Should DoD spend half as much money twice as fast?

Here are a couple slices from a short interview with Palantir’s COO Shyam Sankar on Building the Base podcast with Hondo Geurts and Lauren Bedula.

If I was to sum up my policy advise, it would be to spend half as much money twice as quickly.

I think that’s an interesting phrasing, which at first blush sounds ridiculous but there might be something there. PPBE Reform is about doing $800 billion worth of work with half the money twice as fast. I think that is possible, but very hard to do within a long-standing institution because it means finding and empowering the most capable individuals to do what it takes.

For example, could IBM have reformed itself to create a viable personal computer that was better and cheaper than what Compaq created? It had everything it needed, but the culture was just not right. Twitter is in the midst of a huge experiment, where a large fraction of the employees were fired to control the burn rate after an exodus of advertisers. Yet so far there is no degradation of performance and perhaps even an acceleration of feature development.

The question is whether meaningful DoD reform can be accomplished without such divisive disruption, or whether existing institutions simply have to be side-stepped.

Here’s more from Sankar on how unique the United States tech sector is:

Silicon Valley is a child of the Department [of Defense]. Now the child has become so prosperous — look at Apple, Microsoft, Amazon — American tech is dominant. It is so dominant that sometimes don’t realize that it is American tech, we just think of it as tech. With all due respect to my Israeli friends, Germans, there is no tech scene like American tech scene. The second best is not even on the same chart. So we really have to think about that as a national asset and how we integrate.

One of the things that makes this really hard is that the Department is a technology exporter for some truly exquisite things. So you have to acknowledge the asymmetry. For hypersonics, for a whole class of exquisite engineering, is coming from the Department and its industrial base. For a whole different class of things that tends to be software, all that R&D and scale of innovation is coming from the commercial world. How do we get these two things to match?