Well, one way to examine the issue would be through systems complexity. The approach that is now out of fashion was cybernetics. Now sociophysics seems to be ga

Well, one way to examine the issue would be through systems complexity. The approach that is now out of fashion was cybernetics. Now sociophysics seems to be gaining ground. In a short and vulgar summary, any complex system or systems used by humans should contain examination that includes humans, the irrational, unpredictable meatbags that they are. Software is implemented on automated systems we understand, build, and can predict at some level of performance. Medicine is built with humans, to work on humans (which we can't predict nor understand fully, let alone build like we do computers and software), and on top of that, uses software (and lots more). Software does not exist in a vacuum, some recent examples of the unanticipated effects of systems complexity in software could include a) political radicalization of facebook and youtube users through algorithms designed for a completely different goal, maximizing screen-time b) the simplest case, citibank losing millions because of bad UI design (not taking the human user into the design). So medicine, from a systems standpoint, is not that structurally/conceptually different from software. It is researched, designed and practiced/used by humans, just like software. Some of the reasons medicine is more complex is that it is, at the moment, dealing with more living systems than software (medicine: attempting to influence complex performance on the biological level. Software: for end-users, interact/affect on the psychological, conscious level). At the moment, medicine is more complex than the software industry. As software continues eating the world, that will change, as nanobots, brain-imitating processors, etc, arrive. I am at the moment looking at a cluster-fuck of improving a tiny process for a regional medical complex, and the staggering complexity is outside of anything I had to deal with while working in finance and distributed computing, and those software systems were more complex than what got humans into space. We know the lessons from space and atomic projects - have a mandate, financial and political commitment, theoretical and societal backing. But getting matter to controlled fission or orbit sure appears easier than even understanding cancer, let alone eliminating it, as history so far has shown us.