Problems we don't know the right questions for yet. When we have a good handle on a question, it becomes science. When we have a good answer for the question, it becomes settled science.

Philosophy consists of the questions that we don't understand well enough to even know how to go about answering them, but which, despite that (or because of that), are still really fun to argue about endlessly even in the absence of any new insights about the structure of the problem.

(Basically, I think describing a given problem as "philosophical" is mostly mind projection; from history, it seems that all the qualities that make a given problem a philosophical one have been properties of the people thinking about it rather than of the problem itself.)