DARPA does multiple levels of top-down problem generation and bottom-up solution generation

In some ways the the Program manager is the ‘bottom up’ half of Top down problems and bottom up solutions. When DARPA hires someone, they are usually looking for someone to go after a vague area or problem rather than just hiring someone to fill a slot or because they’re great. So the DARPA director or previous program manager do a first pass of ‘top down.’
The Program manager then synthesizes a program by digging into the current state of the discipline, what people on the bleeding think might be possible, and integrating that with an idea of where they see the world going. From the directors point of view this whole process is very bottom up.

In other ways the program manager is the ‘top down’ half of Top down problems and bottom up solutions. When they are doing initial program design (The DARPA execution framework boils down to showing that thing is not impossible, showing that thing is possible, and then making that thing possible) they talk to tons of researchers about possible directions and solutions. A large part of a DARPA program manager’s job is focused network building. During the actual execution phase, while PMs have a plan and an idea of who they want to deliver on it, they still put out a broad call that anybody can respond to at least nominally keep the door open to other bottom up solutions.

In order to be a nexus of top-down and bottom-up approaches DARPA PMs need to think for themselves, be curious, and have low ego. If they don’t think for themselves, they will be at the whim of either the perceived desires of the director or the scientific community. If they’re not curious, they won’t actually synthesize anything interesting (The best DARPA program managers are the ones who can look at an entire literature in an area and notice a systemic bias) and will leave gems sitting by the roadside. Finally, if they don’t have low ego, they will build a program purely from their own ideas and cut off the bottom-up value of the system.



Web URL for this note

Comment on this note