I once believed in the imminence of superhuman intelligence. In 1981, when I was still in college, I took a science-writing class at Columbia University from the journalist Pamela McCorduck. She had just written Machines Who Think (note the mischievous "Who"), a book about the efforts of Marvin Minsky and other artificial-intelligence pioneers to create conscious, autonomous computers that would leave mere humans in their cognitive dust. This research, which McCorduck often enthused over in class, helped persuade me to become a science journalist. What could be cooler than witnessing this giant leap forward in the evolution of consciousness?
My youthful infatuation with AI gives me a somewhat jaded perspective on the prophecies of some modern scientists, notably the computer entrepreneur Ray Kurzweil, that we are on the verge of a "Singularity." In physics, a "singularity" is an event or place, like the big bang or a black hole, where the laws of physics are stretched to the breaking point. Singularitarians (which some call themselves) have adopted the term to describe a radical transformation of consciousness that will result from breakthroughs in artificial intelligence as well as nanotechnology, biotechnology and neuroscience.
At first, Singularitarians say, we may become cyborgs, as WiFi-equipped brain chips, nanobots and genetic modifications soup up our intelligence, perception and memory. Eventually, we may abandon our flesh-and-blood selves entirely and transform our psyches into giant software programs, like Vista but presumably less buggy. We will then "upload" ourselves into computers and dwell forever in cyberspace. Our transformation into immortal, God-like cyberbeings will supposedly take place not millennia or centuries from now but within the next few decades.
Kurzweil makes his case by showing graph after graph documenting the explosive progress over the past few decades in information technologies, including computers, the Internet, cell phones and so on. These advances are indeed astounding. The problem is, some crucial scientific endeavors have made little or no progress over the past few decades. For example, artificial-intelligence researchers, including Kurzweil himself, have invented programs that can read text, translate languages, recognize speech, interpret cardiograms and play championship-level chess. But they haven't come to close to creating the kind of artificial intelligence depicted in countless sci-fi films, from the cute R2D2 of Star Wars to the evil HAL of 2001: A Space Odyssey. Marvin Minsky, who served as an adviser for 2001, noted recently that computers still lack the common sense of a 4-year-old child; they can't even tell the difference between cats and dogs. "I wish I could tell you that we have intelligent machines," Minsky lamented, "but we don't."
Singularitarians such as Kurzweil insist that scientists will soon "reverse-engineer" the brain so that they understand exactly how it works. Many neuroscientists assume that, just as computers operate according to a machine code, so the brain's performance must depend on a "neural code"; this is the set of rules, syntax or algorithms that transforms electrical impulses emitted by brain cells into perceptions, memories, meanings, intentions. Researchers are trying to decode the brain by probing it with ever-more-powerful technologies, such as magnetic-resonance imaging, positron-emission tomography and microelectrodes.
Cracking the neural code should yield all sorts of benefits. First, the brain's programming tricks could be transferred to computers to make them smarter. Moreover, given the right interface, our brains and computers could communicate as readily as Macs and PCs. Eventually, our personal software could be extracted from our bodies for uploading into computers.
If a neural code exists, however, neuroscientists still have no idea what it is. Far from converging on a solution, scientists cannot agree whether information is represented primarily by signals from individual neurons, or brain cells, by oscillations of many neurons firing in tandem, by even higher-level waves of chaotic electrical activity sweeping through the brain or all of the above.
This lack of understanding is reflected in the slow pace of research on so-called neural prostheses, which replace or supplement capacities lost because of damage to the nervous system. The only truly successful neural prosthesis is the artificial cochlea. More than 100,000 hearing-impaired people have been equipped with those devices, which restore hearing by feeding signals from an external microphone to the auditory nerve.
Artificial retinas, light-sensitive chips that mimic the eye's signal-processing ability and stimulate the optical nerve or visual cortex, have been tested in a handful of blind subjects, but most have been able to see nothing more than phosphenes, or bright spots. A few paralyzed patients have learned to control a computer cursor "merely by thinking," as the media invariably put it, via implanted electrodes that pick up the patients' neural signals—but communicating that way remains slow and unreliable.
Kurzweil has excessive faith not only in artificial intelligence and neuroscience but also in biotechnology. In his book Fantastic Voyage: Live Long Enough to Live Forever, co-written with a physician, he claims that the Human Genome Project and related advances will soon yield cures both for specific diseases and even aging itself. The identification of specific genes underpinning inherited disease—such as Huntington's disease and immune-deficiency syndrome—has inspired researchers to devise therapies to correct the genetic malformations. So far, scientists have carried out hundreds of clinical trials of gene therapy, and not one has been an unqualified success. The treatments have sickened some patients and even killed them.
The record of cancer treatment is also dismal. Since 1971, when President Richard Nixon declared a "war on cancer," the United States has spent more than $70 billion on research, and the annual budget for the National Cancer Institute has increased by a factor of 20, from $250 million to $5 billion. Scientists have gained a much better understanding of the molecular and genetic underpinnings of cancer, but this knowledge has not yielded significant improvements in treatments. Overall cancer mortality rates in the U.S. actually rose from 1971 until the early 1990s before declining slightly over the last decade or so—largely because of a decrease in the number of male smokers. Given the track record of treatments for inherited diseases and cancer, proclamations about the imminence of immortality are absurd.
Part of me—the part that thrilled at prospects for artificial intelligence almost 30 years ago—finds Kurzweil's prophesies highly entertaining. He raises lots of provocative questions: What would be like to be immortal? To have an IQ of 1,000? To exist not as a doomed, flesh-and-blood creature but as a piece of software that can keep redesigning itself and merging with other programs? But another part of me—the grown-up, responsible part—worries that so many people, smart people, are taking Kurzweil's sci-fi fantasies seriously. The last thing humanity needs right now is an apocalyptic cult masquerading as science.