Would you want to serve in the navy on a "Bob Hope" class ship?

The US military has traditionally bought its weapons from a number of enormous military contractors. Contracts involved huge amounts of money, and the contractor that got the contract often depended more on the influence of the senators in the particular states in which the contractors were based then on whether the contractor had the best bid, or whether it was likely to go over budget or not. Components were built to so called milspecs - military specifications that were not the same as the specifications of similar civilian products. The companies that arose to fulfil this market are not always all that nimble, although this varies from company to company.

This is not too bad when the number of components are relatively small, or the production runs are relatively large. However, it fits less well in the modern tech economy, in which the fixed costs of developing products (or components) are enormous, and the marginal costs of producing units are close to zero, and in which products are assembled from an enormous variety of off the shelf components. This article talks about the new generation of companies that are working as defence contactors to build modern high tech weapons. The companies mentioned in the article turn out to not be quite as small and nimble as you think they are when you start reading: they are perhaps only small and nimble compared to the defence contractors that came before (and some of the companies mentioned are in fact spinoffs of the companies that came before), but there is still a definite trend here.

As a caveat though, NASA was traditionally essentially a spinoff of the military, and it traditionally did business with the same defence contractors as the Pentagon. A few years ago, it decided to change its way of thinking: rather than inventing everything from scratch, its unmanned space probes were to adopt a philosophy of "Better, Faster, Cheaper", in which they were to be assembled from off the shelf components. This has had mixed success. Once or twice a mission designed this way has gone spectacularly right (Mars Pathfinder for instance) but in a lot of cases things have gone wrong. In order to make missions cheaper and get the most bang for the buck, one thing that has been done without has been proper testing. While there is nothing wrong with using off the shelf components in principle, NASA's institutional structure did not cope with this change very well. I suspect there are similar questions about the Pentagon's ability to cope with similar changes. That said, the Pentagon is far less besieged than NASA, and I doubt that such things as shirking on testing have gone on there, certainly not since September 11.