It is often said that evolution means that we arose by chance, but this seems to me to be a strange use of the word. There are many possible form of living things, and evolution explores a great number paths through these to select those which are optimised for survival in a given environment. Is this just chance? Well supposing you have an important message for a friend, but are told that he had gone shopping in town, wouldn't be back for some time, and there's no way of contacting him. It's vital he gets the message as soon as possible so you head into town and spend two and a half hours searching through a great many of the shops in turn until you eventually find him. Whether or not you find him in a particular shop might be thought of as a matter of luck, but I don't think that you'd like it if someone said afterwards 'wasn't it lucky that you just happened to bump into him'. I would say that this is the sort of luck that is involved in evolution.

When quantum theory was being developed it soon became clear that it was a very strange subject, and would need a different mindset to interpret it. The idea took root that each time a measurement was made a random process took place which selected the result of the measurement from a set of possibilities. Now over time people get used to some strange ideas, and I think this it would have been the same here were it not for one big problem - it's wrong. Einstein, Podolsky and Rosen implied this in their 1935 paper - if you generated two particles which were linked in some way, then measurements made on each of them would be likewise linked, showing the two measurements couldn't be individually random. However this was dismissed, and the idea of intrinsic random events remained as strong as ever in many people's minds. Indeed popular opinion seems to be that quantum theory implies randomness everywhere, although in fact the theory is wholly deterministic on microscopic scale - it's when you try to link it to the macroscopic world which we experience that you encounter problems (and quantum theory isn't like thermodynamics - its not simply that randomness can be a stand in for a lack of precise knowledge) .

In the 1960's John S. Bell realised what a mess quantum theory was in and decided to see if it could be put right. If individual random events were wrong, then was Einstein right in that there had to be an underlying deterministic theory? Somehow this didn't seem to work either. Bell showed that the question was not one of randomness v determinism as one of locality v nonlocality. Any way of explaining quantum behaviour has to involve faster than light influences, although these can't be used to send a signal faster than light. This is where it become useful to propose an intrinsic randomness. If two measurements are found to be correlated then in a deterministic theory there would be some way of using this to send a signal. If you say that the correlation is linked to a random event, then you can exclude the possibility of signalling. Many physicists think that this explanation is satisfactory and that there is no reason to look for alternatives.