A great time for building great companies: “there are really four types of predictions based upon two criteria: accurate and inaccurate, consensus and non-consensus. The problem is that only accurate non-consensus predictions are valuable.”