- Shopping Bag ( 0 items )
In this compelling book, Hans Moravec predicts that machines will attain human levels of intelligence by the year 2040, and that by 2050, they will surpass us. But even though Moravec predicts the end of the domination by human beings, his is not a bleak vision. Far from railing against a future in which machines rule the world, Moravec embraces it, taking the startling view that intelligent robots will actually be our evolutionary heirs. "Intelligent machines, which will grow from us, learn our skills, and share our goals and values, can be viewed as children of our minds." And since they are our children, we will want them to outdistance us. In fact, in a bid for immortality, many of our descendants will choose to transform into "ex humans," as they upload themselves into advanced computers.
This provocative new book, the highly anticipated follow-up to his bestselling volume Mind Children, charts the trajectory of robotics in breathtaking detail. A must read for artificial intelligence, technology, and computer enthusiasts, Moravec's freewheeling but informed speculations present a future far different than we ever dared imagine.
Progressive change sculpted our universe and our societies, but only very recently has human culture seen beyond the short cycles of day and night, summer and winter, birth and death, to recognize it. No sooner was universal change noted in the traces of history than its accelerating pace became discernible in a single lifetime. By almost any measure--energy, information, speed, distance, temperature, variety--the developed world is growing more capable and complex faster than ever before--a statement that has been true for at least half a millennium, and mostly true since the agricultural revolution and the invention of writing over five thousand years ago.
Many of the products of this accelerating process--written language, city-states, and automation, for instance--sped it further. Today the pace strains the limits of human adaptability: the lessons of a technical education are often obsolete before the education is complete. Nevertheless, the acceleration continues, as machines take over where humans falter. In the 1970s photographic patterns for manufacturing integrated circuits with dozens or hundreds of components were designed and drafted by hand, on plastic sheets. Today's computer chips contain tens of millions of components, placed by design programs running on older computers. Not only did one generation of machines make possible the next, they enabled it to appear in less than a year, compared to an average of three years for purely human designs.
Self-accelerated computer evolution affects other technical fields, and almost every design engineer's work is being amplified and accelerated by computer workstations and communications. The many parts of the new Boeing 777 aircraft, to pick a major example, were designed in parallel, by distantly separated engineering teams, with powerful three-dimensional modeling programs. Subassembly designs were checked for compatibility by programs that put together the aircraft in simulation, detecting major and minor mechanical, electrical, control, and aerodynamic problems while they were easy to correct, long before a physical prototype was built. The result was an aircraft of unprecedented complexity brought into existence in half the time of previous models. In the same way, chemists and biologists are replacing years of wet lab work with weeks of molecular simulation. Architects quadrupled their business capacity by replacing drafting boards and manual bookkeeping with computer workstations in the early 1990s.
"Things tossed up come down" is an early theory of gravity, demonstrably true in everyday life, unquestioned for millennia, until Newton developed a new theory of gravity that gave stable orbits to sufficiently fast satellites, and let slightly faster projectiles escape to infinity. "Wood rubbed warm cools down" may have been a truism for our distant ancestors, until one of them rubbed hard enough to achieve ignition temperature, whereupon the wood flamed hotter than ever on its own. "Machines break down" is a demonstrable truth of industrial society, but as machines increasingly design, diagnose, and repair themselves, it too will be suddenly invalidated. Once given "escape velocity," machines more capable than any we know will, without further help from us, grow more capable still, learning from the world, as we did in our biological and cultural evolution. The wood is already smoldering.
Like passengers in a rising elevator, those riding a developmental curve may be unaware of the altitude already reached--until a passing window shows a glimpse of the ground. In 1930 an Australian gold-prospecting party flew into a supposedly uninhabited area deep in the New Guinea highlands and encountered a human culture separated fifty thousand years from their own. The naked inhabitants, some with stone spears, were driven into paroxysms of confusion and religious fear and awe by the giant roaring silver birds that alighted near their mud-thatch villages to release droopy-skinned white men without genitals who, among too many wonders, captured their souls in small black boxes labeled Kodak.
