> Any bits of data we can collect to make accurate assessments should be collected. This is the kind of thinking that got us into this mess. The data being col

> Any bits of data we can collect to make accurate assessments should be collected. This is the kind of thinking that got us into this mess. The data being collected may not be accurate, may not be interpreted correctly, and may not even be useful, but fearful people like you will happily use it for a tiny sliver of peace of mind. Lets say the pilot on your next flight did call the suicide hotline 15 years ago after his wife and child were killed in a car wreak. Would that make him more likely to be a danger for you today than the pilot on your last flight who was currently suicidal but never called the hotline? Lets say the pilot on your flight wasn't just going through a rough time 15 years ago, but was actively feeling suicidal on the day of your flight. Is it more likely that he would also be a mass murder who wanted to kill everyone on the plane, or is is more likely that he would kill himself later that day while alone in a hotel room at your destination city? Thoughtcrime is a dangerous thing and our algorithms aren't so good at predicting human behavior that we should support handing out extrajudicial punishments to people based on what they might do. Those are things you should fear far more than what the mental health history of your pilot might suggest.