# A private DARPA riff both can and must do multiple experiments at once [[Running multiple organizational experiments flies in the face of common wisdom]] (especially in the startup world) that you should only change one thing at a time. [^1] I agree that this is absolutely true in a business sense. However, businesses *mostly* work - you might be satisfied with how a company thinks about HR or design, but basic organizational structure and best business practices are pretty effective. People have tried millions of fairly open[^2] experiments with organizational practices over many decades in everything from HR to management to employee compensation. Research organizations have simultaneously had many fewer experiments (just compare the number of companies that have been started to the number of university labs that have been started), less selective pressure, looser feedback loops, More concretely, if you simultaneously believe [[It is worthwhile and possible to riff on ARPA]] and that you should only change one thing at a time you’re limited to creating new government organizations like [[ARPA-E]] and [[IARPA]]. If you buy the argument that it’s possible to riff on the ARPA model outside of government ([[Can the ARPA model exist outside of the government?]]) and that in fact [[21st century riffs on ARPA should be private]], that single shift forces you to rethink multiple things (legal structures and funding for a start) whether you like it or not. At that point, clinging to existing structures could become more of a liability than wisdom because the assumptions that made those structures good ideas in the first place no longer hold. [[A private ARPA requires new legal organization structures]]. So an ARPA riff will *need* to do some institutional experiments to survive at all. On top of that, [[There are a number of experiments that seem like they can only be done in the context of an organization]]. Since the goal of an ARPA-riff is not just to enable more <awesome sci-fi shit> but to try to demonstrate a new institutional model ([[PARPA should create a replicable institutional model]]) guinea pigging should be part of its role. [[A laboratory for experiments in 21st century research management]], if you will. This *isn’t* to advocate throwing everything away just because it smells like the past. [[Chesterton’s Fence]] is important! Instead, it’s important to learn from the past (and that is a large part of the research that I will present) but since you’re in a strongly disequilibrium state, you need to evaluate each component on its own merit and perhaps grab components from disparate parts of history or the organizational landscape. Pulling this off will be like sailing between the Scylla of extraneous experiment and the Charybdis of similar incentives leading to similar outcomes. ### Experiments * [[You can create a program design discipline that enables better research and development]] is a hypothesis that needs to be tested in an organizational context. It’s not absolutely necessary to riff on ARPA - from what I can tell, the actions DARPA program managers take when they’re setting up a program is not a super legible process. My hunch is that this is one of the [[Things about ARPA that can be improved]]. ### Possibly extraneous analogy! An analogy might be the difference between setting up a new colony in Massachusetts vs. on Mars. Environmentally, Massachusetts is pretty similar to England so it makes sense to - sure you might want to change some things to deal with Native Americans but you still want to use roughly the same tools, farm the same way, build houses the same way, even the extreme religious stuff the Pilgrims set up the colony to be able to exercise buffered back to an equilibrium state. Mars, on the other hand, is unlike *anywhere* on earth (except maybe Antarctica.) People will need to live there in very different ways than they do on earth. [^1]: I like to think of this like how polynomial approximations are valid in a local area around an equilibrium. [^2]: There are few things more fun to talk about than how messed up your old company was. [Web URL for this note](http://notes.benjaminreinhardt.com/A+private+DARPA+riff+both+can+and+must+do+multiple+experiments+at+once) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/A+private+DARPA+riff+both+can+and+must+do+multiple+experiments+at+once)