The idea that human beings might be able to make objects that act in human ways goes back a very long time. Even the medieval golem–a mass of clay animated by the name of god–was not the first. The objects became machines in the nineteenth century, with clockwork as the animating principle. The word "robot" was coined by Karel Capek in 1921. In 1943, researchers began describing the brain as a switchboard-like electrical network. By 1950, the first computers were stimulating their makers to think of "artificial intelligence" and Alan Turing had devised the "imitation game" (better known today as the Turing test) to tell whether a machine was in fact intelligent.
A definition of intelligence was clearly in order, but that has been as hard to come by as artificial intelligence itself. Initially, it was taken to mean whatever people do with their brains. Math? Computers were too good at that from the get-go. Logic? Alas, that fell early to the computers, with Logic Theorist (1956) and the General Problem Solver (1957). Chess? The basic algorithms were worked out by Claude Shannon in 1950, and it wasn’t long at all before a machine could play well enough to qualify as a "patzer" (as well as most humans). Deep Blue trounced world champion Garry Kasparov in 1997, and in 2002, Deep Fritz played champion Vladimir Kramnik to a draw to uphold the title.
There has been a distinct tendency of critics to say that if a machine can do something, that something is not due to real intelligence. But we can do better than that: computers typically do one thing well; humans do many, which suggests that multiplicity of ability is essential to a good definition. For computers, that multiplicity has been elusive. Will we ever achieve it? Arguments rage, with people like Roger Penrose and John Searle arguing in effect that if it ain’t meat, it can’t think. Marvin Minsky, Hans Moravec, and Ray Kurzweil, among others, think it’s only a matter of time. Bill Joy is afraid they’re right, and we should therefore put a stop to the research before our creations do us in.
That is a brutal compression of the history of A.I., but it gives the gist. For more, read Sam Williams’s Arguing A.I.: The Battle for Twenty-First-Century Science, which tracks the origins of A.I. to 1900 (when mathematician David Hilbert laid the foundations of twentieth-century mathematics), follows it through the extraordinarily fertile period of the 50s, and on to today. Williams’s focus is less on the technicalities than on the general ideas, personalities, and debates, and at the end, he encourages us to think of artificially intelligent machines as mirrors of ourselves and of A.I. research as–in part–an exercise in vanity.
Indeed, it is easy to see the critics of A.I. as motivated by their own vanity. Many–like Penrose and Searle–seem offended at the very idea that a machine might be able to duplicate the performance of their refined brains. At the same time, those–like Moravec and Kurzweil–who forecast a time when it will be possible to copy the human mind into a computer always seem to set the day of that possibility at about the time when the forecaster will turn seventy, which suggests a powerful role for wish fulfillment. Or so says Rodney A. Brooks, in Flesh and Machines: How Robots Will Change Us.
But that’s not all that Brooks says. Hans Moravec noted years ago that mental abilities have been embarrassingly easy to match in software, but physical abilities such as walking and navigating have been much harder; he therefore focused his attention on robotics. However, the robots of Moravec (and many others) have been designed to build internal landscape maps and plan ahead for whole sequences of movements, making them slow and clumsy. Brooks has also focused on robotics, but he saw early on that onboard computation to model the world outside the robot was unnecessary and even counter-productive. The world was its own model in a sense, and a robot could be designed to consult that model instead.
How? Reflexes, just the way a bug does it. Don’t worry about where the wall is. Go ahead and bump it. Then back up and change direction. Don’t worry about bumps and holes in the path. Go ahead, take a step, and if the foot stops too soon or goes too far, adjust balance and keep moving. The result has been remarkably insectile robots that scuttle vigorously about the lab and–marketed by Brooks’s company, iRobot, just in time for last Christmas–the Roomba robotic vacuum cleaner.
That’s hardly the end of it. Brooks expects A.I. and robotics to continue to develop. He does not think we need to worry about independent A.I.s taking over or septuagenarians uploading their minds into brainboxes. Rather, he says, we will incorporate the technology to enhance our abilities in many ways and "The distinction between us and robots is going to disappear."
Brooks has a fascinating vision made credible by impressive success in creating autonomous and semi-autonomous devices. Read the book, and keep an eye on him–perhaps especially if you like to play the market.
