In technology, the odds of making something truly new and popular have always tilted toward failure. That was why Mervin Kelly, the president of Bell Labs, let many members of his research department ...