The computer says...

Following on from yesterday's post, it has been pointed out that the projections of the extent of future global warming have to be taken seriously because they are based on computer models, as though somehow the fact that they have been produced by computers gives them the status of Holy Writ. I think I've mentioned before that I go back to the early days of computers, when for the most part you had to write your own programs. Stuck to the side of my computer used to be a piece of paper with GIGO written on it. It was an acronym much used at the time (although there never seemed to be any agreement as to whether it should be pronounced jee-go or ghee-go) but it stood for "Garbage in, Garbage out" and was a constant reminder that what a computer produces is only as good as the data it is given and the program it is asked to run. Putting it baldly, if you ask a computer the wrong questions, you cannot be surprised if it comes up with the wrong answers. So I think we need to know a little more about how these models were constructed, and what assumptions they were based on before we pass judgment.