I see some sort of internship process in the future for this...where they make high schoolers do this for 40 hours a week <___< . I mean, if humans do better on it, why not make it some sort of job, eh
FoldIt (http://fold.it/portal/) is similar. Humans tend to find a better starting structure while computers are better at refining that structure.
The scientific method wins over mechanical application of top-down rules? Why yes, yes it does :)
Right. The headline should really be "The public, playing a molecule-building game, outperforms computer algorithms made by scientists". Which is distinctly less interesting, since individual people (never mind whole populations) routinely outperform algorithms; otherwise artificial intelligence would be a solved problem.
It is still pretty impressive because these are biology novices, and RNA folding is a somewhat complicated biological process. AI does outperform humans on a regular basis, especially in narrow domains (chess, checkers, calculation, now jeopardy)