random memes }

Welcome to Jim Taggart's cocktail party

In the book "Atlas Shrugged" Ayn Rand wrote a straightforward story -- about the creators of value withdrawing from society -- laden with a lot of her personal philosophy. From the book there is one metaphor I find use for rather a lot lately when framing my opinion of a group or organization:

Would this person fit in at Jim Taggart's cocktail party?

Sorry if the allusion in obscure. Go read the book. I will wait ...

Recently Tim Bray was going on (in ongoing) about "Missile Defense Is A Cult" and how the program just isn't going to work. My opinion is a little different. Intercepting a ballistic missile in flight is an extremely hard problem -- but I would not go so far as call it impossible. When Reagan announced the SDI program in the 1980's my thought was that a lot of hard research was needed. My first concern was that someone would spend too much money deploying a technology before it was ready.

Well, guess what ... twenty years later we have a "Missile Defense" program that is clearly doomed to fail. How can we tell? Simple - whenever you are faced with a extremely hard problem, the best approach is to divide the problem into a series of smaller problems, and work on the smaller problems first. Also you know it is likely that many of your initial solutions will not work, so you plan on trying lots of solutions cheaply.

So how does this translate into solving the difficult problem of intercepting a ballistic missile? Simple - start with the smaller problem of intercepting a ballistic object at short distances. First design, build and test (and test & test & ...) a targeting system that can acquire and intercept a ballistic object thrown from 100 yards away. This stage is pretty cheap, so you can try a lot of different variations to see what works best. When you get good at intercepting at that distance, then you increase to 500 yards, then a mile, then 10 miles, 100, 1000 ... in each stage only moving on to the greater distances (and more expensive tests) when your solution(s) work well at the current distance.

Maybe at some stage we would find that current technology just cannot be stretched to meet the end goal (though I doubt this). Likely there would be useful spin-offs (imagine a system that could intercept incoming mortar shells), though we cannot predict exactly how things would work out.

So instead our current "Missile Defense" program starts with $100 million dollar missile tests -- and fails a lot -- to my absolute complete lack of surprise. Since they skipped the smaller problems we can pretty much assume the folks running the program do not know how to solve hard problems. Since this is indeed a hard problem, we can expect the program to fail (and waste huge amounts of money in the process).

I am thinking the folks in charge of this program would fit in just fine at Jim Taggart's cocktail parties.

Even better, the guy currently in charge of the missile defense program is the Bush administration's candidate for the new head of NASA. Ouch. Is NASA a write-off?