In 1991 Davi Kopenawa took the giant step of being the first to leave the jungle to speak for his people, the Amazonian Yanomami. The Yanomami, with a population of about twenty thousand the largest remaining stone-age tribe, were isolated from the rest of the world for ten thousand years until this century, when missionaries, anthropologists, and, more recently, highway workers and gold miners, began to invade their homeland. Accompanied by a translator, wearing his only possessions, sneakers, jeans, and a sweater given to him for the trip, he visited New York, Washington, and Pittsburgh, to beg to be left alone: foreign diseases, especially malaria, had killed one-fifth of the Brazilian Yanomami in five years.
What he saw in the cities horrified him: crazy ant-people crawling in sky-high huts thinking about cars, money, and possessions instead of relatives and nature. In a zoo he identified with the listless animals among plastic plants, steel vines, and bad air. "If I had to live in your cities for a month, I'd die. There's no forest here."
Kopenawa has a point. The world we inhabit is radically different, culturally and physically, from the one to which we adapted biologically. We were shaped during the last two million years by an ongoing Ice Age--a time of continuous climatic change, as every few tens of thousands of years glaciers advanced and retreated over most of the earth (the current warm spell is but an interglacial period). Such variability favors high adaptability, by making life untenable for the rigidly optimized. In our species the adaptability took the form of a hypertrophied brain and an extended childhood, supporting an extreme cultural plasticity, along with an ever more expressive language to rapidly pass on adopted behaviors: as we grow to puberty we can learn equally well to be fur-clad arctic hunters, robed desert nomads, or naked equatorial gatherers. For almost all of human history, as still in Kopenawa's world, cultural inheritance played a straightforward supporting role: providing the how for the basic needs of life. But somewhere, about five thousand years ago in our cultural history, the relationship between biology and culture began to alter radically.
The Cultural Revolution
Culture lets us rapidly accommodate to environmental changes because it is a medium for a new kind of evolution. Collections of rules for behavior (memes, to use a term invented by Richard Dawkins) pass from generation to generation, mutating and competing with alternatives, just as biological genes do--only much more quickly. A biological trait requires generations of selective replication to become established in a population, but a cultural practice can be altered, and spread through an entire tribe, many times in a single human lifetime. After hundreds of thousands of years of slow cultural meander, our ancestors stumbled into a set of behaviors that catalyzed the creation of ever more behaviors and memories, and physical implements to support them: a self-accelerating cycle that is reaching escape velocity today. What exactly sparked the tinder, apart from simple accumulation of useful skills, is a fascinating question. A baby boom or forced migration in an improving climate may have led to shortages in hunting and gathering resources, forcing would-be survivors into agricultural life, and eventually the first agricultural civilizations, in the Near East and China ten millennia ago.
For millions of years, primates, our ancestors included, have lived in tribes. Among primates (as in canine packs, but unlike herd animals) individuals know one other personally, and maintain long-term one-on-one relationships, involving dominance, submission, friendship, enmity, debts, grudges, and intrigues--the stuff of soap opera. Complex socialization gives the tribe great abilities. In critical circumstances, individuals know who to trust with what tasks. But remembering many things about many individuals ought to take storage space in the brain. Robin Dunbar has indeed found a tightly correlated linear relation between brain size and troop size in monkeys and apes--macaque monkeys, for instance, form bands of about fifteen, while larger-brained chimpanzees and gorillas live in tribes of thirty to forty members. This soap-opera connection likely drove the evolution towards large brains in primates, since tribes compete for food and shelter-providing territory, and a coordinated larger group is likely to beat out a smaller one, giving large tribes, and thus large brains, an advantage. Dunbar extrapolates the primate group/brain ratio curve to human brain size, and finds our natural tribes should have about two hundred individuals. In fact, this is just the maximum size of self-contained nonhierarchical human groups: Yanomami Indian villages and gypsy bands, for instance, and perhaps hippie communes. Modern society's overlapping webs of individual acquaintances muddle but don't eliminate our tribal limitations, evidenced in ubiquitous anecdotes. My wife, involved in many church organizations, notes that growing churches have major crises of identity when their membership reaches about two hundred. The computer science department at Carnegie Mellon University was known for its cooperative, "family" atmosphere in the 1970s, when it numbered about a hundred. It grew rapidly in the 1980s, and in the 1990s the over six hundred members of the School of Computer Science are divided into several departments and projects that are strangers to one another.
