January 26, 2005

Prediction Border Conditions

Thomas Barnett is getting bombarded with the more evidence crowd:


I'm still not convinced! Have you taken into account inflation and the Hubbert's Curve effect on crude oil prices in your calculations? Because if you could just get those numbers right, I'm pretty sure I'd be emotionally invested in your vision of the future! No, really! Just get me those two numbers!

I'll give highly qualified backing to the numbers people on this one. Past a certain point, numbers do change the validity of even grand strategy. Here's an example. Dr. Barnett has predicted that the PRC will be heading to the Middle East in 20 years. If launch costs drop from $10k-$40k per kg down to $100/kg, the PRC doesn't have to go to the ME because enough of the world will be beaming power down from orbital solar stations that they won't have to. Oil will have a new, very clean, energy competitor that is likely to be priced lower than most current producers. The huge new supply overhang will mean that oil will be available to the third world for as long as they need it as the rich will jump to hydrogen very fast.

So who could drop launch costs that much? These guys are predicting launch in 2018, 13 years from now, of a structure that could do exactly that, a space elevator.

The question is what are the numeric border conditions that would knock your predicted future into nonviability? How likely are we to get to those numeric conditions? In Dr. Barnett's case, I think we're talking about pretty wide variances and pretty unlikely outcomes. But then again, you never know.

Until 1991, space elevators were creatures of science fiction and engineers calculated that only mythical "unobtanium" could build the things. One scientist in Japan finds Single Walled Carbon NanoTubes (SWCNT) and we're off to the races with lots of previous eminently sensible predictions becoming very unrealistic in the blink of an eye. Disruptive, transformative innovation is becoming more common so predictions without border conditions are becoming less valuable, especially long term predictions.

Posted by TMLutas at January 26, 2005 09:31 PM