The Age Of Spiritual Machines [Paperback]

If yes, and you're wondering what you might be doing in 2015, read this 1999 classic - the one future book you really should read.

OK, you may hate 'SciFi', you may not buy Kurzweil's grand Laws, or his post-2030 stuff, and Kurzweil's style isn't to everyone's taste, but the surprising thing is that the IT industry seems to agree with his central conclusions for the next fifteen years or so. And that alone is sufficient for us to be living in a world where androids are cleverer than us. That'll change the way you work: most of us will have to respond to our competitors who might be using androids in a competitive situation. Scary stuff for me and most of us, I would think. But why should we take K's particular set of future visions seriously?

Well underlying the whole analysis is Moore's Law, which says that the amount of computing power (memory, processing power, communication speeds etc.) that you can buy for $1000 doubles every two years, and we all know that it has held roughly true for at least the last forty years (K argues for 100 years, but we don't have to agree with him to continue the logic). Current 2003 computers have the processing power of a reptile's brain, and the IT industry apparently agrees with Kurzweil that mainframe computers will equal a human's processing power some time around 2015, with this coming into laptops (or their equivalent) in the following 5-10 years. No serious chip manufacturer or IT player seems to dispute that Moore's Law will hold for another 15 years and that the underlying technologies to achieve this fundamental technological shift exist (3D chips etc).

And that's all that's needed: Moore's Law for another 15 years plus the kind of global economic conditions we now have, and intelligent androids will be here. And if computers are in some important senses equal to us in 2015 then you can bet your entire pension that they will be cleverer than us by 2020. How old will you be then, and what will you be doing? K's predictions don't have to be out by many years for us all to have to think hard about what we will do in a world where androids first evolve separately from us, and then, horrifyingly start to physically merge and evolve with us.

Scared? I was. But I'm coming round to the idea. Think: intelligent assets are already here: a 2003 car already has around 60 components with programmable logic in them; each is upgradeable so that when you take your car in for a service the software can be upgraded. Mobile phones are now so much more than a telephone. Can you think of a seriously good reason why 'intelligent' heart valves, knee joints, or contact lenses will NOT be invented by 2015? And after that?

What shocks me is that when I reflect I realise that the world economy's current competitive forces will compel these 'good' inventions to be created. Then we will have to think about how any ruling Saddam Husseins might develop such machines. I don't see the faintest signs of serious debate, but we're talking about the future of humanity here.

In sum, short of asteroids or human disasters wiping us off the face of the planet, I cannot see how we won't be living in a world populated with androids by at least 2020. Fundamentalist Presidents and Ayatollahs may spout until they're blue, but they won't be able to stop this trend unless they break up the current global economic system, and define, and strictly-enforce, some limits on what computers and Programmeable Logic Circuits are allowed to do in their own countries. And then they might delay things by ten years at most in their own country (think of the Soviet bloc's experiences trying to resist computers and photocopiers). I certainly didn't have the view that intelligent androids are inevitable until I read this book and started asking questions in the IT community about the first fifteen years of K's visions, when I found that it was pretty well taken as read and agreed. Read it yourself and see, five years out, how well on track we are for his 2009 predictions. Then think: what do I want to do in such a world? More importantly, as a race, what do we all want to do?

It's quite possible to nit-pick over much of what Kurzweil says - but that's not the point. The point is the general vision of where we are headed. Kurzweil's view is that there is a 50% plus chance that humanity will make it through this transitory phase (ie the next century), that we will successfully combat the comming threats of self replicating biotech pathogens, software pathogens and self replicating nanopathogens, to complete the process of integration with our technology - and abandonment of our biological roots that we are now in the early/final stages of. Early because we are currently only fractionally fused with our technology (language, books, machines etc). Final because the maybe 40,000 year process is, because of the exponential acceleration of technological development, perhaps only 50-100 years or so away from completion.

I guess this is likely to seem utterly far fetched to 99.9% of the public - just as would mobile phones, the internet and robotic jet travel have seemed beyond belief to a 1900 audience. My belief is that these guys are very much on the right track. Though I'm not sure I'd put our chances of making it through quite as high.

Of course, if they are, it throws up a myriad of philosophical and practical issues which have yet to be satisfactorily explored (at least to my limited knowledge).

To my mind the most interesting of these is what evolutionary process will drive forward the development of knowledge that Kurzweil in common with the others assumes almost as a given, once knowledge seeking is completely delinked from biological survival? Today only a tiny fraction of available (largely biological) computing power is engaged in the quest for knowledge - the majority being used to enhance the immediate(ish)survival and reproductive imperatives of our genes(or to address the psychological mechanisms which have evolved to enhance their reproductive success).

Of course a stream of knowledge seeking activity has evolved and has proved phenomenally successful in terms of impact on the world, if not instantiation. But the successful evolution of this knowledge base has been underwritten by its usefulness in the competition for resources by biological agents. Kurzweil doesn't really focus on how knowledge will evolve once this link is eroded except for the general proposition that good quality knowledge will always be helpful in a competition for scarce resouces. It would be interesting to see how his ideas fitted in with those exploring the memetic theme.

In any event, it would be kind of disappointing if the ultimate fate of intelligence on this planet was an infinately looped orgasmatron or game of Who wants to be a millionaire? But one only has to look at the mass content of the net to see that this is not a trivial concern.

Bottom line - If you are interested in how this century might pan out, this is as good a place to start as any.

However, his enthusiasm and accessibility of his writing tend to gloss over the fact that he uses a rather eccentric interpretation of physics and biology to justify why his futurology is based on solid science, rather than wild speculation.

After you begin to realise that the theories on which the book is founded are flawed, you begin to mistrust the rest of Kurzweil's ideas. For instance, the author suggests that evolution is a bad programmer because there are repeated or apparently extraneous strings of code in our DNA. Surely he must be aware of the AI experiment where the scientists successfully created a functioning neural net and took out the paths which were unused. The net failed to function without the 'unnecessary' elements. Surely, we simply don't know enough about our own DNA to make the assumption that portions are unnecessary?

Despite misgivings over the hard science, I like Kurzweil's optimism and the philosophical questions he posed about the nature of identity. So I can recommend this book, even though Bill Gates has endorsed it!