I always end up second guessing myself and wondering if my prescription could be better. One or Two,? Two or Three? I haven't read through carefully, but you'd think that ML would eventually outperform human choice.
I'm very much interested in optimization and decision-making algorithms, and this is the technical problem that I'm devoting my life to, so "relevant" is an understatement. I'm working in a different regime of plasma right now, but the generalist approach taken in my grad coursework is nice to have under my belt. This article suggests that the direction ITER is taking may or may not bear fruit. I find that terrifying, for a project costing over $10 billion. Either way, like the article says, the parameter space is a nightmare. I'm going to spend a large portion of the next few years looking at other possible magnetic confinement configurations. I don't think I'm going to get payed to do it. So much for the music project :/