Data-mining is the use of mass-data to extract behavior patterns such as food purchases or clothes consumption. That will sound rather innocent compared to this: a scientist at the MIT is willing to learn about individual behavior by analyzing, -- in real time of course -- data collected by our cellular phone. As explained in the last issue of the MIT Technology Review, Sandy Pentland is working on ways to improve social networking (he's trying it with his students and colleagues) by finding where and with whom people spend their time. It can even predict such behavior, using statistical models. Worse, reality mining will even be able to detect the most intimate feeling like depression (through voice analyzer) or Parkinson tendencies (by analyzing data from accelerometers embedded in your phone). Such technology can also be used for a vast spectrum of applications. One example is improving computer models for the spread of contagious diseases. Another reason to be convinced that the mobile phone will be much bigger than the personal computer. This one is unpleasant.