Iāve recently begun researching nuclear risk for Rethink Priorities, but the views in this comment are just my personal views. These views are also pretty tentative, as Iām still at the early stages of learning about these topics.
Iāve read two books focused on nuclear risk:
Command and Control
Iād recommend this for someone whoās pretty interested in nuclear risk.
A lot of this book seemed interesting and useful, but, if I recall correctly, it spent more time on the Damascus explosion than Iād have liked.
Iād recommend this for someone whoās pretty interested in nuclear risk.
Has a narrative feel, which made it quite easy to read.
Iām somewhat skeptical of a decent portion of whatās in the book.
E.g., if I recall correctly, Ellsberg seemed to imply that there was scientific consensus that nuclear winter would be very likely in most nuclear exchanges, and would have very severe consequences (e.g., the death of the vast majority of the global population). But my impressionāpartly based on this Wikipedia articleāis that thereās still substantial scientific debate on these points.
But I still think that reading the book was useful.
Iād recommend this, including for people who are just generally interested in things like war and global catastrophic risk, rather than very specifically interested in nuclear risk.
Iāve read one book focused on trends and drivers of violence more generally, with some parts on/ārelevance to great power war:
This is of course Better Angels of Our Nature
I think this is in general more useful than the above three books. But itās very long, and much of it isnāt very focused on great power war or nuclear risk.
And Iāve read one book focused on existential risks but with ~10 pages on nuclear risk (at the start of chapter 4):
The Precipice
Thisād perhaps be my #1 recommended book for most EAs and EA-inclined people in general.
I think the ~10 pages on nuclear war was probably more useful for that topic per page than the above three books. So I might recommend reading those pages before reading the other books.
I list things Iāve written relevant to The Precipice, and some reviews/āinterviews about The Precipice by others, here.
Iām also aware of but havenāt (yet) read these books focused on nuclear risk, WMDs, and/āor great power war:
Seems to be focused on both nuclear risk and biological weapons.
Atomic Obsession
I learned of this via the post Notes on āAtomic Obsessionā (2009), which I think is itself worth reading, as are Max Danielās comments on that post.
Seems to be focused on nuclear risk.
After Tamerlane: The Rise and Fall of Global Empires, 1400-2000
I believe this is mostly focused on interventions to mitigate how bad nuclear winter would be, if it happened.
Some of the books Luke Muehlhauser lists/ārecommends here
I think that all of the above books are available as audiobooks except Atomic Obsession. But Iād recommend reading The Precipice as an ebook or physical book rather than as an audiobook, as the audiobook doesnāt contain the (huge number of!) interesting footnotes.
(If you want a list of all the EA-relevant books Iāve read since learning about EA, in roughly descending order of how useful I perceive/āremember them being to me, thatās available here.)
Iāve read one book focused on trends and drivers of violence more generally, with some parts on/ārelevance to great power war: This is of course Better Angels of Our Nature.
Only the Dead is basically a pretty effective take-down of Pinkerās analysis of trends in interstate war. Some key points are: (i) Pinker focuses on wars between European states, or wars between (typically European) āgreat powers,ā rather than interstate war generally. (ii) Pinker doesnāt adjust for changes in the number of states in his datasets (Europe used to be a lot more fragmented). (iii) Pinker mostly just eyeballs graphs instead of running statistical tests, which is a bad idea when data is small-n and non-normally distributed.
The author instead analyzes the Correlates of War dataset (which is meant to record all interstate wars since 1816) and concludes that thereās been no consistent downward trend over the past 200 years. The rate of interstate war has bounced up down but supposedly is no lower now than it was in the early 19th century. Thereās also no evidence of a change in the distribution of the deadliness of wars, in terms of the portion of people who die in each participating state.
I also donāt totally buy the analysis, though, since there are significant issues with the Correlates of War dataset. The dataset probably makes the nineteenth century seem more peaceful than it actually was, because it uses very strict/āweird criteria for what counts as a āstateā that leads it to exclude nearly all non-European states in the nineteenth century. (Europe was probably the most peaceful part of the world at the time.) The dataset also begins in what was probably the most peaceful period of European history yet, right after more than two decades of constant war. So a more complete and long-run dataset might actually show something different.
The Causes of War and the Spread of Peace mostly agrees with Pinker, and has some similar issues with rigor, but (IMO) is better argued, discusses a wider range of possible explanations for the apparent decline of war, and also has a really interesting treatment of the causes of pre-modern war. I particularly like the emphasis it gives to the Industrial Revolution as a turning point in the history of war.
Hey Ben, thanks for those recommendations! I hadnāt heard of them, and both sound interesting and potentially useful. Iāve now downloaded Only the Dead, and made a note to maybe read The Causes of War and the Spread of Peace after that.
Iāve recently begun researching nuclear risk for Rethink Priorities, but the views in this comment are just my personal views. These views are also pretty tentative, as Iām still at the early stages of learning about these topics.
Iāve read two books focused on nuclear risk:
Command and Control
Iād recommend this for someone whoās pretty interested in nuclear risk.
A lot of this book seemed interesting and useful, but, if I recall correctly, it spent more time on the Damascus explosion than Iād have liked.
