From the New Yorker, February 12, 2007, page 5.

Kolbert makes a mathematical error when she states that U.S. total
annual energy expenditures, of what she says is roughly a tenth of our
gross domestic product, amount to more than a quadrillion (a thousand
trillion) dollars. Since our actual G.D.P. is thirteen trillion
dollars, it is overstating our energy expenditures a thousandfold. A
quadrillion is more than twenty-one times the world's output, of 46.6
trillion dollars, and to reach this figure we'd need a period of
hyperinflation. This is not impossible; after the Second World War,
hyperinflation reduced the value of all the Hungarian currency in
circulation to one-thousandth of one U.S. cent.

The article in which the error appeared was probably in the New Yorker two weeks earlier. I don't have the reference handy.

For information about hyperinflation, visit

www.econlib.org/library/enc/Hyperinflation.html

en.wikipedia.org/wiki/Hyperinflation