December 17, 2003

Variables and Constants I

One of the most persistent errors in human affairs when it comes to progress is assuming a variable is a constant. These days this error shows up all the time in virtually every field of human endeavor. So, why do we do this? Because for the vast majority of humanity's history it worked so well and for a lot of circumstances, it still works well most of the time. Constants don't have to be calculated so you save a tremendous amount of work in assuming slow moving variables are constants. Everybody ends up doing it, sometimes to disasterous effect but usually not.

Osama Bin Laden did this when he ordered the attacks on the WTC and assumed that the US would react as it always had before. People act as though the population is and always will be going up. Until recently, inflation was assumed to be a constant of the modern condition. The Democrats and Republicans will always alternate in ruling the US and the Liberals have an electoral lock in Canada.

It used to be that tools were considered a constant. Before the industrial revolution and interchangeable parts, a hammer wasn't much different from year to year and specialization of tools and tool development were slow to happen and slow to spread. Essentially, a whole lot of people could get away with long term planning by setting the quality of your manufacturing tools as a constant. That's no longer the case. We know now that they're variables because tools change frequently and do the same jobs faster, more reliably, and cheaper in a constantly shifting tripod of improvements (as the classic phrase goes, pick any two).

But in the shift from viewing something as a constant to viewing it as a variable, ther is a sticky period where you have something I'll call a shifting constant. When plow innovations happen every couple of hundred years or so, adjusting crop yields upwards, it's more efficient to view yields as constants and simply change the constant as a one off exercise if you happen to be alive when an innovation happens. No problem, you have the efficiency savings of not calculating the variable effects most of the time and adjust once as your forecasts go off because of this rarity called innovation.

In more and more fields of endeavor, we're going through that sticky period at the same time.

Posted by TMLutas at December 17, 2003 12:28 PM