things magazine / about / what's new? / archive / photos / projects / order
things 12
summer 2000
Sony's Aibo 'entertainment robot'
caption link / issue index / archives
Jonathan Bell
Forward thinking

Ray Kurzweil: The Age of Spiritual Machines
London: Phoenix, 1999, 484 pages, £8.99

How often do you find yourself talking to your computer? Does it ever seem to be conspiring against you? As you sit, facing the screen, day in, day out, do you imbue the stark grey boxes (or glossy translucency, if you’re an Apple fan) with something approaching a personality? 

Ray Kurzweil believes that one day, in the not too distant future, computers will overtake human intelligence, and technology will evolve to produce a form of consciousness that is inseparable from our own. Kurzweil is a futurologist specialising in artificial intelligence. Unlike his peers, whose senses, it seems, are often confused by the heady scent of prediction, he practises what he preaches, overseeing a raft of companies (including Kurzweil Applied Intelligence, Kurzweil Music Systems and Kurzweil Educational Systems),(1) all dedicated to emulating and enhancing human behaviour in a number of fields. Futurologists generally work themselves into two camps - utopians and dystopians. Kurzweil firmly places himself in the former camp, and The Age of Spiritual Machines, (2) a follow-up to 1990’s The Age of Intelligent Machines, is a eulogy to the forthcoming silicon age, a glorious future with possibilities far beyond our current limited compre-hension. With the fundamental premise that in the not too distant future, machine and human will live side by side in cybernetic symbiosis, is the Age of Spiritual Machines realism or virtual reality?

The Age of Spiritual Machines is both chronicle and almanac, a combination of straightforward technological history with a speculative – and ever-optimistic – look to the future. The book is best read as a guidebook to the future, a highly educated snapshot of the almost unimaginable. As the author himself states in his ‘note to the reader’, we have ‘choices to make in our path through this book’, and my copy certainly became dog-eared and broken-spined as I flipped backwards and forwards between the copious appendices and footnotes. Such a laborious structure lends itself best to electronic documents, which can be linked by the click of a mouse. In print form, however, it swiftly grows tiresome. Each chapter is terminated by the transcript of a ‘chat’ between us – his readers – and Kurzweil. These conver-sations are an irritating literary device whose main characteristics are a predilection for weak humour, Californian surf-speak and an irritating propensity to state the obvious. Kurzweil is no cyberpunk author or Generation X prophet, but although these intrusive futuristic dialogues smack of a creative writing exercise, they also convey the vast technological leaps that Kurzweil feels sure will transform our lives.

 Kurzweil stakes his predictions on the total acceptance of Moore’s Law, the maxim that computer power roughly doubles every 18 months, with a corresponding fall in price. This equation, which – as Kurzweil demonstrates has held broadly constant since the pioneering days of Charles Babbage in the last century but one – is a form of evolution, a technological evolution that will ultimately result in desktop computers with the same processing power as the human brain by around 2025. Moore’s Law can be thwarted only by the so-called silicon limit, the physical barriers on our ability to go on reducing processor size. Kurzweil deftly side-steps these problems, arguing that by the time this limit is reached, we’ll have long since mastered the delirious indecision promised by the quantum computer, harnessing atoms themselves in a kind of mathematical stew, eschewing the simple binary dualism of the bit for the qu-bit, which simultaneously tries all possible solutions and yet instantaneously arrives at the right answer. The quantum computer, he speculates, could be created at molecular level, meaning that a cup of strong coffee, for example, could, theoretically contain millions of quantum computers, if only we knew how to harness them.(3)

Until now, artificial intelligence has been one of the great sleights of hand of the computing age. Granted, any machine can be made to appear to be intelligent, but it takes more than a pre-programmed set of responses - however dextrous and appropriate – to imitate the human mind.(4) Kurzweil confidently predicts that machine intelligence will far exceed human intelligence by 2020.(5) Other futurologists have not been so optimistic in their estimates. But these scenarios, beloved of sci-fi writers from Mary Shelley’s Frankenstein to Philip K.Dick’s Blade Runner, via the alarmist (and populist) visions of Michael Crichton (Westworld, Jurassic Park), James Cameron (The Terminator) and the Wachowski brothers (The Matrix) tell of a different future – dystopian societies of robotic humanoids which (or who) yearn to throw off the yokes of their tyrannical human masters, or where our alleged mastery of technology is revealed to be a hollow sham, and the systems that humankind has tampered with rise up to reveal their essential unmanageability. 

If machines can be elevated to a form of consciousness, Kurzweil confronts us with the possibility of a roomful of computers, learning their ‘pasts’, and their knowledge of the world, by rote. This brings to mind the psychotic HAL 9000 from Stanley Kubrick’s 2001, taught from ‘birth’ as if it were a child. As HAL becomes increasingly demented, its education is spooled out, randomly: 

Say, Dave... The quick brown fox jumped over the fat lazy dog... The square root of pi is 1.7724538090... log e to the base ten is 0.4342944 ... the square root of ten is 3.16227766... I am HAL 9000 computer. I became operational at the HAL plant in Urbana, Illinois, on January 12th, 1991. My first instructor was Mr. Arkany. He taught me to sing a song... it goes like this... ‘Daisy, Daisy, give me your answer do. I’m half; crazy all for the love of you...’

