The implication behind this argument is that the uncertainty in these models will overestimate potential changes. What will happen in a future where the models underestimate climate changes in generally unmentioned.
Luckily, we don't need to look to the future to investigate that possibility anymore. The above graph (from Neven's excellent Sea Ice Blog) shows the actual decline in summer arctic sea ice, relative to various computer model predictions. As this graph shows, ice is now melting much faster than any of the models had predicted.
Of course, the most persistent pro-pollution propagandists tell us that this proves that the computer models are useless, which means that climate change can't be real, which means that any effects we see must be caused by the warming fairies instead of exhaust gasses.
What we, as scientists, would really like is this: We would like to be able to predict the effects of pollution on the climate before they happen. That's why we get into science. The whole purpose of the field is to make predictions about the natural world and then test them. So if y'all cook the Earth faster than we can make decent predictions about the warming, then we get very disappointed. Not as disappointed as all the retirees on the Jersey Shore who just lost their houses, but still not real happy. So folks, here is a request.
Could y'all please slow down the warming of the planet just enough so that we, the research community, can actually catch up and figure out who is happening to this atmosphere?We would much rather predict doom and gloom for the future than look at last week's disaster and shrug , "Yeah. We should have thought of that."