He's saying that it's difficult to think creatively without any constraints, but it's also difficult to actually create something when there are too many constraints. So the ideal strategy for the entrepreneur is to find an area with the optimal amount of constraints, and then keep working on that problem until you ship. (That is, you need to avoid jumping to a new problem when the first problem gets difficult.)

I'll agree it's a little esoteric, but it's a pretty good recipe for shipping something innovative.