How can deep problem solving be monetized if one is unsure where the research/tech will move next? A good example is AR with a number of potential technical problems that are pressing.