So, the message in this paper is that we should not only discuss the likely outcomes, but also the low-probability, high-impact outcomes. I think this makes a lot of sense. Climate change is probably irreversible on human timescales and so I think we really do want to avoid these low-probability, high-impact outcomes. Hence, it’s important that we discuss this publicly.However, this isn’t really the first time that this has been suggested. Michael Tobis generated a figure illustrating a similar point. The public discussion seems to involve people who think climate change will be beneficial, or have minimal impact, and others who think there could be substantial costs but who avoid discussing the possibility of catastrophe. The latter may be very unlikely, but the impact would be so great that it is an outcome that we should probably not ignore.
I was going to say one more thing about the first figure I included in the post. I realise it’s probably just meant to be illustrative, but it’s not (I think) really correct. The impact really depends on how much we warm, not on climate sensitivity alone. How much we warm depends on climate sensitivity and on how much we emit. The latter probably makes this quite complicated because we can’t assign a simple probability to how much we will emit in future (it depends on what choices we make in future). The paper does mention the transient response to cumulative emissions, and emission pathways, so it’s not completely ignoring this. However, I do wonder if doing this rigorously is actually quite difficult. This doesn’t, however, mean that I don’t think this is a reasonable suggestion.