This pseudo-religious, pseudo-scientific tome by one of the world's leading futurists lays out a timeline to a moment when the pace of technological innovation -- specifically, in the power of computer hardware and software -- becomes so rapid that, from our perspective in the present, it occupies a single point in time. Hence, the singularity.

Kurzweil's predictions are based primarily upon an observation made by Gordon Moore of Intel: that computing power per dollar will double every 18 months. This isn't exactly what he said; it's actually an expression of transistor density. But the effect that consumers and Kurzweil care about is price-performance, which is doubling every 18 months. Kurzweil looks into the future and extrapolates that $1000 will buy you as much processing power as a human brain by the mid 2020s and more than all human brains in existence by the 2030s. More intriguingly, he also looks into the past and postulates an identical exponential progress, starting with the evolution of DNA and continuing on up through homo sapiens, agriculture, the wheel, and investment banking. This observation, that "milestones" in "progress" seem to be happening more frequently as time goes by, is the basis of his Law of Accelerating Returns, which is the cornerstone of his belief system. Basically, Kurzweil believes that all evolutionary systems, of which technology is one, increase their "order" exponentially. Therefore, it's no coincidence that it took 2 billion years for DNA to evolve, but then only 1 billion years for eukaryotes, then only 500 million years for multicellular organisms, and so on, each milestone occurring in half as much time as it took to achieve the one before it. For reference, see:

Canoncial milestones in the Singularity

The foundation for his Law of Accelerating Returns is by far the weakest part of his argument for what the future will be like. He asserts that evolutionary systems inevitably increase their "order" exponentially, but never approaches a rigorous definition of the term. This hand-waving approach to his fundamental theory really comes to a head when he talks about the Fermi Paradox, the observation that intelligent life in the universe seems to be uncommon. This flies in the face of Kurzweil's thesis, which is that intelligence inevitably evolves from DNA up through brains through supercomputers, until eventually a civilization is converting all the matter and energy in the universe into one giant computational device at an exponential rate. If such a project is underway on behalf of some other intelligent race, they're being awfully quiet about it. Therefore, a futurist is forced to either 1) apply a universal anthropic principle: we must be the first intelligent (technology-creating) race in the universe; or 2) accept that there's nothing inevitable about the evolution of intelligence or technology. Kurzweil chooses option 1), and while he makes an eloquent case for his choice it's far from airtight. Beetles have been evolving exactly as long as we have -- they're more successful than us by any evolutionary metric -- and yet they possess no intelligence.

Kurzweil is at his best when he sticks to the predictions that his objective measures of price-performance of information technologies actually support, and these predictions really are awe-inspiring. We can say with great certainty that the years of the 2020s will in fact give rise to truly earth-moving, paradigm-shattering technologies as a result of cheap, ubiquitous computers. Kurzweil has a lot of specific examples, mostly concerning the augmentation of human intelligence and virtual reality interfaces for the brain, that are best read in the book itself.

It's worth noting that Kurzweil doesn't seem to think that peak oil will affect his predictions in the least. James Kunstler's The Long Emergency Surviving the End of Oil, Climate Change, and Other Converging Catastrophes of the Twenty-First Century talks about the diminishing returns of technology, while Kurzweil posits the Law of Accelerating Returns. Clearly, one of these men is a pessimist and one an optimist. The question that concerns all of us is: who is right? I view the singularity and peak oil as a race, with the future of human civilization at stake. Can the exponential increase in technological progress replace our fossil fuel energy before Saudi production peaks? No one knows; Kurzweil is pretty bad at predicting specific applications of technology (he thought we would have glasses that use lasers to paint images onto our retinas by 2009), but his data on the exponential increase in price-performance in information technologies is so predictable as to be considered infallible. There's enough solar energy hitting this planet to replace our fossil fuel use and then some. Kunstler thinks we'll be unable to harness it as a civilization because of the how very little time we have until oil production peaks; but then, he clearly doesn't understand the exponential nature of progress, so his predictions are suspect. Kurzweil, on the other hand, is a self-admitted utopian optimist who sincerely believes that humankind's ascension to the stars is not just likely, but destined, preordained.

So to reiterate: should you be investing in canned food or in nanotechnology companies (and virtual pornography firms, of course)? I can't tell you how to live your life, but my personal perspective on this dilemma is as follows: in 20 years we will either be living in the middle of the fantastic future promised by the Singularity; or we'll be back to an agrarian lifestyle and focused mainly on staying alive and killing our neighbors. Since the latter outcome, the total and near-irrevocable destruction of our way of life, is unthinkable, I prefer to believe in the former. In other words, I choose not to invest in canned food, because I can't imagine living in a world where I'll need it.

I can't say I recommend this book very strongly, except for those who already have a strong interest in the Singularity. The subject matter is often fascinating and thought-provoking, but the presentation leaves much to be desired. Kurzweil rambles and repeats himself, and multiple times lets his narrative devolve into pages upon pages of bulleted lists. I'm sorry, Mr. Kurzweil, but I could have gotten that on Wikipedia.