Savvas draws a hard line between first-principle thinkers who deconstruct problems from the ground up, and people who say 'this worked there, it'll work here.' MrQ's interview process tests for this directly — if you can't name the last time you found an unconventional route to solve a problem, you're not a fit.