Read other letters about this article
The "consensus based on predictions that work" is JUST GOSH_DARNED WRONG!
1) The computer models DON'T work. They constantly drift off their predicted lines and the modelers must then "adjust" (Read FUDGE) their parameters to make the models demonstrate something at all reasonable. Which is to say, to make them agree with what they THINK they should show! This is HARDLY a climate model that works. If the models can't predict accurately what will happen next month, then it is darn sure they aren't accurately predicting--"working"--decades down the road.
2) The models are supposedly predicting something that will happen decades or centuries or eons hence--we have NO WAY TO SHOW THESE "WORK" unless you have an operating time machine. If you do, please demonstrate.