Uncertainty--when you haven't a clue, when it makes little sense to pretend you have information, when probability, markets, and strategy are unlikely to be of great use.
The End of (Economic) Growth--If Robert Gordon is correct that we have been living high on the hog from 1870-1970, plus a decade from 1994-2004, in effect floating on a rising tide, and that tide is now going out not likely to return in the next decades, what does that mean?
Intractable Problems, for which we can do something, but they return or are never worked out. Matters of race, of insufficient education and training, of deliberate attempts to game the system being justified, congestion and the time it absorbs. There are the recurrent issues of nuclear war and epidemic, of disaster and inadequate preparation and resilience, and of young people finding themselves leading lives that are criminal and dangerous. Alcohol and opiates and conventional addictive drugs.
The University of the Future--The notion of teaching is becoming a matter of didactic and instructional efficiency, with matters of modeling and mentoring, of thinking and criticism, falling by the wayside. The research enterprise is almost surely too big, and the incentives for research force people who would otherwise do something useful into a world of weak scholarship. Likely, about 20% of our research productive-scholars would be more than enough. Fischer Black said something to the effect that we ought be paid to teach, in effect to discourage research that is not driven by the need to find out.
"I see our university system as similar to the former Soviet empire, and as having similar problems . . . teaching and research are too uniform. They do not respond quickly to shifts in tastes and technology. . . . And, most important, teaching and research cost too much. . . . The basic problem is that we have too much research, and the wrong kind of research, because governments, firms, foundations, and generous alumni support it.[i]
Counting and Scaling: Often when we enumerate something, perhaps weighting each case by a function of that case's "energy" or significance, we end up with a sum that exhibits scaling--much as a sum the values of identical independent random variables produces a distribution, often a Gaussian, that builds in such scaling. What is going on?
I'm sure there's lots more, but this is just a beginning.