Andrew Keen at TechCrunch talks to Clay Christensen about Escaping the Innovator's Dilemma. Clay describes companies with one business model as something similar to organisms in biology that do not adapt or mutate over time to stay competitive. Thus, most companies like this fail in time. He thinks companies need multiple business units running different business models to survive. Old business units get shut down when they are no longer competitive. He cites IBM as the only company to survive the minicomputer and PC eras. IBM succeeded by forming new business units in different locations for their minicomputer and PC products. The mainframe business unit ran with ~60% gross margins, where as the minicomputer and PC business units ran with ~45% and ~25% gross margins respectively.