I talked with David Goodstein at Caltech once about the youth legend in physics. His basic view was that people took two data points---Newton and Einstein---and over-extrapolated. He pointed to Feynman (among others) who continued to make major contributions throughout their careers. Of course, a couple counterexamples isn't enough to disprove a trend, but it does seem like the legend is build on a rather flimsy foundation.<p>N.B. I found myself quite enjoying the writing in this article, and I realized that the author was a colleague of mine from college. Lila Guterman was a writer for the <i>Harvard Science Review</i> when I was its editor-in-chief. I particularly recall a bang-up article she wrote on naked mole rats. Glad to see she's keeping up the good work!