DARPA, the American government’s military research agency, was set up by President Eisenhower in 1959 to explore the frontiers of technology and science.
The idea was to bring together boffins and bureaucrats to hothouse crazy ideas that might just work – and it turned out quite a few of them did.
In fact, it’s been hugely successful from an innovation standpoint.
Quite apart from military applications, DARPA has played a major part in America’s current preeminence in technology.
The Internet, the modern computer desktop, GPS, weather satellites – all can trace their lineage to Eisenhower’s creation.
In recent years DARPA has been cited by economists and politicians who argue that innovation can’t be left to the market, but instead requires government investment.
Others showcase it as a model for organised creativity that gets astounding breakthrough results.
I came across DARPA when researching cognitive bias training in the workplace.
But I had very little idea of how DARPA actually works until I read this account of it, by Ben Reinhardt.
Two interesting things.
First, it’s tiny!
American government agencies tend to be massive, sprawling entities with layers and layers of management.
Yet DARPA has only 124 staff and three layers.
Everyone knows everyone else, which enables strong connectivity, disciplinary cross-fertilization, and serendipitous discoveries over coffee.
At one point, DARPA’s Director successfully lobbied Congress to shrink its budget because he thought the agency was getting too big.
It’s hard to imagine any other agency doing that yet governments and large companies need to find a way to empower small teams if they want to generate more creative problem-solving.
Second, DARPA is run by generalists.
It does not actually do its own research – it’s more like a venture capital firm which looks for intellectual projects to invest in.
The agency is run by Program Managers (PMs): highly curious critical thinkers who are skilled at communicating and at connecting people and ideas.
They have to be able to run teams of diverse experts and at the same time have the intellectual confidence to be skeptical of domain expertise, since all academic fields have systemic biases and blind spots.
That’s why all of DARPA’s staff have received extensive training in ways to mitigate the troubling effects of cognitive biases.
Dominic Cummings is fascinated by DARPA and his government is seeking to create a British version; it was mentioned in Michael Gove’s recent speech.
Cummings and Gove put great emphasis on getting specialists into the civil service, but DARPA shows us that innovation is often driven by generalists.
The work of government, in particular, is not primarily analytical: often, both problem and solution have already been identified correctly, but have met political and institutional resistance.
The hard part is delivery, and that’s usually a people problem.
Technical skills are crucial of course, but in most departments of government somebody like Louise Casey is 10x more valuable than a PhD in astrophysics.
Do you have innovative generalists in your practice or on your project teams?
Related: For more on the importance of generalists I recommend David Epstein’s truly excellent book, Range.)
Get the Builtworks Letter
In every edition of the Builtworks Letter, you’ll get the behind-the-scenes backstory as to how buildings are designed, built, and brought to life.
You’ll hear compelling stories, learn surprising ideas, meet engaging characters, and discover unique voices.