I’m going through a pile of newspaper cuttings for the next issue of my What’s Next report and found this tiny morsel from an old issue of the FT lurking under a pile of New Scientist magazine cuttings.

Philip Tetlock, author of Expert Political Judgement, found that experts of any professional or practical persuasion make very poor forecasters. At best, expert predictions perform little better than randomly sticking a pin on a list of stocks or picking economic predictions out of a hat. So what can one do if what one wants to do is make good predictions?

The answer, according to Tetlock, is contained within an essay published a very long time ago by Isaiah Berlin, which itself harks back to the thinking of the Greek poet Archilochus. The secret, it seems, is to develop a cognitive style that is highly promiscuous, self-doubting, frantically curious and meddlesome. This will, on most occasions, work better than thinking that is based on a single worldview or a lone idea.

In other words, question everything.