> What is the most important problem (of policy, not of specific technology development) that Computer Scientists should be addressing and currently are not?

My answer would be: the problem that we always do "introducing a new technology" and "thinking about the consequences of it" in the wrong order.