The thing we’re going to have to show is that we can deconstruct a major effort like this — not run it like a major defense acquisition program — have everything come together seemlessly, but there’s no program manager for the internet and somehow it comes together seemlessly. It simply does it through adherence to standards. We’ve talked about doing that for years in the Air Force, but have never really pulled the trigger.
That was Air Force acquisition executive Will Roper at his Ask Me Anything session. He goes on to say, “I want 10-15% solutions that allow us to learn faster. ABMS [Advanced Battle Management System] won’t deliver, it will emerge.” That really is a tremendous statement from Dr. Roper. He is really grappling with what it means to be a program of record in the Pentagon.
Rather than each program creating a new platform with tightly coupled components, Roper is looking to achieve something more like an open system. This often looks like it is a disadvantage because it often implies lower performance. For example, most companies ran proprietary network protocols on the internet like AppleTalk and SNA in the early 90s. Everyone thought we would leap to the information superhighway, broadband. But the open protocol TCP/IP won out by being able to fully exploit the creative developments of the entire community. Open systems grow and drive value because one company or platform isn’t trying to dictate the value-stream too much.
There’s no PM for the internet today, but there certainly was when it was being invented. Larry Roberts (https://www.livinginternet.com/i/ii_roberts.htm) is legendary.
Hey David, good point. While the internet had a DARPA PM, which created the foundations for a good system, I tend to believe Peter Diamandis when he said that the internet would have continued being used by only a small group in govt/academia if not for private people building browsers and other protocols that opened the internet up to the public. Requirements for things like cookies didn’t come from DAPRA, and yet we have them standard. So I guess the internet could be another instance of the transition/scaling problem out of the labs. Often, it seems commercial industry picks up on the S&T and then scales it into a consumer product, only for it to boomerang back later to gov’t — like the internet, AI, robotics. It seems that where S&T actually prototypes at scale itself, prospects for transition into defense programs are better — like stealth and UAVs. All this is speculation though.
Oh, I agree completely. But key standards decisions (like TCP/IP) were made by the government PM explicitly with an eye toward future scalability — and we all benefitted.
Oh, I agree completely. But key standards decisions (like TCP/IP) were made by the government PM explicitly with an eye toward future scalability — and we all benefitted.