The models employed to make projections of global warming are all tuned to enable them to fit data from the instrumental era. Tuning involves fixing the parameterization schemes employed to represent physical processes that are not resolvable explicitly at the spatial and temporal resolution of the coupled structure. It is unclear whether this may significantly prejudice the quality of global warming projections. The use of multi-model ensembles of projections provides one means of alleviating the biases that inevitably arise. A further means of testing the fidelity of these models is to compare the predictions they make of climate system phenomenology that has occurred in the (deep) past. The Younger-Dryas climate reversal that occurred during the most recent transition from glacial to interglacial conditions has become a prime target for such work. A novel hypothesis advanced for its explanation by the Toronto group 5 years ago has recently been fully verified. It involves the reaction of the thermohaline circulation of the oceans to the meltwater released to the oceans during the deglaciation process. I will discuss the current state of understanding.