Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio.
This is: On the assessment of volcanic eruptions as global catastrophic or existential risks , published by Mike Cassidy, Lara Mani on the AI Alignment Forum.
By Mike Cassidy and Lara Mani
Ord 2020, and others have suggested that the existential risk posed by volcanoes is the largest of the potential ‘natural’ catastrophes - 100 times that of asteroids and comets combined. As volcanologists we wanted to delve a bit deeper into the x - risks and also global catastrophic risks from large explosive eruptions, adding insight from the latest volcano and climate science, including challenging some of the assumptions that have been put forward so far. Significant global impacts from large explosive eruptions will be described, which have a ~1 in 6 probability of occurring this century.
Introduction
I’m Mike and new to EA world, having just been introduced to it early this year from a friend in Oxford and having thoroughly enjoyed reading The Precipice, the super-eruption part sparked my interest. I’m a senior research fellow based in the earth science department at the University of Oxford, and specialise in volcano research. I research a range of different aspects in volcanology, but principally I aim to understand what influences the explosivity of eruptions and how best to forecast them. Lara works for the Centre for the Study of Existential Risk (CSER) at the University of Cambridge and has been thinking about extreme volcanic risk scenarios since joining the centre in January 2020. In particular, Lara has been exploring the systemic nature of volcanic risk from the view point of cascading and compound risks.
The idea of this post here is to provide some thoughts that might interest the EA community from a different perspective. I am also slowly catching up with the existential risk literature (this is partly time, but also because I wanted to provide a fresh insight initially without thoughts being diluted by later reading), so if you know literature that may be connected with some of the themes we cover, then please let us know. This really is the start of us thinking about this topic and as you’ll see there’s lots of uncertainty and we’d welcome some thoughts and questions about this generally to assess the gaps going forward.
Challenging assumptions and why we think the current risk may be underappreciated
The focus of impacts from large explosive eruptions has so far focussed on the initial cooling effect (which could be substantial), but the effects are far more varied than this, which we’ll discuss in later in this article and in our modern, connected world, perhaps means that we could more vulnerable to these. In this post we aim to challenge two notions that have been put forward in the existential risk community previously:
1) That the risk from volcanic eruptions is ‘natural’ and hasn’t changed/won’t change going forward
2) That humans have survived 2000 centuries therefore the natural risk must be low
The first assumption is challenged by recent studies which show that anthropogenic climate change increases the climatic cooling effect of large magnitude explosive eruptions (potentially by as much as 60%; Aubry, 2021 & Fasullo et al 2017); climate change itself also may increase the likelihood of triggering eruptions through glacial retreat and sea level change. Furthermore as we explore below, the more extreme climatic effects are not always attributed to the largest magnitude eruptions, in other words, you don’t need a ‘supereruption’ to cause global climatic impacts. All these factors, along with new ice core records showing large explosive eruptions are more common than values Ord 2020 used, mean that our current base rates for global catastrophic risk and existential risks might be higher than initially appreciated.
The 2nd assumption seems to be more based on hu...
view more