desolate beach,
Scattering remains of love.
TOMCAT
A haiku written by Ray Kurzweil’s Cybernetic Poet after reading poems by Randi and Kathryn Lynn
An old yellow tomcat
lies sleeping content,
he rumbles a heart
LONG YEARS HAVE PASSED
A poem written by Ray Kurzweil’s Cybernetic Poet after reading poems by Randi and Kathryn Lynn
Long years have passed.
I think of goodbye.
Locked tight in the night
I think of passion;
Drawn to for blue, the night
During the page
My shattered pieces of life
watching the joy
shattered pieces of love
My shattered pieces of love
gone stale.
SANDALS
A haiku written by Ray Kurzweil’s Cybernetic Poet after reading poems by Kimberly McLauchlin and Ray Kurzweil
Scattered sandals
a call back to myself,
so hollow I would echo.
The Visual Arts
The visual arts lie in between the musical arts and literary arts in terms of the impact of computers. A great deal of visual art is created using paint and illustrator programs, which can simulate the effects of conventional materials such as paint strokes as well as implement a wide range of techniques that could only be executed on a computer. Recently, computers have also taken over most video and film editing.
The Web is filled with the artistic musings of cybernetic artists. A popular technique is the evolutionary algorithm, which allows the computer to evolve a picture by redoing it hundreds or thousands of times. Humans would find this approach difficult—they would waste a lot of paint, for one thing. Mutator, the creation of sculptor William Latham and software engineer Stephen Todd at IBM in Winchester, England, uses the evolutionary approach, as does a program written by Karl Sims, an artist and scientist at Genetic Arts, in Cambridge, Massachusetts.8
Probably the leading practitioner of computer-generated visual art is Harold Cohen. His computerized robot named Aaron has been evolving and creating drawings and paintings for twenty years. These works of visual art are completely original, created entirely by the computer, and rendered with real paint. Cohen has spent more than three decades endowing his program with a knowledge of many aspects of the artistic process, including composition, drawing, perspective, and color, as well as a variety of styles. While Cohen wrote the program, the pictures created are nonetheless always a surprise to him.
Cohen is frequently asked who should be given credit for the results of his enterprise, which have been displayed in museums around the world.9 Cohen is happy to take the credit, and Aaron has not been programmed to complain. Cohen boasts that he will be the first artist in history who will be able to have a posthumous exhibition of completely original works.10
Paintings by Aaron by Cohen
These five original paintings were painted by Aaron, a computerized robot built and programmed by Harold Cohen. These color paintings are reproduced here in black and white. You can see the color versions on this book’s web site, at www.penguinputnam.com/kurzweil.11
PREDICTIONS OF THE PRESENT
With the impending millennium change there are no shortage of anticipations of what the next century will be like. Futurism has a long history, but not a particularly impressive one. One of the problems with predictions of the future is that by the time it’s clear that they have had little resemblance to actual events, it’s too late to get your money back.
Perhaps the problem is that we let just anyone make predictions. Maybe we should require futurism certification to be allowed to prognosticate. One of the requirements could be that in retrospect, at least half of your ten-or-more-year-ahead predictions have not been completely embarrassing. Such a certification program would be a slow process, however, and I suspect unconstitutional.
To see why futurism has such a spotty reputation, here is a small sample of predictions from some otherwise intelligent people:
“The telephone has too many shortcomings to be seriously considered as a means of communication.”
—Western Union executive, 1876
“Heavier-than-air flying machines are not possible.”
—Lord Kelvin, 1895
“The most important fundamental laws and facts of physical science have all been discovered, and these are now so firmly established that the possibility of their ever being supplemented by new discoveries is exceedingly remote.”
—Albert Abraham Michelson, 1903
“Airplanes have no military value.”
—Professor Marshal Foch, 1912
“I think there is a world market for maybe five computers.”
—IBM Chairman Thomas Watson, 1943
“Computers in the future may weigh no more than 1.5 tons.”
—Popular Mechanics, 1949
“It would appear that we have reached the limits of what is possible to achieve with computer technology, although one should be careful with such statements, as they tend to sound pretty silly in five years.”
—John von Neumann, 1949
“There’s no reason for individuals to have a computer in their home.”
—Ken Olson, 1977
“640,000 bytes of memory ought to be enough for anybody.”
—Bill Gates, 1981
“Long before the year 2000, the entire antiquated structure of college degrees, majors and credits will be a shambles.”
—Alvin Toffler
“The Internet will catastrophically collapse in 1996.”
—Robert Metcalfe (inventor of Ethernet), who, in 1997, ate his words (literally) in front of an audience
Now I get to toot my own horn, and can share with you those predictions of mine that worked out particularly well. But in looking back at the many predictions I’ve made over the past twenty years, I will say that I haven’t found any that I find particularly embarrassing (except, maybe, for a few early business plans).
