Review | Ray Kurzweil is (still, somehow) excited about humans merging with machines

Review | Ray Kurzweil is (still, somehow) excited about humans merging with machines

There is a meme, popular among tech world insiders, that distinguishes between two types of people: wordcels and shape rotators. Wordcels are humanists, effete creatures who trade in anachronisms like writing and philosophy. Shape rotators, in contrast, are staunchly modern, possessed of a ruthlessly practical intelligence. They are the movers and shakers — the engineers and programmers who congregate in Silicon Valley in hopes of remaking the world.

The wordcel/shape rotator dichotomy — a 21st-century update of the right-brained vs. left-brained taxonomy — may appear laughably reductive, but it is only half ironic. When tech billionaire, venture capitalist and Silicon Valley darling Marc Andreessen tweeted disparagingly about the other half — the wordcels — he was expressing a view that he seemed to hold sincerely. His cohort of digital disrupters is disdainful of an entire domain of human endeavor, which may explain why its denizens are so eager to shed their personhood and transform into machines.

Computer scientist and famed transhumanist Ray Kurzweil is less openly scornful of the arts than his peers are, but he is still a shape rotator — and by all accounts, an adept one. His not-so-subtle title at Google is “principal researcher and AI visionary,” and he is the author of a number of nonfiction books beloved by the technology fetishists of Silicon Valley. In “The Age of Spiritual Machines” (1999) and “The Singularity Is Near” (2005), he offered controversial prognostications about a strange near-future in which machines take over, solving all of the world’s ills and perhaps even having transcendent experiences along the way.

His impressive grasp of computing is on display once again in his disjointed and occasionally delusional new book, “The Singularity Is Nearer: When We Merge With AI.” Kurzweil is a refreshingly lucid expositor of complex technical concepts, but he suffers from the shape rotator’s characteristic deficiency: an incapacity to recognize the limits of his own understanding.

“The Singularity Is Nearer” is daunting to summarize — even for a wordcel like me — because it is so careless and careening. Kurzweil makes strident assertions about a wide range of subjects, from personal identity to the nature of consciousness to the future of medicine. Far from the sort of disciplined treatise we might expect from a veteran programmer, this book is a welter of free associations and shameless simplifications.

That is not to say every piece of it is sloppy or misguided. Kurzweil’s basic thesis — that “information technologies like computing get exponentially cheaper because each advance makes it easier to design the next stage of their own evolution” — is sensible, even wise. As he points out, “One dollar now buys around 600 trillion times as much computing power as it did when the GPS was developed.”

Unfortunately, Kurzweil rarely restricts himself to claims about the mechanics and history of AI. Instead, he ventures into foreign territory, with unfortunate results. History, he announces in the book’s first chapter, is nothing but the evolution of information processing, and it can be neatly and unproblematically delineated into “six epochs, or stages, from the beginning of our universe.” At the sixth and final stage, “our intelligence spreads throughout the universe, turning ordinary matter into computronium, which is matter organized at the ultimate density of computation.” Kurzweil is spared the indignity of attempting to explain what this means, because the sixth epoch and its mysteries remain in the distant offing. At present, we are approaching the fifth, when “we will merge with AI and augment ourselves with millions of times the computational power that our biology gave us.”

One chapter of “The Singularity Is Nearer” is devoted to the thesis that life has been “getting exponentially better” for centuries, but Kurzweil expects that life will improve even more dramatically when the singularity occurs: when machines surpass their human makers. In this brave new world of super-intelligent computers, 3D printers will enable us to produce enough clothing and housing for everyone, and AI will pioneer techniques that allow us to grow crops more efficiently. Meanwhile, sophisticated machine-learning programs will design innovative new medicines, and nanorobots will enter our bodies and kill all the wayward cells, effectively curing cancer. “As AI unlocks unprecedented material abundance across countless areas,” Kurzweil writes, “the struggle for physical survival will fade into history.” There will even be a remedy of sorts for death itself: We will go on chatting with the robot analogues of our loved ones after they pass away.

Kurzweil graciously concedes that there may be a few growing pains through these final epochs. Jobs will be automated before new forms of markets emerge, and violence may erupt in the interim. Humans struggling to adapt might be plagued by a sense of uselessness or inferiority. Eventually, however, we will stop complaining and rejoice at our liberation from the burdens of physicality: “Once our brains are backed up on a more advanced digital substrate, our self-modification powers can be fully realized.” And when we incorporate AI directly into our brains, it will not be a competitor but “an extension of ourselves.”

