The first rule of modelling is that "All models are inaccurate, though some of them are useful". I have a bit of experience with modelling. (I constructed my first model on a vacuum tube analog computer in the mid 1960s.)
Pat Frank's point about the growth of error bands as the GCMs (Global Climate Models) project into the future is well taken. I admire his perseverance in getting it published. Mere mortals would have given up long before he.
Another big problem with the models is their use a value for the equilibrium climate sensitivity (ECS) that results in positive feedback in the model. This is the value for the rise in temperature due to doubling the amount of CO2 in the atmosphere. The values used in the GCMs result in positive feedback in the control of temperature. Positive feedback always results in out of control performance of the model. It would be the same if your thermostat turned the furnace off rather than on when the set temperature was reached. Your furnace would be off all winter, and on all summer. This is the dreaded "tipping point" described by the Warmists. Once the tipping point is reached, temperature increases exponentially to disaster. Earth would become either a boiling Venus, or a frozen Pluto.If the earth were so precipitously balanced on the edge of collapse, it certainly would be there already, after four and a half billion years of CO2 varying from zero to thousands of PPM.
One learns quickly about positive feedback with vacuum tube analog computers. Ours had about a hundred operational amplifiers, the basis of electronic integrators. Each had a handful of vacuum tubes which would burn out regularly. When an op amp would fail, a loud horn would sound, everyone in the room would hear it, and the repair technician would run over to find and fix the problem. The problem was that the same alarm would sound if the "tipping point" in the model was reached. This was almost always caused by inadvertent positive feedback somewhere in the model. The tech would laughingly point this out to the embarrassed operator (me) who would the slink off to his desk to find the error in his model.
An excellent point. One would think that the people developing the models would think of that as well. In fact, I am sure they have thought of it. So, it is either willful blindness on their part, or intended deception. What a bad stain they are putting on science.