Me, I’ll keep watching because I’m a teensy bit jealous. Back in the 70s and early 80s, I wrote several papers (including two in Robotics Age in 1984) which discussed using reflexes for robotic control. I never tried to turn the ideas into actual devices. If I had . . . Well, "what might have been" is thin comfort, not worth dwelling on, and I really am happy to know the approach works. So what will Brooks do next?
Jack Cohen is a British biologist, enthusiastic and insistent in manner, who has consulted with a great many SF writers who wished to get their aliens right. Not that he knows what aliens are like, but he is quite good at considering worlds and the environment and evolution, fitting the bits together, and coming up with something that feels like it might really exist somewhere. He has worked with Anne McCaffrey, Harry Harrison, Larry Niven, David Gerrold, and many more.
It is therefore hardly surprising that he and friend, mathematician, and collaborator Ian Stewart–or perhaps the entity they refer to as "Jack&Ian"–should write a book called What Does a Martian Look Like? The Science of Extraterrestrial Life, but the book began long before Jack&Ian got together. Indeed, its roots are in a lecture Jack started giving way back in 1958, "What Does a Martian Look Like?," which was later and more soberly renamed "The Possibility of Life on Other Planets" and given a great many times over the years, always incorporating new material and staying up to date.
Jack&Ian do not care for the modern business of "astrobiology," which they say is far too firmly based in the familiar. That is, astrobiologists look for life as they know it here on Earth. Better, perhaps, is "xenoscience," which admits the possibility of other forms of life, rooted in virtuality ("artificial life") or solar plasmas or . . . who knows? The key, they say, is pattern and complexity and even chaos (the subject of another Jack&Ian book).
But our imaginations are necessarily limited in such matters. We serve the cause better if we bear in mind that the key word is variety, not sameness, and recognize that even here on Earth there is a quite astonishing variety of shape and habit. If one wishes to advise an SF writer, one need only find a terrestrial bizarrity or two, mix a bit of this with a tot of that, and then remember that since every variation is fitted to its environment by natural selection, there must be an environment in which our mix makes genuine sense. Far too many devisers of aliens forget that requirement.
What about intelligence? Or "extelligence," which refers to technology and culture? Again, variety is likely, and we must struggle to escape our parochial biases. And within another century or so–perhaps after we (following Rodney Brooks’s suggestion) "change our selves so that our descendants are very different, very alien compared to what we are now"–we may be able to visit other worlds to see what the neighbors (if any) are like.
Jack&Ian have done an admirable job of pulling together a great many developments from modern biology, chemistry, and astronomy to show the vast breadth of living possibility that awaits discovery. At the same time and as excellently, they remind us that man is not the measure of all things: there are universals, but they are not our familiar forms and chemistries. Those are mere parochials, not to be held too high just because they are all we have at the moment.
This is a book for everyone who writes or reads SF and fantasy. Both will learn from it; the writers will then create better worlds and aliens, which the readers will then appreciate. As for the writers who ignore the book–I expect they will find their readers grown much more demanding, which will serve us all well.
Crossfire,
Nancy Kress, TOR, $24.95, 364 pp. (ISBN: 0765304678). |
The aliens in Nancy Kress’s Crossfire might well have been created using Jack and Ian’s book as a guide. The bad guys are the humanoids; the good guys look sort of like plants, sort of like slime molds, unfamiliar in both form and chemistry.
But first the setup: Jake Holman is a lawyer with a dark secret in his past, but that hasn’t kept him from organizing an interstellar colony. His colonists are not the persecuted minorities who once fled Europe for the Americas, but wealthy faux-Cheyenne, a deposed Arabic royal family, a delegation of Quakers, and assorted millionaire eccentrics. Their ship is the Ariel, their destination the world of Greentrees, certified free of indigenes by a robotic probe. The journey is marred only by a single psychotic crisis and a possible sighting of a strange–alien?–starship, and then they are there, breathing clean air, gazing upon open spaces, and listening to the news of chaos and breakdown back home. They seem to have escaped just in time.
They have brought some problems with them, of course. The Quaker leader, Dr. William Shipley, finds his daughter, Naomi–full of rebellion and antipathy–among the colonists. The Arabs keep their women out of sight. The Cheyenne are planning to leave the colony to return to Nature and live the way their Earthly predecessors once lived. And then they discover the "natives."