The Age of Intelligent Machines, which I wrote in 1987 through 1988, as well as other articles and speeches I wrote in the late 1980s, contained a lot of my predictions about the 1990s, which included the following:12
• Prediction: A computer will defeat the human world chess champion around 1998, and we’ll think less of chess as a result.
What Happened: As I mentioned, this one was a year off. Sorry.
• Prediction: There will be a sustained decline in the value of commodities (that is, material resources) with most new wealth being created in the knowledge content of products and services, leading to sustained economic growth and prosperity.
What Happened: As predicted, everything is coming up roses (except, as also predicted, for long-term investors in commodities, which are down 40 percent over the past decade). Even the approval ratings of politicians from the president to the Congress are at an all-time high. But the strong economy has more to do with the Bill in the west coast Washington than the Bill in the east coast Washington. Not that Mr. Gates deserves primary credit, but the driving economic force in the world today is information, knowledge, and related computer technologies. Federal Reserve Chairman Alan Greenspan recently acknowledged that today’s unprecedented sustained prosperity and economic expansion is due to the increased efficiency provided by information technology. But that’s only half right. Greenspan ignores the fact that most of the new wealth that is being created is itself comprised of information and knowledge—a trillion dollars in Silicon Valley alone. Increased efficiency is only part of the story. The new wealth in the form of the market capitalization of computer-related (primarily software) companies is real and substantial and is lifting all boats.
The U.S. House Subcommittee on Banking reported that in the eight-year period between 1989.and 1997, the total value of U.S. real estate and durable goods increased only 33 percent, from $9.1 trillion to $12.1 trill
ion. The value of bank deposits and credit market instruments increased only 27 percent, from $4.5 trillion to $5.7 trillion. The value of equity shares, however, increased a staggering 239 percent, from $3.4 trillion to $11.4 trillion! The primary engine of this increase is the rapidly increasing knowledge content of products and services, as well as the increased efficiencies fostered by information technology. This is where new wealth is being created.
Information and knowledge are not limited by the availability of material resources, and in accordance with the Law of Accelerating Returns will continue to grow exponentially. The Law of Accelerating Returns includes financial returns. Thus a key implication of the law is continuing economic growth.
As this book is being written, there has been considerable attention on an economic crisis in Japan and other countries in Asia. The United States has been pressing Japan to stimulate its economy with tax cuts and government spending. Little attention is being paid, however, to the root cause of the crisis, which is the state of the software industry in Asia, and the need for effective entrepreneurial institutions that promote the creation of software and other forms of knowledge. These include venture and angel capital,13 widespread distribution of employee-stock options, and incentives that encourage and reward risk taking. Although Asia has been moving in this direction, these new economic imperatives have grown more rapidly than most observers expected (and their importance will continue to escalate in accordance with the Law of Accelerating Returns).
• Prediction: A worldwide information network linking almost all organizations and tens of millions of individuals will emerge (admittedly, not by the name World Wide Web).
What Happened: The Web emerged in 1994 and took off in 1995 through 1996. The Web is truly a worldwide phenomenon, and products and services in the form of information swirl around the globe oblivious to borders of any kind. A 1998 report by the U.S. Commerce Department credited the Internet as a key factor in spurring economic growth and curbing inflation. It predicted that commerce on the Internet will surpass $300 billion by 2000. Industry reports put the figure at around $1 trillion, when all business-to-business transactions conducted over the Web are taken into consideration.
• Prediction: There will be a national movement to wire our classrooms. What Happened: Most states (with the exception, unfortunately, of my own state of Massachusetts) have $50 to $100 million annual budgets to wire classrooms and install related computers and software. It is a national priority to provide computer and Internet access to all students. Many teachers remain relatively computer illiterate, but the kids are providing much of the needed expertise.
• Prediction: In warfare, there will be almost total reliance on digital imaging, pattern recognition, and other software-based technologies. The side with the smarter machines will win. “A profound change in,military strategy will arrive in the early 1990s. The more developed nations will increasingly rely on ‘smart weapons,’ which incorporate electronic co-pilots, pattern-recognition techniques, and advanced technologies for tracking, identification, and destruction.”
What Happened: Several years after I wrote the Age of Intelligent Machines, the Gulf War was the first to clearly establish this paradigm. Today, the United States has the most advanced computer-based weaponry and remains unchallenged in its status as a military superpower.
• Prediction: The vast majority of commercial music will be created on computer-based synthesizers.
What Happened: Most of the musical sounds you hear on television, in the movies, and in recordings are now created on digital synthesizers, along with computer-based sequencers and sound processors.
• Prediction: Reliable person identification, using pattern-recognition techniques applied to visual and speech patterns, will replace locks and keys in many instances.
