works
Carl Shulman Comment on "Nuclear war is unlikely to cause human extinction" online I agree it’s very unlikely that a nuclear war discharging current arsenals could directly cause human extinction. But the conditional probability of extinction given all-out nuclear war can go much higher if the problem gets worse. Some aspects of this: at the peak of the Cold War arsenals there were over 70,000 nuclear weapons, not 14,000, this Brookings estimate puts spending building the US nuclear arsenal at several trillion current dollars, with lower marginal costs per weapon, e.g. $20M per weapon and $50-100M all-in for for ICBMs, economic growth since then means the world could already afford far larger arsenals in a renewed arms race, current US military expenditure is over $700B annually, about 1/30th of GDP; at the peak of the Cold War in the 50s and 60s it was about 1/10th; Soviet expenditure was proportionally higher, so with 1950s proportional military expenditures, half going to nukes, the US and China could each produce 20,000+ ICBMs, each of which could be fitted with MIRVs and several warheads, building up to millions of warheads over a decade or so; the numbers could be higher for cheaper delivery systems, economies of scale and improvements in technology would likely bring down the per warhead cost, if AI and robotics greatly increase economic growth the above numbers could be increased by orders of magnitude, radiation effects could be intentionally greatly increased with alternative warhead composition, all-out discharge of strategic nuclear arsenals is also much more likely to be accompanied by simultaneous deployment of other WMD, including pandemic bioweapons (which the Soviets pursued as a strategic weapon for such circumstances)and drone swarms (which might kill survivors in bunkers); the combined effects of future versions of all of these WMD at once may synergistically cause extinction

Comment on "Nuclear war is unlikely to cause human extinction"

Carl Shulman

Effective Altruism Forum, November 7, 2020

Abstract

I agree it’s very unlikely that a nuclear war discharging current arsenals could directly cause human extinction. But the conditional probability of extinction given all-out nuclear war can go much higher if the problem gets worse. Some aspects of this: at the peak of the Cold War arsenals there were over 70,000 nuclear weapons, not 14,000, this Brookings estimate puts spending building the US nuclear arsenal at several trillion current dollars, with lower marginal costs per weapon, e.g. $20M per weapon and $50-100M all-in for for ICBMs, economic growth since then means the world could already afford far larger arsenals in a renewed arms race, current US military expenditure is over $700B annually, about 1/30th of GDP; at the peak of the Cold War in the 50s and 60s it was about 1/10th; Soviet expenditure was proportionally higher, so with 1950s proportional military expenditures, half going to nukes, the US and China could each produce 20,000+ ICBMs, each of which could be fitted with MIRVs and several warheads, building up to millions of warheads over a decade or so; the numbers could be higher for cheaper delivery systems, economies of scale and improvements in technology would likely bring down the per warhead cost, if AI and robotics greatly increase economic growth the above numbers could be increased by orders of magnitude, radiation effects could be intentionally greatly increased with alternative warhead composition, all-out discharge of strategic nuclear arsenals is also much more likely to be accompanied by simultaneous deployment of other WMD, including pandemic bioweapons (which the Soviets pursued as a strategic weapon for such circumstances)and drone swarms (which might kill survivors in bunkers); the combined effects of future versions of all of these WMD at once may synergistically cause extinction

PDF

First page of PDF