Teaching data science
AA: Speaking of students, one of the reasons you left NASA was to establish the world’s first data science program. What do you think are some of the key elements of a modern program these days? What should we really be teaching students that we’re not teaching them at the moment?
KB: That’s a really good question because I’ve been involved in a number of panels where they’ve discussed this, and some of them have gone on for days, and some of the reports have gone on for dozens of pages. You can get really into the weeds with this one.
The key elements are the higher-level things. There are mathematical foundations. There’s modeling. People underestimate the power of modeling. A model is a representation of a thing. It’s not the thing. We use data as a representation of a customer. The data is not the customer. What you purchase is not you; it’s a representation of the things you like or are interested...