When it comes to typing people - how can sensors have much of a hope of being accurate.. being that they're most likely typing from superficial behaviours/words, which are all too often sullied by a person's experiences etc? Perhaps we sensors should vacate the area, and leave the intuitives to do the dirty work?