The agricultural civilizations were able to grow far beyond village size because of a series of social inventions, among them institutional roles like King, Soldier, Priest, Merchant, Tax Collector, and Peasant, clearly marked by costumery, ceremony, and standard rituals, substituting for the impossible task of remembering thousands of individual relationships. New solutions bring new problems. Cheaters in villages are easily recognized and punished but find many opportunities and hiding places in the anonymity of large society. Enforcement institutions--Moral authorities, Lawmakers, Police, and Criminal labels--partially countered the breakdown of cooperation. The problem of keeping track of who owes what to whom, a matter of memory in a village but a criminal opportunity in a city, encouraged the invention of recordkeeping: tokens, tally marks, a number system, and eventually writing. The new social functions involved complicated procedures unlike those of tribal life, many thus slow and difficult to learn. Enter extended formal training periods, eventually Teachers and Schools.
Like villages, civilizations compete with one another for resources and may gain advantage from institutions that foster innovations--and incidentally put cultural evolution into higher gear. Agriculture benefits from precise knowledge of the season, and thus of celestial cycles, and military and civil projects go better with professional thinkers and builders on the job, so the positions of Astronomer/Astrologer, Philosopher/Magician, and Engineer/Artisan become part of the picture. The innovations of professional thinkers, transmitted by increasingly effective written language over huge distances and times, accelerate innovation itself. The result is a process far, far faster than biological evolution that produces ever more elaborate places for humans to live, ever swifter ways for them to move and to communicate, ever larger storehouses of previous thought, ever more territory occupied, ever more energy controlled. It also produces a world increasingly unlike the villages, fixed and nomadic, in which human behavior evolved, and so makes ever greater demands on our adaptability.
Strange Ducks, Out of Water
Today, as our machines approach human competence across the board, our stone-age biology and our information-age lives grow ever more mismatched. Work in the developed countries has become increasingly specialized and esoteric, and it now often takes a graduate degree, representing half a working lifetime of sustained learning, to master the necessary unnatural skills. As societal roles become yet more complex, specialized, and far removed from our inborn predispositions, they require increasing years of rehearsal to master, while providing fewer visceral rewards. The essential functions of a technical society elude the understanding of an increasing fraction of the population. Even the most successful individuals often find their work boring, difficult, unnatural, and unsatisfying, more like a sustained circus performance than a real life. Caffeine substitutes for natural adrenaline. Those original activities that do remain--eating and child raising, for instance--are often squeezed by the strange new tasks. The mismatch between instinct and necessity induces alienation in the midst of unprecedented physical plenty.
By the standards of our inherited tribal psychology, we are sick and crazy. Physically, however, we are healthier and live longer than ever, and we have vastly more options in every sphere of activity. Few city-dwellers would be prepared to adopt the circumscribed life in a stone-age forest village, despite uneasiness with their own. On the contrary, much of the third world is rushing to overcome its physical problems by adopting the patterns of the developed nations (Davi Kopenawa himself is now a regular speaker at ecological meetings around the world, and his stories and commentary are extensively represented on the World Wide Web!). The urbanized, meanwhile, have devised substitutes for some tribal experiences, for instance, churches and other social organizations that bring together village-sized groups with a common sense of purpose, a shared experience, a defining mythology, and uniform behavioral expectations. Others find release in competitive sports (very like tribal wars), outdoor vacations, or even backyard barbecues. Some business trips resemble mammoth-hunting forays but lack the scenery, quiet stalks, and satisfying physical marksmanship--and a golfing weekend fills the void. But, as the pace, diversity, and global geographic interconnectedness of life continues to increase, even such occasional imitations of our ancestors' lifestyles are crowded out and may be becoming less satisfying. The world is rushing away from our ancestral roots ever faster, stretching the limits of both our biological and institutional adaptability.
Some individuals and communities have tried to isolate themselves from the problem. The Pennsylvania Amish live in a perpetual state of early-nineteenth-century rural industrialization. Some cloistered religious orders operate like isolated tribes. Countercultural rural communes of the 1960s and 1970s deliberately resembled villages. Yet, industrialized society's increasing population, accessibility, and competitive vigor in all fields seems to erode such communities, who cannot reasonably, or legally, deny members in need the benefits of modern medicine, inexpensive food, clothing, building materials, useful machinery, and especially empowering, but distracting, education.
There are unhappy voices today calling for a worldwide rollback to an earlier state of affairs. They are outvoted by the demands of billions for food, housing, and civilized comforts. Yet, paradoxically, as our cultural artifacts achieve self-sustaining maturity, they will provide the means to restore humanity and nature to an imitation of the wild past.
