It is extremely important to recognise the nature and limitations of the computer models on which the case for anthropogenic global warming (AGW) is based. In short, they are flawed:
They make the basic assumption that atmospheric carbon dioxide is responsible for atmospheric warming. It is therefore inevitable, from the initial design of the model, that they will predict that an increase in CO2. will lead to an increase in global temperatures. If, on the other hand, they made the initial assumption that cosmic ray flux was responsible for global cooling, we would not be surprised to find that the models predicted that temperatures increase when cosmic ray flux decreases. But, of course, governments cannot tax cosmic ray flux and pretend that, by doing so, they can control it.
These computer models based on CO2 do not adequately describe our climate history; in other words, they are very obviously wrong.
They have also introduced a philosophical flaw. Those who rely on them treat their output as empirical "data". But they are not empirical data, except in some species of virtual reality. In real science, data are produced from observation or from experiment in the real world. To quote Hendrik Tennekes (former Director of Research, Koninklijk Nederlands Meteorologisch Instituut), an expert in scientific modelling: "The blind adherence to the harebrained idea that climate models can generate 'realistic' simulations of climate is the principal reason why I remain a climate skeptic."