“Ray Kurzweil is the best person I know at predicting the future of artificial intelligence,” Bill Gates gushes in a blurb on the back cover of “The Singularity Is Nearer.” Despite this endorsement, it is reasonable to wonder whether the oracle of Silicon Valley is really so reliable. At one point, he goes so far as to speculate that, “as nanotechnology takes off, we will be able to produce an optimized body at will: we’ll be able to run much faster and longer, swim and breathe under the ocean like fish, and even give ourselves working wings if we want them.” At times, Kurzweil’s prophecies read like passages from messianic religious texts.

Yet even if his most outrageous technological predictions came true — even if we could cure cancer with nanobots and program ourselves to grow gills — paradise would remain elusive. Suppose 3D printers really did disgorge enough food to sustain the world’s population, or AI really did devise cures for the most obstinate diseases. Why should we expect these lifesaving goods to be distributed equitably, given that our already ample supply of food reaches so few and that so many of us lack access to the most basic health care? Kurzweil’s insipid answer — “we’ll need smart government policies to ease the transition and ensure that prosperity is broadly shared” — is not satisfying or particularly surprising.

From the first page of “The Singularity Is Nearer,” it is clear that, for Kurzweil, technological problems are the only ones that matter. Questions of political will and even ethical permissibility are afterthoughts. If a technology is available, then Kurzweil regards its adoption as not just inevitable but probably advisable. Progress is cast as an unstoppable force, rather than a product of effort and ingenuity.

He observes, for instance, that renewable-energy sources “accounted for about 1.4 percent of global electricity generation” in 2000 and 12.85 percent in 2021 — then sanguinely announces, “This progress will continue exponentially.” Never mind that reactionary regimes, many of them in denial about climate change, are poised to take (or have already taken) power in much of the Western world. Or that scientists warn that we are already past the point of no return. As long as the line on the graph has been climbing, Kurzweil believes it will go on climbing.

“History gives us reason for profound optimism” when it comes to the expansion of the political franchise, too, as illustrated by a neat little chart tracking the “Spread of Democracy Since 1800.” But this crude quantitative measure does not register the many qualitative reasons for alarm, among them the erosion of LGBTQ rights and expanding restrictions on abortion. Democracy is a fragile achievement, a perpetual work in progress, not the sort of thing that can be vouchsafed by a trend line — and certainly not the sort of thing that technology automatically generates.

It is, however, an obvious good, unlike many of the more disturbing advances that Kurzweil welcomes with open arms. What if Kurzweil and his Silicon Valley peers are alone in finding the prospect of merging with computers appealing? What if the rest of us are horrified by the idea of chatting with AI facsimiles of our dead loved ones? Indeed, many people have already chosen to abandon even the technologies this book heralds as irresistible, such as the Metaverse. Though Kurzweil frets over the most dramatic and fanciful ways that AI could betray us — by enabling bioterrorists to engineer a deadly pathogen, by self-replicating until nanobots take over — he never spares a thought for the million banal ways that recent technological breakthroughs have already made life unbearable. To him, it is irrelevant that misinformation spreads like wildfire, or that neo-Nazis are radicalized on YouTube, or that teenage girls swept up in a stream of diabolically addictive images often develop eating disorders. We can use all of these radical new technologies — and by his lights, that is enough to prove that we should.

There is no question that AI will soon outstrip humans in many respects, and Kurzweil is right to suspect that society will change considerably when computers can do our laundry, diagnose our diseases and conduct trials of new drugs. A sober assessment of what computers can and cannot contribute to the human project is long overdue, but unfortunately “The Singularity Is Nearer” is no such thing. A better guide to the future will have to take stock of the many aspects of personhood that Kurzweil ignores or dismisses. He is right that life has improved in many ways, of course, but what’s left out of his relentlessly futuristic picture are the many precious things that cannot be said to “advance.”

What about art? Kurzweil’s answer is that it, too, will benefit from an upgrade. “Actors can now convey what their character is thinking only through their words and external physical expressions,” but in virtual reality we will “have art that puts a character’s raw, disorganized, nonverbal thoughts — in all their inexpressible beauty and complexity — directly into our brains.” If he had been paying attention for the last century, he might have realized that modernist novelists employing stream-of-consciousness methods, among them Marcel Proust and Virginia Woolf, solved this problem long ago.

What about physical experience? In the “digital universe,” Kurzweil effuses, “many products won’t even need a physical form at all, as simulated versions will perform perfectly well in highly realistic detail.” Among these: “a sensory-rich virtual beach vacation for the whole family.” Only someone with no regard for sensory pleasures could imagine that digital simulations will ever be adequate substitutes for luxuries we can taste or touch.

The smug conclusion that “computers will be able to simulate human brains in all the ways we might care about” only serves to reveal which parts of the brain Kurzweil himself does not care about. Perhaps the shape rotators are convinced that computers can outthink us because their own minds are so impoverished. If they have ever encountered a painting or a poem, they no doubt blinked at it with affable perplexity, wondering how it might be optimized.

The Singularity Is Nearer