Danny Truell is one of the most gifted investment managers in the country and has made literally billions for the Wellcome Trust since he took over the running of its funds a few years back, enabling it to maintain its position as one of the largest charitable funders of medical research in Europe.
On Thursday I was lucky enough to hear him give one of his rare public speeches at an investment seminar organised by Charity Times.
His approach to investing for the long term would have been music to the ears of the Government, but it is his comment on governance I want to share.
Companies should have scientists on the board if at all possible, he counselled. The first reason is that they are experts in their own field, and therefore respect people who are experts in turn in another fields. The respect means they don't interfere or second-guess him as the expert in investment management. Second, he said, scientists are usually good at mathematics which comes in useful when the board has to grasp some of the more arcane aspects of investment management theory. Third, he says, they love a good debate.
But what really struck home was his comment that board members are allowed to say "no" when there is something they really don't like, but not allowed to say "yes".
His reason? "If you try to get a consensus among intelligent people you end up with the lowest common denominator."
It is probably why boards often go spectacularly wrong.