What Happened: Person-identification technologies that use speech patterns and facial appearance have begun to be used today in check-cashing machines and to control entry into secure buildings and sites.14
• Prediction: With the advent of widespread electronic communication in the Soviet Union, uncontrollable political forces will be unleashed. These will be “methods far more powerful than the copiers the authorities have traditionally banned.” The authorities will be unable to control it. Totalitarian control of information will have been broken.
What Happened: The attempted coup against Gorbachev in August 1991 was undone primarily by cellular telephones, fax machines, electronic mail, and other forms of widely distributed and previously unavailable electronic communication. Overall, decentralized communication contributed significantly to the crumbling of centralized totalitarian political and economic government control in the former Soviet Union.
• Prediction: Many documents never exist on paper because they incorporate information in the form of audio and video pieces.
What Happened: Web documents routinely include audio and video pieces, which can only exist in their web form.
• Prediction: Around the year 2000, chips with more than a billion components will emerge.
What Happened: We’re right on schedule.
• Prediction: The technology for the “cybernetic chauffeur” (self-driving cars using special sensors in the roads) will become available by the end of the 1990s with implementation on major highways feasible during the first decade of the twenty-first century.
What Happened: Self-driving cars are being tested in Los Angeles, London, Tokyo, and other cities. There were extensive successful tests on Interstate 15 in southern California during 1997. City planners now realize that automated driving technologies will greatly expand the capacity of existing roads. Installing the requisite sensors on a highway costs only about $10,000 per mile, compared to $1 to $10 million per mile for building new highways. Automated highways and self-driving cars will also eliminate most accidents on these roads. The U.S. National Automated Highway System (NAHS) consortium is predicting implementation of these systems during the first decade of the twenty-first century.15
• Prediction: Continuous speech recognition (CSR) with large vocabularies for specific tasks will emerge in the early 1990s.
What Happened: Whoops. Large-vocabulary domain-specific CSR did not emerge until around 1996. By late 1997 and early 1998, large-vocabulary CSR without a domain limitation for dictating written documents (like this book) was commercially introduced.16
• Prediction: The three technologies required for a translating telephone (where you speak and listen in one language such as English, and your caller hears you and replies in another language such as German)—speaker-independent (not requiring training on a new speaker), continuous, large-vocabulary speech recognition; language translation; and speech synthesis—will each exist in sufficient quality for a first generation system by the late 1990s. Thus, we can expect “translating telephones with reasonable levels of performance for at least the more popular languages early in the first decade of the twenty-first century.” What Happened: Effective, speaker-independent speech recognition,
MY LIFE WITH MACHINES: SOME HIGHLIGHTS
I walked onstage and played a composition on an old upright piano. Then came the yes-or-no questions. Former Miss America Bess Myerson was stumped. But film star Henry Morgan, the second celebrity panelist on this episode of I’ve Got a Secret, guessed mysecret: The piece I had played had been composed by a computer that I had built and programmed. Later that year, I got to meet President Johnson with other high-school science winners.
In college, I ran a business matching up high-school kids with colleges using a computer program I had written. We had to pay $1,000 an hour to rent time on the only computer in Massachusetts with an extraordinary million bytes of core memory, which allowed us to fit all the information we had about the nation’s three thousand colleges into memory at the same time. We received a lot of letters from kids who were delighted with the college that our program had suggested. A few parents, on the other hand, were furious that we had failed to rec
ommend Harvard. It was my first experience with the ability of computers to affect people’s lives. I sold that company to Harcourt, Brace & World, a New York publisher, and moved on to other ideas.
In 1974, computer programs that could recognize printed letters, called optical character recognition (OCR), were capable of handling only one or two specialized type styles. I founded Kurzweil Computer Products that year to develop the first OCR program that could recognize any style of print, which we succeeded in doing later that year. So the question then became, What is it good for? Like alot of clever computer software, it was a solution in search of a problem.
I happened to sit next to a blind gentleman oa a plane flight, and he explained to me that the only real handicap that he experienced was his inability to rad ordinary printed material. It was clear that his visual disability imparted no real handicap in either communicating or traveling. So I had found the problem we were searching for—we could apply our “omni-font” (any font) OCR technology to overcome this principal handicap of blindness. We didn’t have the ubiquitous scanners or text-to-speech synthesizers that we do today, so we had to create these technologies as well. By the end of 1975, we put together these three new technologies we had invented-omni-font OCR, CCD (Charge Coupled Device) flat-bed scanners, and text-to-speech synthesis to create the first print-to-speech reading machine for the bilind. The Kurzweil Reading Machine (KRM) was able to read ordinary books, magazines, and other printed documents out loud so that a blind person could read anything he wanted.