It has become a familiar sight; speakers will talk and fill the allotted time, but we will learn nothing new. A panel of grandees is likely to be even worse. Put five of them together and, however smart they are as individuals, it is almost guaranteed that none will say anything interesting.

This form of speech is akin to blowing bubbles, and professional politicians, business leaders and diplomats become very skilled at doing it.

Events like Davos or the UN General Assembly confirm this rule. Often the fringe events are full of fascinating detail, as are the one-to-one conversations. But the keynotes and the plenaries tend to be extremely dull.

I've been intrigued by this phenomenon for a while, and now view it as an example of collective stupidity—which raises the question, how do you turn a group that is individually smart into something that is collectively dumb? This question is clearly part of a broader research programme that also applies to nations, firms and other organisations. Some make average people collectively brilliant, whilst others do the opposite—at Nesta's Centre for Collective Intelligence Design we've begun searching for insights into how we can make groups of individuals collectively smart by combining human intelligence with machine intelligence.

Yet, when attending big conferences, I’ve also realised that this norm of blowing bubbles clearly does perform a useful function; it restores (perhaps subliminally) our confidence in the fact that the world is predictable, legible and stable. The very dullness of the speeches may be precisely their point.