theorywatch

Via Allan Cho’s blog: do computing power and massive datasets spell the end of the scientific method as this article in Wired suggests? The Google guy’s twist on George Box’s aphorism (“All models are wrong, but some models are useful”) is clever, but I’m quite sure that no “Big Blue” “cloud” of computers is ever going to make the maxim “Correlation is not causation” obsolete. (Read the 60+ comments at the end of the Wired piece for some rather trenchant remarks concerning all of this, as well as a more measured response from John Zimmer at Ars Technica.)