I started my physics career as a condensed matter experimentalist at
MIT, and as such I was brought to Cornell as a Xerox fellow. I went
once down into the bowels of Clark Hell, where there was a professor
with an army of people slaving away in dark
cubicles, and I promptly decided to join instead the field theorists who owned
a beautiful rooftop view of the Ithaca hills and Ithaca
skies.
One fateful day Toichiro
Kinoshita came up with a Feynman integral and asked me whether I could
evaluate it for him. No sweat, I worked for a while and not only did I
integrate it, but gave a formula for all Feynman integrals of that
topology. It was only a bait. He came up with the next integral on
which my general method miserably failed. Then he came with the next
integral, and then it was like Vietnam - there was no way of getting out
of it. I was spending nights developing algebraic languages disguised as
editor macros so that synchrotron experimentalists would let me use
their computer; we
were flying in small planes to Brookhaven, carrying suitcases of
computer punch-cards; and by four years later we had completed what
at that time
was the most complicated and the most expensive calculation ever carried
out on a computer, and the answer was:
At the very end, I dreamed that I was a digit towards the end of the long string of digits that we had calculated for the electron magnetic moment, and that I died by being dropped as an insignificant digit. I was ready to move on.