They are humanoid and furry, but neither curious nor communicative. Call them "Furs." Naomi Shipley adopts them as her own pet project and soon thinks she’s learning to communicate with them. But then another group turns up, physically identical but aggressive, to get into a killing fight with the Cheyenne. Then another, still Furs, but acting permanently zoned, as if on drugs.
What’s going on? It’s a mystery, until the "Vines" show up. They respond well to Dr. Shipley’s Quakerish approach–he insists the humans sit in silence–and when they produce a translator device, it doesn’t take long to reveal that the Furs have been genetically modified by the Vines. The Furs are their age-old enemy, and they are trying to find a way to make them harmless.
This quite upsets Naomi, who calls it genocide, neglecting that the Vines don’t want to kill. That’s when the unmodified Furs show up and slaughter the Vines, kidnap a party of humans, and insist quite firmly that the humans help them slaughter the entire Vine species. After all, they say, the Vines are evil, abominations. Like Naomi, perhaps, they think it worse to modify than to destroy.
Is it? The Vines are not destroying the Fur species, but they are surely destroying the Fur-ness of the Furs. We might have an Earthly parallel if someone came up with a virus that robbed all people of color of their color. Most of us would surely be horrified. Yet the Vines are as much the targets of the Furs’ destructive urges as the Jews were of the Nazis’. Perhaps a better parallel would be–set this parable about 1940, if you will–a Jewish scientist who came up with a virus that robbed the Nazis of their racism. This too destroys the essence of who the person is, but surely in this case few of us would object.
To do such a thing gratuitously, just to bring the world more in line with our prejudices, would be heinous. To do it, lacking any alternative save annihilation of the foe, to stave off one’s own annihilation, must be counted regrettable but necessary. Form and chemistry, say Jack and Ian, are parochials, only locally relevant. Universals are such things as the requirement for inputs of material and energy, natural selection, and of course self-preservation.
Should the kidnapped humans help the Furs? Or defeat them and help the Vines? I won’t tell you my thought, nor even Nancy Kress’s. Think about it for a bit, remember that this author has been consistently thought-provoking through many novels (most recently Probability Moon, Probability Sun, and Probability Space, reviewed here in October 2000, January 2002, and February 2003, respectively), and then go out and find a copy.
But first, bow your head for a moment. Nancy Kress’s husband, Charles Sheffield, died November 16, 2002. His fiction was long an ornament to the genre. He will be missed by many more than his family and friends.
Engine City,
Ken MacLeod, TOR, $24.95, 304 pp. (ISBN: 076530502X). |
Ken MacLeod’s The Engines of Light trilogy began with Cosmonaut Keep, which introduced us to a cosmos dominated by ancient life forms who reside in asteroids and comets, seething planetary interiors, and Oort-cloud iceballs. Biological creatures that reside on planetary surfaces–like we do–are rare. Yet here are krakens and saurs, who powerfully resemble the greys of saucerite lore and are in fact responsible for all those alien abductions we hear about. And others as well, reaching back to the days of Homo neandertalensis and Paranthropus robustus and more, all of whom now coexist more or less happily on a host of worlds among the stars.
In Keep, a group of modern astronauts (or cosmonauts in this future) discovered one of the gods and was handed the blueprints for a stardrive. They equipped their space station with the gadget, took off to discover the above, and scattered among the hominid worlds. Quite intriguingly, they turned out to be immortal, and at least a few couldn’t resist stirring the political pot. In Dark Light, that meant chaos among the descendants of the Roanoke colonists (some folks always knew, you know!) while the folks from Mingulay, who have learned how to build starships they can control, interview one of the local gods and learn some disturbing things.
Now, in the final volume, Engine City, cosmonaut Grigory Andreievich Volkov and his allies come to Nova Babylonia, the capital world of galactic civilization, bent on prodding the local savants into discovering the secret of immortality and building defenses against the aliens who are on their way. He succeeds, at least in the latter, if not quite in the way he anticipated. Meanwhile, others are meeting the aliens and finding them less alarming than anticipated. Soon the Mingulayans have incorporated them into the Bright Star Cultures and are off to bring enlightenment to a quite paranoid, heavily armed, and renamed New Babylon.
