Volume 1, Issue 3 
3rd Quarter, 2006


How We Can Manage Our Way Through the Intertwined Promise and Peril of Accelerating Change

Ray Kurzweil

page 5 of 15

Image 8 shows the smooth curve of average transistor price.

Kurzweil Image
Image 8: Average Transister Price

You could buy one transistor for a dollar in 1968. When I was a high school student in New York, I would hang out at the surplus electronics shops on Canal Street, which are still there, and buy a telephone relay with support circuitry for $40. This was a million times slower than a transistor, but otherwise equivalent to one transistor. You could buy 10 million transistors in 2002; it is now about 50 million transistors for a dollar. It is a very smooth curve. 

This is not the output of some table top experiment. This is the result of measuring worldwide activity that is very chaotic. There have been bankruptcies, IPO's, accusations of one country dumping products in another, and wars and we nonetheless have this very smooth progression. Unlike Gertrude Stein's roses, it is not the case that a transistor is a transistor. As we have made them cheaper, they have actually gotten better because they are smaller. The electrons have less distance to travel, so we have exponential growth in the speed. The cost of a transistor cycle has been coming down by half every 1.1 year. If you add in other levels of innovation, it is about one year now to double the price-performance of electronics. That is 50% deflation.

This actually affects every aspect of information technology, including biology. It took us 15 years to sequence HIV; we sequenced SARS in 31 days. The economists say that that is a danger. They have been worrying as much about deflation recently as inflation. We had deflation during the Depression - a completely different phenomenon, that was the collapse of confidence, collapse of the money supply. This deflation is due to improved productivity, but the economists say that's all very good, but if you can get the same capability for half the money a year later, you may increase your purchasing somewhat, but you are not going to keep up with doubling consumption every year.  You are not going to buy twice as much every year. You will therefore have a contraction of the economy at least as measured in dollars, which would be a bad thing. 

Actually, what we find is that we more than keep up with it.  There has been 18% growth for the last 50 years, 18% per year, compounded, in electronics and information technologies in general in dollars, despite the fact that you can get twice as much per dollar each year. The reason is as new capabilities come to be cost-effective, it opens up whole new applications. People did not buy iPods for $10,000 five years ago. Again, we have very smooth exponential growth; information technology is already 8% of the economy. It will be the majority of the economy by the 2020s. 

It is very pervasive. Image 9 shows Magnetic Data Storage.

Kurzweil Image
Image 9: Magnetic Data Storage

This is not Moore's Law; this is not shrinking transistors on an integrated circuit. It i’s shrinking magnetic spots on a substrate: a different technological problem, with different engineers, different companies, different countries, yet the same exponential progression. 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Next Page>