ARTS & IDEAS

        The Soul of the Next New
        Machine: Humans
 

        How the Wedding of Brain and Computer
        Could Change the Universe

        By ROB FIXMER

              When Ray Kurzweil discusses human destiny, it is
              not always clear whether he's talking about
        technology or theology.

        It is technology that
        defines his résumé; he
        has spent 34 of his 51
        years inventing
        ingenious uses for
        artificial intelligence.
        But like a priest
        caught playing in a
        physics lab, he keeps
        coming up with
        inventions inspired by
        aesthetics and social
        conscience.

        For instance, when he
        was still in high
        school, he wrote a
        program that composed music, while his latest
        software, available at his Web site, writes poetry. In
        between he created machines that read to the blind,
        software that draws and paints, electronic keyboards
        that produce the sounds of acoustical instruments and
        one of the most advanced and commercially successful
        forms of computer speech recognition.

        All were concrete products of a restless mind
        consumed by the question of what will be. A more
        abstract product of that vision is his latest book, "The
        Age of the Spiritual Machine: When Computers Exceed
        Human Intelligence" (Viking, 1999). In it, he looks at
        the exponential increase in calculating power since the
        turn of the century and concludes that within 50 years,
        machines will not only be smarter than humans but also
        smart enough to persuade us that they are conscious
        beings in their own right.

        That assertion, not surprisingly, has drawn the wrath of
        several prominent philosophers who question his
        definitions of both intelligence and consciousness.

        John R. Searle, the renowned University of California
        at Berkeley professor of philosophy, wrote in The New
        York Review of Books that the fatal flaw in
        "Kurzweil's entire argument" is that "it rests on the
        assumption that the main thing humans do in their lives
        is compute." In a phone interview this week he added:
        "Kurzweil's proposals are preposterous science. I think
        he got a little carried away and made massive
        philosophical errors, too."

        But while the debates about what defines intelligence
        and consciousness have gained the most public
        attention, the far more compelling idea in the book is
        his prediction that our progeny -- if not some of us alive
        today -- are destined to be human-machine hybrids.
        Based on current trends in computer and biological
        sciences, he says that a superpowered intelligence will
        result from such a hybrid. The merging of human brains
        and computer circuits, he asserts, will enable the
        species to redesign itself and determine not just its own
        destiny but the fate of the universe. To become, in other
        words, God-like.

        "In my view," Kurzweil writes in the book's
        conclusion, "the primary issue is not the mass of the
        universe or the possible existence of antigravity or of
        Einstein's so-called cosmological constant. Rather the
        fate of the universe is a decision yet to be made, one
        which we will intelligently consider when the time is
        right."

        In his review Searle writes in an
        amazed tone: "Kurzweil does not
        think he is writing a work of science
        fiction or a parody or satire. He is
        making serious claims that he thinks
        are based on solid scientific
        results."

        Over dinner at a Boston hotel,
        Kurzweil insisted, "This stuff isn't
        science fiction." His voice was soft,
        almost emotionless, but the ideas
        emerged fully formed, articulated
        with machinelike precision.

        On this night, he was clearly multi-tasking -- answering
        questions and expounding on complex technologies
        without so much as a pause for a breath, his eyes all the
        while parsing the salmon entree before him. (He takes
        eating very seriously indeed. His previous book, "The
        10 Percent Solution for a Healthy Life," was about
        nutrition, specifically about a low-fat diet he put
        together to battle his own congenital heart disease.)

        The "stuff" he's talking about is no less than a physical
        hybrid of human beings and their technology. He says
        the machines being created today are the beginning of
        our metamorphosis from thinking mammal to
        all-knowing hybrid.

        Biological evolution has already given way to much
        more rapid and less random technological evolution,
        Kurzweil argues. And within 30 years, he predicts,
        direct links will be established between neurons in the
        human brain and computer circuitry. The implications
        are mind-boggling. Such links would mean that the
        entire contents of a brain could be "insubstantiated" --
        copied (and preserved) in an external database. Not
        only would the brain's biological capacity be
        supplemented with enormous amounts of digital
        memory, it would also be linked to vast information
        resources like the Internet at the speed of thought.

        And it would produce, through direct neural
        stimulation, a virtual reality indistinguishable from
        objective reality.

