He's written either a book or lengthy paper about this every few years since he wrote The Age of Intelligent Machines in the late 80s. At that time mainstream opinion was that he was nuts. The book was especially ridiculed for its predictions of a computer AI unseating a world chess champ by 1998 and of the human genome being mapped out by the early 2000s.
Both happened ahead of schedule, and that's the primary reason I took his later books seriously.
I'd be interested to see a chart of something trying to more directly measure "standard of living" alongside "information storage per unit volume, communications bandwidth, and computation speed". I also wonder if the relationship will change in the future. I guess standard of living depends on more than just technological change though, in particular government policy.
In http://en.wikipedia.org/wiki/The_Singularity_Is_Near, Ray Kurzweil also claims that growth is not only exponential (e.g. doubling every x years), but that this rate is itself increasing - which I guess is superexponential.
It's curious that Moore's Law (http://en.wikipedia.org/wiki/Moore%27s_law) seems to be just plain ol' exponential. That is, doubling the number of transistors on a chip every 2 years. That rate itself hasn't increased; it's still 2 years, 47 years on. (BTW the 18 month figure is related, see link).
Why isn't Moore's Law superexponential?
Quick guess, Moore's law applies for capacity of silicon industry to improve storage and processing of one production line.
If you include the increasing number of production lines, the total number of simultaneously connected computers, capacity of producing software that solves complex problems and the many others factors that are included in the global processing power and relevancy of calculus, you might explain this super exponential growth.
I wonder to what extent this is a case of a technology platform (integrated circuit density) improving at an exponential rate, but then further technologies built on top of that (e.g efficiency of computational algorithms) are are themselves improving exponentially. The exponential rates multiply to give a super-exponential rate in aggregate.
We're definitely still on the fast part of the curve, but I am one of those that thinks inevitably the rate of advancement in some and eventually all of these technologies will subside. There's nothing magical or special about IT. All technologies eventually mature, but the world will be a very different place than it was before the IT revolution by the time that happens. In fact, it already is.
Article: Gonzales says this research suggests an accelerating rate of improvement in technology... If these trends continue, he says, “in some ways things will continue to get better. In that sense it is a hopeful paper.”
Abstract: In contrast, one cannot reject the hypothesis of superexponential growth with decreasing doubling times. This raises questions about whether past trends in the improvement of information technology are sustainable.
These seem to be two different conclusions, don't they?
I wonder how the number of people working on progressing technology affects this - i.e., crudely, what (rate of technological progress) / (world population) is.
11 comments
[ 3.0 ms ] story [ 20.8 ms ] threadBoth happened ahead of schedule, and that's the primary reason I took his later books seriously.
It's curious that Moore's Law (http://en.wikipedia.org/wiki/Moore%27s_law) seems to be just plain ol' exponential. That is, doubling the number of transistors on a chip every 2 years. That rate itself hasn't increased; it's still 2 years, 47 years on. (BTW the 18 month figure is related, see link). Why isn't Moore's Law superexponential?
If you include the increasing number of production lines, the total number of simultaneously connected computers, capacity of producing software that solves complex problems and the many others factors that are included in the global processing power and relevancy of calculus, you might explain this super exponential growth.
We're definitely still on the fast part of the curve, but I am one of those that thinks inevitably the rate of advancement in some and eventually all of these technologies will subside. There's nothing magical or special about IT. All technologies eventually mature, but the world will be a very different place than it was before the IT revolution by the time that happens. In fact, it already is.
Abstract: In contrast, one cannot reject the hypothesis of superexponential growth with decreasing doubling times. This raises questions about whether past trends in the improvement of information technology are sustainable.
These seem to be two different conclusions, don't they?