This isn't just about being able to decompress video on your smartphone so you can watch youtube without draining your battery, by the way. All compression algo

This isn't just about being able to decompress video on your smartphone so you can watch youtube without draining your battery, by the way. All compression algorithms, whatever domain they operate in, are about understanding the entropy stored in a stream of data and using that understanding to extract just the useful information. (Lossy compression incorporates understanding of our perceptual systems to only transmit what is necessary to transmit relevant entropy from the source to our brains.) The primary use is to display media streams with reduced network, CPU, and battery usage. But it's a much more general field, replete with all kinds of fascinating insight about the nature of visual/aural manifestations and human perception. Anything you can remove yet still recover the relevant parts of the original signal is an opportunity to understand how our brains, and the world in general with its physical laws, work. There's a ton of interesting and useful research that could be done -- but would you really want to look into how texture is encoded, processed, and reacted to in our neurons when you have to worry about various mafia-like patent licensing consortiums? The whole licensing setup has a chilling effect on many different area of inquiry, even apart from its impact on classic compression algorithms. (Though that is harmed too -- if you come up with an incremental improvement that works really well for some type of input, and you'd like other people to build on your work, what do you do?)