As the 21st Century proceeds, tasks facing our species will become increasingly complex. Many problems that in an earlier era might have been easily addressed by a single person will now require a sophisticated set of abilities contributed by different people. The individual contributions must be deeply complementary; the whole must be more than the sum of its parts.
This much seems obvious. But no part of contemporary formal education—at any point from kindergarten through post-graduate work—is designed to teach people how to interact effectively with other people when they participate in goal-oriented groups. When such a group functions well it synergizes the talents and abilities of its members. But at present, such synergy occurs because of a lucky combination of people—people who happen to have complementary skills and abilities that are relevant for the task at hand and happen to be able to interact effectively. It is not obvious how best to compose a group to facilitate such synergy. But most people don't seem aware that there's a problem here.
Consider a similar situation: Many people who interview applicants for jobs think that they are good at picking the "right" applicant. They think that they know how to pick appropriate employees based not just the content of the applicant's answers, but also on his or her nonverbal behavior while answering. But it has been repeatedly shown that interviewers who rely on intuition and "feeling" generally are not good at picking job applicants.
So, too, with selecting people to work together in goal-oriented groups. People have intuitions about how to assign individuals to groups and how to organize them, but decisions based on such intuitions are not necessarily any better than chance. Relying on the luck of the draw won't be very effective as task-oriented groups face increasingly complex challenges. We must overcome such intuitions. We need to realize that understanding how best to select the right people for the right group is a very hard problem, and understanding how they should interact most effectively in a group is itself a very hard problem.
To compose a group that will work effectively to tackle a complex problem, we need to know: (i) how to analyze the nature of tasks in order to identify which skills and abilities are necessary to address a particular task, (ii) how to identify such skills and abilities in individual people, and (iii) how different sorts of people can interact most effectively when working on a particular sort of task. Much research will be required to crack these problems (and such research is already underway), but the results of such research will not be widely used as long as people don't recognize the nature of the problems and why they are important.
Science can do better than intuition—but we first must understand that intuition isn't good enough. And this isn't intuitively obvious.