As change accelerates, so does the danger of holding on too tightly to yesterday's certainties - a lesson that many senior executives have yet to learn. Again and again I've seen great companies laid low by their reverence for precedent. Nokia, for example, created a big opening for Samsung when it was slow to question its long-held belief in the design superiority of "candy bar" shaped phones (vs. the two-piece flip-phones preferred by Asian customers).

To keep learning, you constantly must keep unlearning. First, this requires an effort to identify precisely those beliefs that seem most indisputable; usually that means beliefs that are the most widely held. Second, it demands a continuous search for disconfirming evidence. For example, like most B-school professors, I long accepted the notion that big companies can't manage without managers. After all, in any complex undertaking there are lots of things that need managing - and who else to do this work but, well, managers. Then I started looking for counterexamples, like the open-source movement. According to Sourceforge.net, there are currently over 1.4 million volunteers contributing to more than 130,000 open-source projects, with nary a manager in sight. Turns out you don't always need managers to aggregate human effort in productive ways - a revelation that undoubtedly came as a bit of a shock to a legion of project managers at Microsoft, Sun Microsystems, and other software incumbents.