In "The Age of Spiritual Machines," Ray Kurzweil recaps the history of computation, outlines the most important trends seen in our present computer technology, then boldly extrapolates forward to imagine a future where computers achieve human-level intelligence. It's a simple matter to show that by 2029, $1,000 (in 1999 dollars) will buy the computational power of about one thousand human brains. Once we have that kind of raw computation at work, other attributes of intelligence are sure to emerge. What would that mean?, he asks.
It's an intriguing premise, and indeed, this book has gained a certain mythical stature among cybernetics researchers and others who think deeply about the blurring of the human-computer interface. Although I consider myself one of these people, I was rather disappointed in the book and give it a solid 0, with one qualification.
The qualification is that this book rates more like a weak + if you are new to the whole concept of cyborgism. If you have never thought about:
However, that quality is exactly what makes this book so annoying for one who has kept up with the recent literature and philosophy. Kurzweil's book would have been brilliant and prescient in 1979, and exciting in 1989, but in 1999, it is stale stale stale. It sounded just like the late-night dorm-room chats I was having over a decade ago, except it includes updated technology like the Internet and wireless broadband. I have evolved beyond dorm-room chats to see a larger picture of computer development within societal (and indeed, global) development. I found that "The Age of Spiritual Machines" felt..... immature.
I'll get into my specific objections below-- let me first mention what I enjoyed about the book, for there were several good things. First, the structure. We're all used to reading things online now, with hyperlinks. Books can seem ploddingly linear compared to the richness of the Internet's information. Kurzweil has attempted to write his book like an online document, not with actual hyperlinks, of course, but with a structure that allows chapters to be stand-alone. In the Introduction, he encourages readers to skip around, reading interesting sections in whatever order appeals. Although the book does build from start to finish, he has constructed it so that a reader arriving from any other point will still get something out of it. It's hard to describe, but it ends up working well. Kudos to Kurzweil for looking beyond a serial presentation.
Second, I'll give Kurzweil credit for excellent depth and breadth of knowledge in the area of computer technology. He's not just a programmer or a silicon jockey, he's a technological historian and a better philosopher than many self-described geeks. "The Age of Spiritual Machines" includes a detailed timeline of the evolution of computation on Earth, actual software algorithms explained in plain English, as well as creative conversations between a human and a machine intelligence. Plus plenty of easily readable text. As noted above, if you are new to the topic, this book is a good one. I highly recommend it for junior high or high school students who are just getting enamored with computers and how we humans interact with them.
But frankly, most of the ideas are not that new. Carl Sagan wrote beautifully about the human mind within the physical universe over twenty years ago. Sherry Turkle has written much about the sociology and psychology of computer-using humans. Marvin Minsky, Alan Turing, and other pioneers of cybernetics have extensively explored the possiblities and implications of machine intelligence since the 1960s. Eric Drexler helped create much of our modern vision of nanotechnology. Kurzweil quotes all of these people; I am not accusing him of using their ideas without reference. His synthesis of the history and proposed futures of computation is fine. But we must ask, what did Kurzweil add to make "The Age of Spiritual Machines" his own?
The answer seems to be his specific scenarios for life in 2009, 2019, 2029, and 2099. Although many have proposed that computers will someday match and exceed human intelligence, Kurzweil quantifies these predictions and dares to write detailed snapshots of the future state of computing technology and human society far richer than those that others have ventured forth. In the first part of the book-- the synthesis of past and present thought-- he imagines "scanning" the 1999 human reader's brain into a PC. The 1999 Kurzweil holds regular conversations with this entity. In the scenarios of the second half of the book, Kurzweil lets this entity "evolve" through the changes that occur in humans through 2099. In the ensuing conversations, it acts as tour guide and historical interpreter.
This is actually the part of the book that I enjoyed the least, which is why I ended up rating the book poorly. The very part that Kurzweil himself came up with was, to me, very disappointing. I am trying to frame my criticism carefully because I don't want to sound like I am objecting to the fact that I don't believe his scenarios.
In fact I don't believe his scenarios, but that's not the point. Anyone who boldly writes what specific things will happen by a certain time is going to be wrong in some way. That's not my criticism of Kurzweil's visions. In fact, if anything is correct, it is some of his details-- the fact that computers will perform more computations per second than the sum total of human brains within a few decades, the fact that humans and their machines are becoming more like each other (that's been going on forever), and various other particular items. All of that is fine.
My problem concerns the assumptions behind Kurzweil's visions. He has made a common, and completely fatal, error: he assumes that his worldview is the one that will propagate into the future. "The Age of Spiritual Machines" is the story of what they world would be like over the next century if the 1999 human population consisted of 6 billion people like those who live at the MIT Media Lab, write software for Silicon Valley startups, and build high-tech gadgets for IBM.
There is nothing wrong with these people, and indeed, much of what they work on is affecting the rest of the world whether it wants to be affected or not. They should not be dismissed as agents of change. But neither should we forget about 1 billion Chinese citizens who may have different ideas about how to use the Web; the rise of religious fundamentalism; the fact that over half the population has never made a phone call; the stubborn pervasiveness of racism, classism, and sexism; and so many other political, social, and religious caveats.
Self-described geeks like to treat these matters as unnecessary complications that will get swept away in the utopian future of bots, virtual reality, personal-assistant software, and an economy where physical goods just take care of themselves so humans don't have to bother with laundry, cooking, housekeeping, and other tasks that self-described geeks aren't very good at.
Kurzweil doesn't quite give in to the traditional vision of computer fanatics, but he comes close. He is at least aware of larger societal structures like government and business, as well as intellectual concerns like philosophy and law. All of these are addressed in his series of 21st-century scenarios, but only in the most cursory way. And religion, race relations, sexism, and geopolitics are almost completely ignored. I wasn't expecting a full treatise on political and social ideology from a book about the advancement of computers, but nonetheless, I was disappointed with Kurzweil's limited grasp of how these things interrelate with technology and how they might evolve in the coming decades. This was the part that seemed especially immature.
It wouldn't be fair to try to summarize Kurzweil's scenarios in a few short paragraphs. What I will say is that nearly every page I found myself shaking my head at the gaping holes in the story.
Kurzweil's book is an excellent summary of where we are and where we can potentially go with our computational technology. The facts in it should not be ignored by anyone who is paying attention to the rapid pace of change, and who wants to think deeply and seriously about what these developments mean for humans and human culture. But don't rely on Kurzweil himself for the interpretation of his facts. For that, you must leave the geek world and see what real people are doing with computers around the globe.
Copyright © Kim Allen 2000
