I work on the ITER project, specifically I develop algorithms for multiscale simulation of fusion plasmas.
I consider it not unlikely that nuclear fusion will be a useful source of energy sometime in this century. And if some game-changing event happens before that, well so be it (included here are FAI, one of a large variety of small-scale fusion devices turning out to be practical, the greens finally getting their act together and starting to produce truly large amounts of "green" energy, ...).
On a meta level: why are comments in this thread being upvoted? Should I conclude that the LW community thinks data fitting is 5/3 more interesting than precambrian atmospheric dynamics?
I'm curious as to whether you use Markov Chain Monte Carlo methods in your work.
There must be quite a few undergrad/graduate/post-doc/???-level researchers on LessWrong. I'm interested in hearing about your work. I'll post about myself in the comments.