Early in his research (influenced by 'fail fast' startup advice), Matuschak didn't think deeply about problems, instead rapidly prototyping ideas. This led him to miss opportunities for developing strong theoretical understanding before building.