[mhh__]

By definition the new is not always obviously the next, so "not relevant" requires a crystal ball for anything other than basically already established industries.

Deep learning is beyond what VLSI was for the transistor, for example.