I don't think UNIX is a virus, it's not like Windows is radically different. These types of operating systems have shapes largely defined by the hardware. Stil

I don't think UNIX is a virus, it's not like Windows is radically different. These types of operating systems have shapes largely defined by the hardware. Still, I definitely think you can do way more exciting stuff with OS design than what anyone currently does. It would be at the higher levels rather than how the kernel works, but you can do it. Academia is a particular disappointment in this regard - that's the one place that should be able to do really interesting stuff here, yet academic OS research is dead. There's less scope to do new stuff in PL research, IMO. Most of the radical ideas there have hit dead ends. The whole pure FP thing turned out to be a dead end, when was the last time you thought about Haskell? Probably even Clojure has done better. It's not clear where you can go from here beyond usability improvements which is probably why modern languages all look the same. Hardware: there's definitely a ton of scope to do cool stuff with AR/VR goggles in the work environment. Web: Java isn't less closed than the web. Arguably it's actually more open as there are interestingly different implementations of the JVM, it's a way more modular ecosystem that's way less dependent on the absurd army-sized dev team that Chrome requires.