The Age Of Spiritual Machines ~ Ray Kurzweil

Reading this fascinating book at the moment.

Kurzweil is famous for making the prediction (in the 70s) that computers would be able to beat the best human chess players by 1998. This occured in 1997, when IBM's Deep Blue beat Gary Kasparov. There is still some dispute about that game (Deep Blue apparently had a team of grandmasters "advising" it in real time), but Kurzweil's point was there would be an increasingly exponential growth in computing power from the 70s onwards, and that this event would signify merely a footnote in this inexorable growth.

In this book he makes further predictions about this growth. He makes startling claims at the outset - that by 2099, computing power will be at such a level that 99% of the intelligence in the world will be machine-based; that this progression is the next level of evolution; that this evolution is unstoppable - claims you'd be tempted to dismiss as speculation. But Kurzweil follows each claim up with thorough justification - he buttresses each claim with a helping of evolutionary biology, computer science, mathematics, philosophy - thought-provoking and all considered, hard to disagree with.

I was immediately dubious of the central claim - that machine intelligence will overtake human intelligence as the superior intelligence on Earth in the near future. I thought of creations like Deep Blue - incredible feats of programming - but not intelligent. I don't think you can program "intelligence" - that human attribute that we aren't intelligent enough ourselves to understand, let alone set out in code or algorithm.

Kurzweil comes from two angles - the first is an outline of possible paths that computing power will take to get to this superior "intelligence". He discusses the nature of intelligence, "neural net" computing, quantum computing. The concepts are amazing, but I'm still not convinced. As I understand it, the intelligence we possess arose through millions of years of blind evolution, with branches hither and thither, pruned constantly by chance - till we arrive here battered and bruised, intelligent, and to borrow Bonnie "Prince" Billy - "with a drive to live I won't let go". Can intelligence exist without the wiliness that the will to live bestows? Even if we program "intelligence", can we program the will to live that we possess in spades and that I would contend is the basis of intelligence? We are intelligent, in order to survive, to outwit nature, to impress the ladies. Where is the prerogative for machines? Are we to give them this prerogative? Do we (or can we) understand enough our own prerogative?

The second angle is, for me, much more convincing, if fantastical. This angle conjectures that over the next century we will become more dependent on ever-developing technology, and eventually subsume it.

This, for me, is an astonishing claim, and much more plausible than a programmed intelligence. This conjecture says that machine intelligence will emerge because we will become machines. This sounds nuts! But is it so far-fetched? Consider hearing aids - they have become the most normal apparatus - a little machine that sits on your ear and augments sound quality. These have developed over the decades and are now very sophisticated, having the ability to analyze patterns and detect and amplify sounds in certain frequency wavelengths (like human speech). I wouldn't claim that these machines are intelligent, but the intelligent human wearing it is. It is not inconceivable that more of these machines will be developed - like for instance the ability to see for the blind (these are actually at advanced prototypes - a camera is wired directly into the part of the brain that deals with sight giving rudimentary vision). So Kurzweil contends that these technologies will be developed first to correct disabilities - then, when their individual abilities exceed the human sense they are designed to mimic - healthy people will use them to augment their own (perfectly functioning) senses. This will become as normal as wearing clothes. Eventually we will be using machines to aid memory and critical thinking. Why not - a box like a hearing aid wired to the brain to expand your intelligence to previously unimaginable proportions? Who'd decline?

(A not uninteresting corollary to this thinking is that with parallel advances in medical technology - such as genetic engineering of organs - we could achieve immortality this way.)

I'd be much more inclined to believe that superior intelligence will emerge in this way. We use machines as tools, first externally, then internally. A world of cyborgs awaits.

A blind watchmaker spent hundreds of thousands of years honing the human mind - I don't believe it can be recreated in C++ any time soon. But to become machine, its not so far-fetched? Fascinating to consider in any case.

The strengths of the book are in the fascinating summaries of the frontiers of computer science. For the purposes of the argument, this is neccessarily related to questions of philosophy. Issues like this - while they seem at first questions of speculative science fiction - actually go straight to the meat of the deepest philosophical questions; what is intelligence? what is human? are they mutually exclusive? It reminds me of Phillip K Dick's unfortunately titled novel - "Do Andriods Dream of Electric Sheep?". A world where humans and machines are indistinguishable, a bounty hunter is tasked to track down and kill these imposters; but they feel pain and have thoughts and feelings - is it moral to kill them? Eventually the bounty hunter - in an existential crisis - begins to question whether he himself is human or machine? The big questions can be posed in exhilarating narrations and in popular form, make no mistake.

The weaknesses of the book are in the closing chapters where the author has imaginary conversations with an inhabitant of the distant future - ironically, they are just too far-fetched and in a narrative conceit that is just too false. Also, the book - published in the early 2000s - makes a number of claims about the year 2010 that are - as far as I know - well off; like a telephone that translates languages in real time.

Highly recommended - nourishing food for thought.

Mission Statement

So many times, in discussion with friends or wherever, I’ve found myself lost for words.

Not lost for words in the sense that I don’t know what to say on a subject. Lost in the sense that I either haven’t refined my opinions to myself and they remain nebulous in my head – or, I know exactly what I want to convey, but it can’t be expressed in a sound-bite. To express it would require me to speak in pages and paragraphs and punctuations – and this is not how people talk. (In fact, I wouldn’t want to talk to anyone who spoke in paragraphs.)

So I have resolved to start a blog where I expound on what I think about stuff.

Subjects will be eclectic – I am as much establishing what I think about things. Being as prone to error as anyone else – I’m always interested in other views. If I’m wrong, tell me why.

Hairycakes