We’re drifting again, back and forth, comparing fiction with fact and speculation with reality. Essentially, however, that is how this book reads, dazzling you with clarity and logic to underpin the fantastic, then plummeting back to reality.

Can computers be taught to comprehend the nuances and dual meanings inherent in everyday speech? Kurzweil’s celebration of the brain concludes with his opinion that one day, computers will be configured in a similar, but vastly superior, manner to the human brain.(6) They will therefore be able to learn. The now celebrated sentence, ‘Time flies like an arrow’,(7) used to highlight the complexity that context brings to simple language, is cited here as an example of how the things we take for granted might not be immediately apparent to a computer. Kurzweil quotes Marvin Minsky: ‘Birds can fly, unless they are penguins and ostriches, or if they happen to be dead, or have broken wings, or are confined to cages, or have undergone experiences so dreadful as to render them psychologically incapable of flight.’ If our machines are to make the same intuitive leaps in their comprehension of how the world works, our understanding of how we do it – how our own brains function – must make similarly bold advances.

So where is all this leading? One potential use for the imminent wholesale miniaturisation of our means of communicating, creating and archiving is in the development of virtual reality, that holy grail of computer technology. Ever since the author Howard Rheingold described our tentative steps in this brave new imaginary world,(8) virtual reality has attained global recognition, if not usage. Will the ‘haptic’ tactile interfaces of the future resemble the ‘utility fog’ described here? (9) How will governments cope with freely available massive computing power? And how will society respond to the wide availability of their snail-like trace through their daily digital meanderings? Kurzweil touches on the cryptographic con-undrums and the problems besetting the internet’s quest for privacy (it’s illegal to export anything stronger than 40-bit cryptography from the US), but seems content to meekly accept that in the future, ‘15 minutes of fame’ will be willingly traded for ‘15 minutes of privacy’.

For Kurzweil, the crux of his book is that evolution, now represented by technological evolution, is continuing at an exponential rate. Currently, we’re in the ‘knee’ of this development curve, so while today’s programs and systems might appear to be smart, they will be exponentially outpaced by their descendants in a few years time. Our current ventures into the world of artificial intelligence include Brutus.I, an electronic author that specialises in short, snappy and somewhat open-ended tales of betrayal, as well as Kurzweil’s own Ray Kurzweil’s Electronic Poet. All of these rely on existing knowledge, building up databases through ‘reading’ other texts. But we’re a long way from seamless integration – even Kurzweil admits that the dreaded grammatical pedantry of the widely used word processor Word is not for him. Ray Kurzweil’s Electronic Poet’s output might smack of the satisfactory composition that those of us with neither rhyme nor inclination can snap together in five minutes on the front of a fridge, but, nonetheless, the ‘language model’ used – whereby the program reads, learns and inwardly digests the author it is to imitate – is a sign of things to come. For now, machines seek us to flatter us through imitation. When will these roles be reversed? 

In a section entitled ‘Predictions of the present’, Kurzweil repeats the well-known blunders of scientists and businessmen attempting to predict the future. It’s easy to laugh with the benefit of hindsight, amazed that, from their position of knowledge, these people didn’t ‘know better’. Alvin Toffler’s seminal Future Shock (1970), subtitled ‘a study of mass bewilderment in the face of accelerating change’, wildly under-estimated our potential for absorbing and accommodating the then nascent technological revolution and bombarded its readers with scenarios that challenged their existing systems, claiming, for example, that ‘long before the year 2000, the entire antiquated system of degrees, majors and credits will be a shambles. No two students will move along the same educational track.’(10) And that’s precisely the dilemma this book places the reader. To criticise its claims would be to add to futurism’s ‘spotty reputation’, as Kurzweil calls it. His argument that evolution occurs exponentially appears to be borne out by the extensive appendix charting technological developments. At first, inventions are sparse and few, but gradually they increase, coming thick and fast – the Law of Accelerating Returns. 

Kurzweil’s relentlessly utopian prognosis contrasts strongly with a darker streak running through contemporary technological writing, especially in America. Kurzweil himself presents us with a piece of anonymous anti-technological writing. This is a game anyone can play: ‘You can’t get rid of the “bad” parts of technology and retain only the “good” parts. Take modern medicine, for example. Progress in medical science depends on progress in chemistry, physics, biology, computer science and other fields’.(11)  When taken out of context, sentences like these don’t sound so strange. We are forced to admit, grudgingly perhaps, that, yes, they contain some elements of truth. The author, Kurzweil reveals, is Theodore Kaczynski, otherwise known as the Unabomber, the solitary scientist who wreaked death and mutilation on America’s scientific community via a series of elaborately hand-built letter bombs. In any case, Kaczynski’s extremist reaction is redundant, Kurzweil notes – ‘we’ve passed the point of no return’ and society is wholly dependant on computers. 