Back to the Future
Productivity rose during the Industrial Revolution, as steadily improving machines outperformed and displaced ever more human labor. Simple diffusion, and social innovations like labor unions and profit taxes, widely distributed the consequent wealth. The wealth expressed itself both in increased public and private consumption and increased leisure time. During the last three centuries in the industrialized countries, slave and child labor and hundred-hour factory work weeks have given way to under-forty-hour weeks and mandatory retirement.
Short-term fluctuations in the trend notwithstanding, as machines assume more--eventually all--of essential production, humans everywhere will be left with the options of the idle rich. Work time is yoked to the strange needs of the enterprise, but idle time can be structured to satisfy hunter-gatherer instincts. The human population will regain the opportunity to organize its life in more natural patterns. A greener planet is a likely result of this ongoing process. As societies industrialize and become wealthy, increased consumption manifests itself in deforestation, pollution, and the like--to a point.
Further wealth reduces the manifestations of industry, by making the luxury of a greener environment affordable. Advancing technology widens the options from which individuals sculpt first their personal lives, but then also their communal world. The developed countries of America, Asia, and Europe began their green return in recent decades, as per-capita annual income grew beyond about $15,000. Many developing countries are just reaching this turnaround point. Advanced robots will reinforce the trend indirectly, by tremendously accelerating technological evolution and, for instance, allowing extreme processes to be moved to outer space. They will contribute directly by substituting for energy- and chemical-intensive industrial separation and shaping processes. A robot population far exceeding the human one will achieve the same end much more efficiently by tirelessly sorting and rearranging matter on a tiny scale with myriads of microscopic fingers.
Any choice has consequences: by comfortably retreating to its roots, biological humanity will leave the uncomfortable, uncharted but grand opportunities to others.
No less than today's organizations, fully automated companies will compete with one another not only in routine manufacture and distribution, but also in planning, development, and research. To robots built for it, outer space will offer unprecedented energy, materials, room, and perhaps freedom from taxation for these activities--a tremendous competitive advantage. Sooner rather than later, automated industry will grow away from earth. The space industries will continuously devise their own improvements, gaining rapidly in size, efficiency, diversity, and intelligence. The earthbound "consumer outlet" parts of the operation, while not shrinking in absolute size, will represent an ever-decreasing fraction of the whole. Old earth will become insignificant on the ever grander scale of earth-spawned activity.
Robot industries will start as conversions of existing enterprises, retaining their institutional, legal, and competitive structures. But then they will explore and exploit expanding non-traditional options, some very unhuman. Our artificial progeny will grow away from and beyond us, both in physical distance and structure, and in similarity of thought and motive. In time their activities may become incompatible with old earth's continued existence. Even so, it is likely that we, the historical root of their transcendence, will be preserved in some form--though, to us, the form may seem extremely strange. Just possibly, human personalities could participate in some way in the mainstream of this future activity, either under the wings of superintelligent hosts, or by being transformed into a compatible form--surely becoming very unhuman in the process.
There is an analogy between the evolution of the first living organisms from simpler chemical processes several billion years ago and the development of technical civilization from human manipulative and learning skills. Technical civilization, and the human minds that support it, are the first feeble stirrings of a radically new form of existence, one as different from life as life is from simple chemistry. Call the new arrangement Mind. Unlike Life alone, which learns from its past, but is blind to its future, Mind can choose among alternatives to imperfectly select its own destiny--even to amplify that very ability.
Chapter 2 reviews the state of the robot art, like a baby poised for sudden growth. The following chapters mix predictions with suggested actions. One of the lessons of chaos theory is that sensitive systems are impossible to predict but often easy to control. Under that model, the future can sometimes be predicted, if one steadily nudges events toward the prediction! Believable and physically possible predictions can themselves nudge, by inspiring work. When such proactive predictions miss, it is often because they overlook even more potent possibilities, rather than because they are unachievable.
