Thanks, Dave! I love that you brought this up during Career Week. I’ve found that regularly tapping into my gratitude for getting to work on high-impact projects not only strengthens my own “why”, but also inspires others—several peers have told me they’re now taking concrete steps to increase their impact. Some ask me if working on such heavy topics is depressing, and I tell them I personally find much more hope & optimism by contributing as opposed to fretting about them without taking action. Our gratitude creates a positive spillover effect, encouraging more intentional helpers. EA isn’t a monolith, and it’s wonderful to see this ethos celebrated alongside the rigorous analytical thinking our community values.

Thanks for the explanation! I think I see what you're saying now. I bet you're right, deaths attributed to the change in temperature would likely remain low. I think most of the deaths would come from food shortages (as of the current data, a lot of land in Europe would no longer be arable) and that could affect a more significant portion of the population.

I probably shouldn't have cited the averages like I did here because it hides the seasonal extremes, also it's really the variance that shows what society will need to contend with if this were to happen. This recent paper by van Westen published just a few weeks ago illustrates the differences in winter/summer temps as well as how "unusual" events will become more frequent. London for example, could face -40C winter weather. https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025GL114611

Curious, how did you calculate your estimate of European population decline?

Wow, that's beautiful. What a majestic, complex planet we live on! 

Thanks for your comment! I’m actually attending the Global Tipping Points Conference right now, and the leading scientists still say there are mixed results regarding net temperature change in post-collapse scenarios and that overall change hides what could be very different summers and winter results. It very much depends on how much warming we have already experienced at the time of a potential AMOC collapse, as well as the state of other interconnected regional climate systems—like the ice sheets. Suffice it to say it’s tough to model. I think it’s fair to say, though, that it wouldn’t just “cancel out,” as there are the ITCZ shift, the monsoon changes, Northern Hemisphere drying, and other knock-on effects outside of surface air temperature.

Thank you for the offer to talk! I’ll message you privately to coordinate a call.
You’re absolutely right, there are skeptical voices in the field and perhaps I did not adequately highlight them above—that’s something I’ll update. Would you mind sharing the specific paper you’re referring to?

I should also note that the preliminary food system data I mentioned was from our model, where we used much lower probability estimates than the more alarming studies I mention in the post—that includes, conservatively, the lower end of the IPCC estimate. I too hold those more extreme concern estimates with a grain of salt.

My goal was to highlight under-resourced risks, not cherry-pick the scariest scenarios. There is definitely a lot of uncertainty, much like nuclear winter as you point out. I just think it is worth doing more detailed scenario planning IF this were to occur, since we certainly can’t rule out that it won’t (we estimate a 0.1%–3% annual chance of AMOC collapse occurring this century).
 

FWIW, Johan Rockström has categorized AMOC as a global catastrophic risk (using the CSER University of Cambridge definition) and believes scientists need to shift the narrative to communicating this more clearly and get away from focusing so much on the wide range of uncertainties, which gives politicians permission not to act.