[ToJoh]

A couple of thoughts:

* The models are continuously updated. Recent measurements of land surface temperatures and polar ice sheets have been tracking worst case scenarios, not the previous expected mean (see e.g. Arctic ice sheet melt-off in 2020). * There are considerable confidence intervals around predicted temperature increases and related forecasts, e.g. GDP growth. * There are considerable nonlinear dynamics such as thawing permafrost or sea-floor deposits of methane, or death of carbon sinks such as forests. This means that the outcomes of the models can diverge wildly from the mean expected outcome. * I believe IPCC models do not take into account the political upshot of massive loss of farmland and habitable land in the tropics (i.e., refugees and wars).