We want to BUILD systems that can generalize to a new task. Being able to do things with much LESS data and with much less computation is GOING to be interesting and IMPORTANT.
I think one of the things about reinforcement learning is that it TENDS to REQUIRE exploration. So using it in the context of physical SYSTEMS is somewhat hard.
That's the thing with ELECTRONICMUSIC, you set up systems to BRING in an ACCIDENT, to bring in quirks you didn't choose, but you STILL will have had to set up systems.