• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« 3D Atom-level Analysis Tool | Main | Global Goals for Year 2050 »

September 08, 2005


Feed You can follow this conversation by subscribing to the comment feed for this post.

michael vassar

Of course, that same "law of accelerating returns" implies as much change between 1985 and today as between 1885 and 1985. I know of no reasonable metric for which this is true, not even Absolute (not log) economic growth. MNT might make things change that fast, but don't please support sound arguments with nonsense.

Mike Treder, CRN

Actually, no. The rate of acceleration between 2005 and 2025 will be *much greater* than between 1985 and 2005. You can't go back in history and say that each 20 year period should have as much progress as the previous 100 years. That's not how the math works.

If Ray is right, we are now roughly at the "knee of the curve", the point when rapid acceleration should start to become apparent.

michael vassar

Last I checked, Ray wasn't claiming hyperbolic curves, just double exponential, e.g. exponential with a slow exponential change in the base. What math do you think he is claiming? What math could he possibly claiming which fits the past 30 years, the 30 years before that, and his idea of the next 30 years into a single, simple, smooth function? At least numerology uses numbers.

Tom Craver

Calling progress exponential is kind of problematic, in that there really is no reasonable metric for progress. E.g. computing power doubles every 18 months - but what if each doubling only enables one "significant" new computational task? Is progress exponential? Or linear?

I do expect we're in for some major life-altering changes over the next 20 years. We're facing some tough issues, some of which are "adapt or die" - and technology is how humans adapt.

One issue that isn't commonly mentioned: China has a new generation coming up with 13% (17 million!) more boys than girls. I would expect those males to be getting rather bitter in 5-10 years.

Rather than let that anger focus on itself, the Chinese government may decide to put those excess males to use in a major war of expansion, especially if it is hurting for resources. Kazakhstan would be one potential early target.

Janessa Ravenwood

Tom: If it wasn't for that possibility (use 'em for war), I'd be laughing myself to death over China's predicament. Truly something they brought on themselves. We girls being just so valueless compared to all those boys. Unless they're all gay, they're certainly going to be frustrated. Same thing's happening in India, and again I have no sympathy at all. Idiots, they deserve what they get. The kids, unfortunately, get stuck for idiocy of their elders. Well, hopefully they'll take it out on them and maybe, just maybe, girls won't be quite so devalued next generation.

Tom Craver

Janessa - in my paranoid moments, I wonder if China planned it this way - certainly they've had years to try to crack down and enforce the 1 child rule more stringently, or reward mothers for getting sterilized after having a daughter.

Chris Phoenix, CRN

Janessa, Tom, don't forget that one of the hepatitis viruses causes males to be born preferentially, and is especially prevalent in areas of China with large oversupply of men. So it's not certain that China is responsible for the whole 13%.

Michael, the math Kurzweil is using is:
"The paradigm shift rate (i.e., the overall rate of technical progress) is currently doubling (approximately) every decade; that is, paradigm shift times are halving every decade (and the rate of acceleration is itself growing exponentially). So, the technological progress in the twenty-first century will be equivalent to what would require (in the linear view) on the order of 200 centuries."

He doesn't give enough numbers to know the rate in 1985, much less 1885. But with so many variables in Kurzweil's formulas that we don't know the value of, your initial numerical claim isn't supportable by his statement. By the same token, we can't check his work.

Intuitively, I'm a bit suspicious of his figures. But there are a couple of arguments... First, my impression is that back in 1985, people still used 10- or 20-year old scientific papers; today, anything older than 3 or 5 years is out of date in a lot of fields.

Second, it wouldn't surprise me much to see a rapid and self-reinforcing takeoff of computer-aided or even computer-managed design in the next couple of decades. This doesn't prove that Kurzweil's methodology is right; he might for all we know have massaged his numbers to make them come out to match a destination chosen by another methodology. But it does tell me that we can't dismiss his results on the grounds that they predict too much change in the future.


The comments to this entry are closed.