That this salient fact underpins both Kaczynski’s infamous manifesto as well as Kurzweil’s book is indication of the variety of paths and interpretations open still open to us. The traditional dystopian view, where machines enslave weak humans, is somehow interchangeable by a technology-filled future, with benefits for all. At this stage in the game, it’s hard to tell which way the road will fork. 

As befits someone who has clearly profited from his ability and prescience, Kurzweil’s faith in the free market and the economy of boundless choice also guides his visions – ‘the road we’re going down is a road paved with gold,’ he notes, injecting a note of misplaced materialism into an ostensibly intellectual debate. At times, his relentless optimism seems misplaced: 

It may seem that will have too many choices. Today, we only have to choose our clothes, makeup, and destination when we go out. In the late twenty-first century, we will have to select our body, our personality, our environment – so many different decisions to make! But don’t worry – we’ll have intelligent swarms of machines to guide us.(12)

There is no doubt that things are going to go on getting smaller and smarter. There is probably little doubt that as wireless data networks become ubiquitous, the need for information to be associated with a physical object will rapidly decrease.(13) ‘The accelerating pace of change is inexorable,’ Kurzweil claims, and to argue seems Luddite and ignorant. But the questions and themes raised in this book, as well as the scenarios suggested, are, for the most part, discomforting. If my Mac crashes three times in one day, am I likely to want to trust my brain, or any other part of my body for that matter, to a marauding swarm of self-replicating nanobots? 

There are those of use for whom the creativity involved in imbuing inanimate objects with feeling and personality is a positive decision. The idea that our frequent cathartic rants at inanimate objects, especially technology, might one day elicit a coherent, reasoned and even emotional response is frightening. When do computers stop being ‘things’? We already anthropomorphise our computers to a shocking extent, and that’s even before robotic pets have made a widespread impact on our lives (the forerunners, such as Sony’s Aibo,(14)  induce levels of devotion amongst owners equivalent to their flesh and blood counterparts). Heaven knows how we’d suffer if our ‘mine is bigger than yours’ mentality had to be halted out of sensitivity to the machine.

If society today is totally dependent on computers – for they hold our collective memories, records and files – then in the future this dependency can only increase, as we hand over more and more control to our silicon masters. Will we become a nation of specialists, increasingly de-humanised by the vast knowledge needed to understand and survive in a complex society? Kurzweil counters these fears by pointing out that we’ll be able to artificially enhance our capacity for learning, memory, vision, etc. At the same time, the chasm between rich and poor will yawn ever wider, as the oft-cited ‘information underclass’ finally find themselves almost entirely cut-off from meaningful interaction with society. Kurzweil touches on the problem of these information paupers, as well as the increasing problem of large numbers of intelligent people without anything to actually do. Regardless of the continuously falling prices and rising power, will any computing device be created that has a truly positive impact on all echelons of society? As Kurzweil’s fictional ‘reader’ says of her nephew, ‘It’s not just that he blocks out real reality, it’s that he seems to avoid interacting with real people.’ When the definition of a real person becomes blurred and confused, will we notice? In 1960, there were 6,000 computers in America. In 2060, will there be only 6,000 humans?


1). A list of companies founded by Ray Kurzweil can be found at
2). The book has it’s own website at, including selected chapters as well as the vast array of links cited in the appendixes. Kurzweil himself can be contacted at
3). The author Douglas Adams predicted similar in his science fiction satire The Hitchhiker’s Guide to the Galaxy. The starship Heart of Gold’s Infinite Improbability Drive required a strong Brownian motion generator in order to function – in this case, a cup of revolting tea made by the useless on-board vending machine. 
4). Experiments in ‘evolving’ computer circuitry are already underway in halls of academia (see this article in the New Scientist for details). The New Scientist has a large archive on AI topics. 
5). Coincidentally around the same date that software programmer Bill Joy asserted, in his lengthy essay ‘Why the future doesn’t need us’, that we will be running scared from a new breed of ‘intelligent’ machines, scuttling along imperviously, flocking, grouping and swarming and generally making our lives a misery. Archived here
6). The visible human project is located at here
7). There are generally agreed to be five definitions of this phrase:

1) Time passes as quickly as an arrow passes. 
2) One should time flies in the same manner in which one would time an arrow. 
3) One should time flies in the same manner in which an arrow would time flies. 
4) One should only time flies that resemble an arrow, that is to say ‘time flies’.
5) The species of flies known as ‘time flies’ (see 4) like to have an arrow. 

8). Virtual Reality, Howard Rheingold, Touchstone Press, 1992.
9). Utility Fog is the name given to a highly advanced theoretical form of VR, proposed by BC Crandall. A dense crowd of nanobots, able to form themselves instantaneously into any shape or form, completely re-creating environments.
10). Future Shock, p241. 
11). Excerpted from The Unabomber Manifesto, point 121. The full text is available on the web
12). The Age of Spiritual Machines, page 184.
13). However, as anyone who works with computers knows, paper consumption is vastly increased. In 1959, the US consumed 7 million tons of paper. By 1986, this figure had risen to 22 million tons. 
14). The second generation has his (?) very own website

things 12, summer 2000

to top of page