on 3 October 2012
Don't be fooled by the cover of the book (which bills the author, Kevin Kelly, simply as cofounder of "Wired") into thinking that it's only about recent digital technology. Inside, there is a refreshingly long view of technology, starting with stone scrapers. I heartily recommend What Technology Wants because I think it's awesome and important. Awesome (in both senses of the word) because it demonstrates an impressive breadth and depth of knowledge and is very thought-provoking. Important because everyone uses technology and is affected by other people's use of it.
Kelly sets the scene by confessing to contradictions in his own relationship with technology. For example, he runs websites but has no TV or smartphone. But he points out that all of us accept some technologies while rejecting others. In order to know how to respond to technology, it's no good starting with specific questions, such as whether we should allow human cloning or cars that park themselves. Rather, we have to look at past trends in technology as a whole and then try to extrapolate from those trends to see where technology is going. He frames this process as figuring out what technology wants.
To ascertain past trends, Kelly examines biological evolution and technological development. The central thesis is that the development of technology mimics the evolution of genetic organisms in that both share several traits, including moving from the simple to the complex, from the general to the specific, and from energy waste to efficiency. Both are seen as information systems with major transitions in the level at which information is organised. Biological organisation started at the level of one replicating molecule and moved through various levels, including RNA, DNA, cells, asexual reproduction, sexual recombination, etc., until reaching language-based societies. Technology started with language and moved through printing, the scientific method, and mass production until reaching ubiquitous global communication. Language bridges the two sequences and unites them into one continuous sequence.
On the biological side, Kelly puts the case for the idea that evolution is not random drift. Rather, it has a direction and is more or less reproducible, and this is because evolution is constrained by the laws of geometry and physics. I found this part convincing (though I don't know whether this view is generally accepted by biologists). The evidence presented here includes the frequency of convergent evolution (the evolution of similar traits, based on the same proteins, in far flung branches of the tree of life) and the results of experiments on cloned bacteria. For me, this was the most fascinating part of the book because it gave me a whole new way of looking at evolution. I'd recommend What Technology Wants for this part alone.
On the technological side, there is again a case for an inevitable direction, again supported by lots of interesting evidence. The evidence here includes simultaneous inventions, Moore's Law and similar laws, and the use of technology by the Amish. Kelly points out that while many people say they disagree with the idea of technological determinism, they don't act that way; all inventors are in a hurry to patent their ideas or get them into distribution before someone else does. Again, I'd recommend What Technology Wants for this part alone.
Unfortunately, I think there's a missing link between the biological and technological. Every time an inevitable direction is perceived in biological evolution, it is asserted that there is an equivalent inevitability in the direction of technological development, due to the continuity between the two. However, the earlier claim that biological evolution and technological development constitute a continuous sequence relies on the assertion that they are bridged by language. But no proof of this is offered. Only two pages are devoted to the origin of language and the discussion is somewhat speculative. In a way, the speculative nature is not surprising. As a former student of linguistics, I know that the scientific study of language evolution only took off in the 1990s, is very difficult due to the paucity of data, and is highly controversial, with many competing theories. But Kelly does not acknowledge that his whole argument rests on one briefly stated perspective among many on the origin of language. Therefore, I treated the comparisons of biological evolution and technological development merely as a metaphor. A powerful metaphor, but no more than that. (That's why I say "technological development", whereas What Technology Wants says "technological evolution".)
Moving on, Kelly continues the trajectory into the future, predicting the general direction of technology in terms of facets such as complexity, specialisation, ubiquity, and structure. I don't agree with some of the predictions here, though it's possible that I just don't want to believe them. As a wise person once said, prediction is very difficult, especially about the future. So I can happily agree to disagree.
If there is an inevitable direction of technology, where does that leave human choice? Kelly claims that understanding the inevitable direction enables us to predict where technology as a whole is going, which enables us to steer the specifics of a given technology. I'm not convinced by this. His answer to any problem with a technology is always another technology. For example, he cites psychological research showing that the more choices people have, the less likely they are to make any choice because they become overwhelmed. Answer? Choice-assist technologies, such as search engines, recommendation systems, tagging, and social media. Yes, but how to choose which of the many choice-assist technologies to use? In any case, we have no choice but to use certain technologies. As Kelly himself points out, even dropouts and the Amish depend on others adopting technology. Nonetheless, What Technology Wants is still interesting up to this point.
However, I was very disappointed with the last chapter. The subject matter suddenly changes to theology and the style changes from meticulous to something veering on the mystic. This is bad form. A conclusion should never come as a surprise. It should bring the work full circle, showing how the objectives set out in the introduction have been met.
So, anyway, I said at the beginning of this review that I heartily recommend this book. That recommendation still stands despite the missing link in the argument and the disappointing last chapter because (1) the argument for a technological direction stands on its own merit, (2) the book works perfectly well without the last chapter, and (3) my overall impression is still that it's awesome and important.