Research projects have a halting problem

# Research projects have a halting problem The Halting Problem is the question of whether a program will complete or run forever. Research is similar - it is shockingly hard to look at a research program in progress and know whether it is a giant money sink that will last forever or whether it will look like a giant money sink until it produces something incredible. This property of research is terrible from an efficiency standpoint - you can make a strong argument to continue funding things that look like losers. At the same time those things could actually be losers. Unlike computational halting problems, the research halting problem might become more tractable if you are willing to put time and effort into actually thinking about the work. My hunch is that this problem becomes less the case the more trust there is between the researcher and the manager and the more skilled the researcher. [[Research requires more trust than other disciplines]]. At the same time, from personal experience sometimes it is completely unclear how *your own* research is going to turn out, so there are obviously limits even when you are as close to the work as it is possible to be. Any apparently stochastic process will have some examples of great outcomes coming after a long time from something that looks terrible. However, the nature and ‘underlying’ distribution is important here. If you buy the definition of project from [[raoGUTSGrandUnified2018]] or [[Areas programs and projects are distinct but often confused]], most interesting research “projects” are not actually projects. Almost by definition, because of the [[Knightian Uncertainty]], “research” is not a project. You can do projects *as part of research* but the research itself is not a project. *This idea originally comes from* [[José Luis Rincon]] in [[riconFundPeopleNot2020]]. ### Related * [[You can create a program design discipline that enables better research and development]] * [[For a certain class of research, deliverables can be detrimental]]  [Web URL for this note](http://notes.benjaminreinhardt.com/Research+projects+have+a+halting+problem) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Research+projects+have+a+halting+problem)