exercise we may as well start right now. It won't be long before we will need the answers.