Recommended by Rob Wiblin, by Nick Beckstead, and as āOutstandingā by Luke Muehlhauser.
The Doomsday Machine
Iād recommend this for someone whoās pretty interested in nuclear risk.
Has a narrative feel, which made it quite easy to read.
Iām somewhat skeptical of a decent portion of whatās in the book.
E.g., if I recall correctly, Ellsberg seemed to imply that there was scientific consensus that nuclear winter would be very likely in most nuclear exchanges, and would have very severe consequences (e.g., the death of the vast majority of the global population). But my impressionāpartly based on this Wikipedia articleāis that thereās still substantial scientific debate on these points.
But I still think that reading the book was useful.
One of Rob Wiblinās top 9 book recommendations.
See here for a review from an EA.
Iāve read one book focused on great power war:
Destined for War
Iād recommend this, including for people who are just generally interested in things like war and global catastrophic risk, rather than very specifically interested in nuclear risk.
Recommended by Rob Wiblin.
Iāve read one book focused on trends and drivers of violence more generally, with some parts on/ārelevance to great power war:
This is of course Better Angels of Our Nature
I think this is in general more useful than the above three books. But itās very long, and much of it isnāt very focused on great power war or nuclear risk.
Highly recommended by Nick Beckstead and by Luke Muehlhauser, and recommended by Rob Wiblin.
And Iāve read one book focused on existential risks but with ~10 pages on nuclear risk (at the start of chapter 4):
The Precipice
Thisād perhaps be my #1 recommended book for most EAs and EA-inclined people in general.
I think the ~10 pages on nuclear war was probably more useful for that topic per page than the above three books. So I might recommend reading those pages before reading the other books.
I list things Iāve written relevant to The Precipice, and some reviews/āinterviews about The Precipice by others, here.
Iām also aware of but havenāt (yet) read these books focused on nuclear risk, WMDs, and/āor great power war:
The Dead Hand
I learned of this via the post Key points from The Dead Hand, David E. Hoffman, which I think is itself worth reading.
Seems to be focused on both nuclear risk and biological weapons.
Atomic Obsession
I learned of this via the post Notes on āAtomic Obsessionā (2009), which I think is itself worth reading, as are Max Danielās comments on that post.
Seems to be focused on nuclear risk.
After Tamerlane: The Rise and Fall of Global Empires, 1400-2000
Arden Koehler /ā 80,000 Hours lists this as a useful book in relation to great power conflict
Feeding Everyone No Matter What
I believe this is mostly focused on interventions to mitigate how bad nuclear winter would be, if it happened.
Some of the books Luke Muehlhauser lists/ārecommends here
I think that all of the above books are available as audiobooks except Atomic Obsession. But Iād recommend reading The Precipice as an ebook or physical book rather than as an audiobook, as the audiobook doesnāt contain the (huge number of!) interesting footnotes.
(If you want a list of all the EA-relevant books Iāve read since learning about EA, in roughly descending order of how useful I perceive/āremember them being to me, thatās available here.)
I would recommend Only the Dead and The Causes of War and the Spread of Peace over Better Angels.
Only the Dead is basically a pretty effective take-down of Pinkerās analysis of trends in interstate war. Some key points are: (i) Pinker focuses on wars between European states, or wars between (typically European) āgreat powers,ā rather than interstate war generally. (ii) Pinker doesnāt adjust for changes in the number of states in his datasets (Europe used to be a lot more fragmented). (iii) Pinker mostly just eyeballs graphs instead of running statistical tests, which is a bad idea when data is small-n and non-normally distributed.
The author instead analyzes the Correlates of War dataset (which is meant to record all interstate wars since 1816) and concludes that thereās been no consistent downward trend over the past 200 years. The rate of interstate war has bounced up down but supposedly is no lower now than it was in the early 19th century. Thereās also no evidence of a change in the distribution of the deadliness of wars, in terms of the portion of people who die in each participating state.
I also donāt totally buy the analysis, though, since there are significant issues with the Correlates of War dataset. The dataset probably makes the nineteenth century seem more peaceful than it actually was, because it uses very strict/āweird criteria for what counts as a āstateā that leads it to exclude nearly all non-European states in the nineteenth century. (Europe was probably the most peaceful part of the world at the time.) The dataset also begins in what was probably the most peaceful period of European history yet, right after more than two decades of constant war. So a more complete and long-run dataset might actually show something different.
The Causes of War and the Spread of Peace mostly agrees with Pinker, and has some similar issues with rigor, but (IMO) is better argued, discusses a wider range of possible explanations for the apparent decline of war, and also has a really interesting treatment of the causes of pre-modern war. I particularly like the emphasis it gives to the Industrial Revolution as a turning point in the history of war.
Hey Ben, thanks for those recommendations! I hadnāt heard of them, and both sound interesting and potentially useful. Iāve now downloaded Only the Dead, and made a note to maybe read The Causes of War and the Spread of Peace after that.
Yes, there is more detail on the nuclear risk in this paper. And this paper on a fault tree model of the chance of nuclear war.
Oh, and I think one chapter of Global Catastrophic Risks is on nuclear risk, but I havenāt read it.