There are characters, of course. They even have names and roles. But they are every one forgettable. What carries this book–and the series–is the big picture: The sweep of history from before the age of dinosaurs into the future, the encounters of species, the struggle to form a pluralistic culture with room for very different folks. MacLeod doesn’t extend this quite as far as the gods in the asteroids, etc., but perhaps one day . . .
He is, after all, optimistic enough to end the book with a rather neat reversal of the last line of Arthur C. Clarke’s "The Nine Billion Names of God."
Darwin’s Children,
Greg Bear, Ballantine Del Rey, $24.95, 385 pp. (ISBN: 0345448359). |
Among the fascinating discoveries that have followed the elucidation of the human genome (and others) has been the presence of numerous viruses–in whole and in part–embedded among the genes that define us. Why are they there? One idea is that they are the remnants of ancient infections, in particular of retroviruses which like herpes and HIV have the ability to insert themselves in our DNA strands and lie doggo until some event (such as stress) awakens them once more, perhaps even generations later. These "endogenous retroviruses" do not, however, seem to be all bad. True, they may give ancient plagues a new chance at life, and they may swap genes with other viruses (much as flu viruses when duck and pig flus meet in Chinese pigs). They may also transport genes between species, as well as provide a sort of "surplus" DNA which can be modified by mutation and selection without risking more established or normal genes. They may thus aid evolution. They may also play essential roles in our biology without being subsumed.
Is it all random happenstance? Or is there pattern, even design? Greg Bear’s Darwin’s Radio suggested at least a pattern: Even as disease erupted, a new kind of child was born, equipped with pigment-flashing skin and pheromonic communications. And there was not just one such child: They were born by the thousands, without obvious causative connection.
Call them "virus children," and watch how they empower fear, panic, and political repression. As the sequel, Darwin’s Children, opens, Mitch and Kaye Rafelson (pivotal figures from the earlier book) are worrying about their daughter, Stella, a virus child. They have kept her safe by keeping a low, reclusive profile in rural Virginia, but now, on the verge of puberty, Stella is rebellious. She wants to be with others of her kind.
So she runs away and is kidnapped by bounty hunters, who plan to turn her over to the government-run "schools," or concentration camps, where the virus children are sequestered and feared and studied. Meanwhile, Christopher Dicken is visiting Fort Detrick to hear "People died . . . Isn’t that enough to make us . . . crazy?" Mark Augustine is visiting the Office of Special Reconnaissance to hear "We’ve found Kaye Lang" or Kaye Rafelson, who once had been the first to discover the problem; they’ve also found Stella, and they’re gloating at the thought of getting her into a camp. Mitch is visiting a politician who has long resisted the paranoid right and hearing that it is getting worse.
Bear spends enough time with Stella–and in time, with other virus children at the camps–to make it clear that they are human, humane, in some ways much more so than the older model. Yet, he says, the paranoid right cannot see anything but fear, and the road to political power that it opens. There are also unscrupulous scientists who relish the thought of dissecting children in search of mysteries and–perhaps even better–new plagues for use in biowar.
To Bear’s credit, he knows that though the pattern is familiar, it is by no means permanent. Political insanity waxes and wanes; sanity returns, for excess is its own foe–people come to their senses, resistance takes shape, science better reveals truth. But it takes time, and events are painful for Mitch, Kaye, Stella, Dicken, Augustine, and many more. The pain is what maintains the reader’s sympathies throughout the tale, until the pattern can be fulfilled.
I asked above whether design could possibly be behind the viruses. It seems unlikely that the viruses would have the same effects–produce the same children–in thousands of independent cases, if there were no such thing. Bear has Kaye Rafelson experience an epiphany–a sense of immanence, of all-accepting love and approval. There is no divine message, but there is a hint of intent or a plan.
I found this hard to accept, but that’s me, not Bear or his story. Given the world in which we live, such speculations have as much validity as those about viruses that are essential to development and evolution. Grant him that, and you will find this one a very satisfying book.
Jubilee,
Jack Dann, TOR, $27.95, 442 pp. (ISBN: 076530676X). |
Jack Dann offers a collection of seventeen prize items from the last twenty-five years of his career in Jubilee. Many are familiar, all are excellent, and you cannot possibly consider the price of this one wasted. Dann is a master storyteller who writes flawlessly and movingly, often of characters in existential or spiritual jams (as in Counting Coup, reviewed here in May 2002).
|
