The Right Coast
December 13, 2005
Quite possibly the last grumpiness on the Singularity in 2005
By Tom Smith
Over at instapundit, Ray Kurzweil's book "The Singularity is Near" is rated the most important (by far) non-fiction book of 2005. This gets my grump up, mainly because the book has no business in the non-fiction category. Where it belongs is fantasy.
I read the summary of the book's arguments in New Scientist. I am not going to read or buy the whole book. K's argument seems to boil down to this: we are going to experience not only growth, but exponential growth in various key technologies for the next few decades, and so by 2040 or so we will be immortal, have computers that are smarter than we are, and so on and on. But this is just a fallacy. For two reasons. First, with a few exceptions, we don't know how to measure technological growth. We can measure the processing speed of computers, but beyond that, K's choices of things to measure struck me as opportunistic and sometimes bizarre. I should give examples here, but really, just buy the book and see for yourself, and pay particular attention to what is on the y axis, where the x is time. It's not quite "units of coolness," but that was the impression I took away. Economists can't even measure with any accuracy the rate of return on capital invested in R&D, which makes R&D investment a very entrepreneurial activity, lots of Knightian uncertainty, if you know what I mean. So it is far from clear what it is we are projecting will grow exponentially in fields such as biotechnology.
Here's a scientist who says we are entering a dark age of innovation, that innovation is slowing. His evidence strikes me as much more concrete than K's, though also very problematic. Does number of patents really measure anything?
But even if we did know what to measure, and found it had been increasing exponentially for the last t years, that gives us no ground whatever for projecting that trend indefinitely into the future. It might, or might not. We don't even know what inflation is going to be next year, let alone the psychohistory (isn't that what Asimov called it in the Foundation trilogy?) of economic and technological development. Yes, I know, Moore's law says chip speed will double every 18 months or whatever, and at that rate by 2040 or whatever, a computer the size of a breadbox will be smarter than everybody in the world put together, or something. But that is at least a good a reason for thinking sometime before 2040 some barrier will be encountered to make such a computer impossible, as it is for thinking we will have such a computer. Back in the 60s as a lad, I remember reading projections of increases in the speed of transport from the year 0, to the year 1965. You had only horses for the longest time, then trains, then cars, planes, jets, space rockets, you get the idea. An exponential (roughly -- just fiddle with the axes) increase. By 2001 or so we would be zipping around the stars at the speed of light or much greater. But alas, now the curve has bent over. Fuel is heavy, metal is only so strong, and people have better things to do than see how fast they can go.
I'm afraid for Ray and the rest of us, we will just have to deal as best we can that in more like 20 or 40 years than 400, we are going to be into whatever afterlife may exist, or barring that, existing only in the memories of those who remember us, and I don't mean consciousness creating computers here, either. Ray apparently takes many dietary supplements because he does not "plan on dying." It's a free country, but from seeing him on TV I would say weight watchers is probably a better bet for him (and for many of us) than little nano-bots that will scrub out his arteries for him. Bran is cheaper, and available at your local grocery store. And don't forget the red wine.
I suppose everybody needs a dream and living forever is a popular one, though most people think the medium will be spiritual not mineral. By all means, if big pharma comes up with some little white pill I can take that will allow me to remember my colleagues' names and make me feel frisky at 90, I'm all for that. In fact, bring it on. And I suppose if I could retire and spend 40 years travelling around, seeing my great grandchildren, actually learning how to fly fish, well, sure, why not. But live for 200, 300 years? Not that anyone will have to, but spare me of that, please. Oblivion beats boredom any day, and if there is some kind of spiritual existence, it seems likely to be better than whatever they will dream up at MIT or Microsoft. The aesthetics of Windows are bad enough. Am I to believe they would do better with eternity?