In the thirteenth century Roger Bacon imagined high-speed worldwide travel--via seven-league boots, rather than flying conveyances. In the sixteenth century Leonardo da Vinci designed aircraft--powered by human muscle, rather than combustion engines. In the nineteenth century Jules Verne anticipated submarine warfare--against wooden sailing vessels, rather than armored battle fleets guarded by electronic senses and aircraft. Shortly thereafter, H. G. Wells anticipated a world of the distant future with humanity radically transformed--by Darwinian evolution, not directed engineering. Science fiction of the early twentieth century, inspired by the theories, inventions, and speculations of rocket pioneers like Konstantin Tsiolkovsky, Robert Goddard, and Hermann Oberth, is filled with spacecraft--guided by slide-rule-wielding human navigators, not digital computers (telephone, radio, and computers, in particular, and their dramatic applications, seem to have taken prognosticators by surprise). There are no large fleets of dirigible airships ferrying transatlantic passengers; faster and more manageable heavier-than-air craft displaced them.
Barring cataclysms, I consider the development of intelligent machines a near-term inevitability. Chapters 3 and 4 offer a scenario. Like airplanes, but unlike spaceships or radio, machine intelligences will be direct imitations of something already existing biologically. Every technical step toward intelligent robots has a rough evolutionary counterpart, and each is likely to benefit its creators, manufacturers, and users. Each advance will provide intellectual rewards, competitive advantages, and increased wealth and options of all kinds. Each can make the world a nicer place to live. At the same time, by performing better and cheaper, the robots will displace humans from essential roles. Rather quickly, they could displace us from existence. I'm not as alarmed as many by the latter possibility, since I consider these future machines our progeny, "mind children" built in our image and likeness, ourselves in more potent form. Like biological children of previous generations, they will embody humanity's best chance for a long-term future. It behooves us to give them every advantage and to bow out when we can no longer contribute.
But, as also with biological children, we can probably arrange for a comfortable retirement before we fade away. Some biological children can be convinced to care for elderly parents. Similarly, "tame" superintelligences could be created and induced to protect and support us, for a while. Such relationships require advance planning and diligent maintenance. Chapter 5 offers suggestions.
It is the "wild" intelligences, however, those beyond our constraints, to whom the future belongs. The available tools for peeking into that strange future--extrapolation, analogy, abstraction, and reason--are, of course, totally inadequate. Yet, even they suggest surreal happenings. Chapter 6's robots sweep into space in a wave of colonization, but their wake converts everything into increasingly pure thinking stuff. A "Mind Fire" will burn across the universe. Inside the Mind, considered in Chapter 7, physical law loses its primacy to purposes, goals, interpretations, and God knows what else.
Posted December 9, 2003
Hans Moravec is both a practical robotics engineer and a transcendent dreamer. Robot: Mere Machine to Transcendent Mind is a work of pessimism delivered by an optimist. It¿s complex, compelling, naïve and frightening. Is this the world we¿re building for our children? I mean human children, not mind children. Robot begins with a good overview of robotics, outlining the work of cyber-pioneers such as Alan Turing, John McCarthy and Marvin Minsky; then progressing into the late nineties. Moravec is a first-class robotics scientist and engineer, explaining technical issues and solutions in a concise, interesting manner. Good stuff, if you¿re interested in understanding robotics (why else would you be reading this book?). Moravec then projects the growth of robotics and artificial intelligence employing a model similar to that of Ray Kurzweil in The Age of Spiritual Machines. Both men base this growth on the exponentially increasing power of inexpensive computers, which they believe will match the computing power of the human mind by approximately 2020. They both present strong arguments that the human mind is fundamentally a complex machine; therefore, it¿s not a stretch to believe an equally complex mind can be developed in silicon. Moravec then provides his assessment of robot capabilities for each decade of the twenty-first century. It¿s fascinating and not unreasonable. Okay, we all know the future is not going to look like Star Trek, but Moravec¿s vision for the coming centuries is just too unbelievable: robot corporations in outer space, some planet-sized, virtually all of humanity living on a dole provided by taxing robot corporations, ¿execs¿ with almost supernatural powers ¿ you get the idea. Who knows ¿ maybe he will be right ¿ but these speculations don¿t fit well with the practical, science-based tone of the bulk of the material. Nevertheless, this is an excellent work if you enjoy thinking about the near-future. Read Kurzweil and Moravec back to back ¿ throw in Flesh and Machines by Rodney Brooks ¿ and you¿ll have a persuasive picture of what the next few decades may hold.Was this review helpful? Yes NoThank you for your feedback. Report this reviewThank you, this review has been flagged.