        Kurzweil cites medical treatments in which silicon
        chips have been successfully implanted in human
        brains, for example, to alleviate symptoms of
        Parkinson's disease, or been made to communicate with
        neurons, as they do in cochlear implants for the deaf, as
        examples of primitive steps toward his predictions.

        While these sorts of visions might seem far-fetched,
        other respected futurists find Kurzweil's ideas
        compelling. Marvin Minsky, Toshiba professor of
        media arts and science at the Massachusetts Institute of
        Technology, has called him "a leading futurist of our
        time," while the techno-guru George Gilder says, "This
        book makes all other roads to the computer future look
        like goat paths in Patagonia."

        That Kurzweil's theories are given a serious hearing is
        testament to his credentials. Since his teenage years he
        has been harnessing computer power to "pattern
        recognition," which Kurzweil describes as "that part of
        the artificial intelligence field where we teach
        computers to recognize abstract patterns, a capability
        that dominates human thinking." In 1965, at age 17, his
        music composing program won a Westinghouse science
        award, a visit to the White House and a spot as a
        contestant on the old television game show "I've Got a
        Secret." (Young Ray's secret stumped the former Miss
        America, Bess Myerson, but was guessed by the second
        panelist, the actor Henry Morgan.)

        By the time he graduated from M.I.T. in 1970, Kurzweil
        had already achieved his first business success, having
        founded a computer database service that helped high
        school students choose the right college. That endeavor,
        which he sold for $100,000, was followed by a string
        of successful businesses built on his inventions. Among
        his best-known is the Kurzweil reading machine for the
        blind, which was a true marvel when it was introduced.
        CBS News was so impressed with the device that
        Walter Cronkite used it to deliver his signature
        sign-off, "And that's the way it was, on Jan. 13, 1976."

        Along the way, Kurzweil has won a raft of accolades in
        both business and academia and received nine honorary
        doctorates. Today, in addition to writing his next book,
        he is developing Fat Kat, an artificial intelligence
        system that applies evolutionary algorithms to
        securities investment decisions.

        In Kurzweil's vision of the future, the man-machine
        hybrid will be accomplished not through some
        Frankenstein-like amalgam but through an elegant
        technology: microscopic, self-replicating robots called
        nanobots that will be introduced through the
        bloodstream and will interact with individual neurons
        throughout the brain.

        "The idea," Kurzweil said, "is to direct nanobots to
        travel through every capillary in the brain, where they
        will pass in very close proximity to every neural
        feature." This, he says will enable the tiny machines to
        scan each neuron and "build up a huge database" --
        basically, the entire contents of a brain.

        "And all these nanobots would be communicating with
        each other," he said. "They'd basically be on a wireless
        network. They could also be on the Web and
        communicating with computers maintaining the
        database outside the brain."

        The breakthrough in nanotechnology came several years
        ago with the discovery of the nanotube, a carbon
        molecule of enormous strength. Just about anything can
        be fabricated from nanotubes, at many times the strength
        and at a fraction of the weight possible with
        conventional materials.

        What's more, nanotubes have a far greater capacity for
        raw computing power than the commonly used silicon.
        This combination of features holds out the possibility of
        building machines no larger than a blood cell that are
        fully programmable and perhaps even able to construct
        replicas of themselves from carbon atoms.

        The size of the technology is shrinking so rapidly,
        Kurzweil says, that "within 30 years, both the size and
        cost of this scenario will be feasible."

        Of course, such technology would inevitably be
        accompanied by terrifying dangers. By scanning a brain
        into a database, a person's most private thoughts and
        memories would be vulnerable to intrusions by
        hackers. And wouldn't the brain also be vulnerable to
        external control of information, thought processes and
        even perceptions of reality?

        "Those are real concerns," Kurzweil admits.
        "Organizations like governments or religious or
        terrorist groups or just clever individuals could put
        nanobots in food or water supplies, trillions of them.
        These would then make their way inside people and
        would monitor their thoughts and even could control
        them and place them into virtual environments."

        But he adds: "We won't be defenseless. We have these
        concerns today at a primitive level with Trojan horses
        that make their way into our computers."

        In any case, Kurzweil says there is no turning back.
        Once evolution produced a technological species, it
        locked us into a relentless quest for understanding and
        control of our universe.

        "I tend to be optimistic, but that's more of a personal
        orientation than something I could scientifically argue
        for," he admits. "There definitely are dangers, and we
        do tend to address them imperfectly, so there is some
        possibility that this will fail."