Great interview, I enjoyed both Blink and Tipping Point. However, I bristle at this: WSJ: Do you worry that you extrapolate too much from too little? MG: No. I

Great interview, I enjoyed both Blink and Tipping Point. However, I bristle at this: WSJ: Do you worry that you extrapolate too much from too little? MG: No. It's better to err on the side of over-extrapolation. I definitely disagree. Overextrapolation is a problem in our society. People try so hard to find signal where there's only noise, then find some conclusion and pass it as fact. Eventually someone of consequence believes it and it ultimately pollutes the information landscape. Most people don't understand science or statistics, so it frightens me that spurious correlations (with terrible r^2 values) seem to become new "facts". I used to work in a genome sequencing lab. DNA has lots of secrets for us to uncover. DNA might contain clues for discovering genes related to diseases like Alzheimers. One might extrapolate that genetic code contains the possibility of understanding or even curing some disease. The problem is thats a dangerous extrapolation and when it becomes commonplace to believe that human genome research can rid the world of disease (which some people somehow subscribe to), we've overstepped some boundaries. Overextrapolation, as a rule, is not the way to go.