Climate Models Fail

Predictions of global warming, scary enough to cause countries to change their policies, have all failed to predict both the last year's cooling trend and the last three years' cooling of the ocean.  Those in the global warming camp are talking around the failures of the models with discussion of temporary downturns and El Ninos, but the fact remains that the only evidence for those dire predictions in 50 or 100 years, the only evidence causing whole countries to try to change their energy infrastructure, are the models that have failed to predict the current conditions.  Isn't the only value of a computer simulation that it be able to predict what will happen? 

Are they really trying to get us all to believe that, "Well, yes, the models might fail to predict yearly trends but they are correct for the long term."  When you look at the model's progression above, do you see a plateau for 1998 through 2007?  Do you see any points on the graph in 2007 falling down to 2001 levels?  Are we to expect the temperature to radically jump higher in the future just because the chart says it will?  And look at the range of the models' predictions:  anywhere from  1.5 degrees to nearly 6 degrees hotter in the next 100 years.  If a 6 degree increase can cause global catastrophe and the models have a 5 degree span of prediction -- what does that tell you about their worth?