A number of people have claimed that a full-scale nuclear war is likely to cause human extinction. I have investigated this issue in depth and concluded that even a full scale nuclear exchange is unlikely (<1%) to cause human extinction. 

By a full-scale war, I mean a nuclear exchange between major world powers, such as the US, Russia, and China, using the complete arsenals of each country. The total number of warheads today (14,000) is significantly smaller than during the height of the cold war (70,000). While extinction from nuclear war is unlikely today, it may become more likely if significantly more warheads are deployed or if designs of weapons change significantly.

There are three potential mechanisms of human extinction from nuclear war:

1) Kinetic destruction

2) Radiation

3) Climate alteration

Only 3) is remotely plausible with existing weapons, but let's go through them all.

1) Kinetic destruction

There simply aren't enough nuclear warheads to kill everyone directly with kinetic force, and there likely never will be. There are ~14,000 nuclear weapons in the world, and let’s suppose they have an average yield of something like 1 megaton. This is a conservative guess, the actual average is probably closer to 100 kilotons. With a 1 megaton warhead, you can create a fireball covering 3 km², and a moderate pressure wave that knocks down most residential houses covering 155 km². The former kills nearly everyone and the latter kills a decent percentage of people but not everyone. Let's be conservative and assume the pressure wave kills everyone in its radius. 14,000 * 155 = 2.17 million km². The New York Metro area is 8,683 km². So all the nuclear weapons in the world could destroy about 250 New York Metro areas. This is a lot! But not near enough, even if someone intentionally tried to hit all the populations at once. Total land surface of earth is: 510.1 million km². Urban area, by one estimate, is about 2%, or 10.2 million km.² Since the total possible area destroyed from nuclear weapons is ~2.17 million km² is considerably less than a lower bound on the area of human habitation, 10.2 million km², there should be basically no risk of human extinction from kinetic destruction.

The circle with the white border indicates the zone of moderate blast damage radius (5 psi): 7.03 km (155 km²) from a 1,000 kiloton warhead, link to nukemap

If you want to check my work there, I was using nuke map

The even more obvious reason why kinetic damage wouldn't lead to human extinction is that nuclear states only threaten one or several countries at a time, and never the population centers of the entire world. Even if NATO countries and Russia and China all went to war at the same time, Africa, South America, and other neutral regions would be spared any kinetic damage. 

2) Radiation

Radiation won't kill everyone because there aren't enough weapons, and radiation from them would be concentrated in some areas and wholly absent from other areas. Even in the worst affected areas, lethal radiation from fallout would drop to survivable levels within weeks.

Here it's worth noting that there is an inherent tradeoff between length of halflife and energy released by radionuclides. The shorter the half life the more energy will be released, and the longer the half life the less energy. The fallout products from modern nuclear weapons are very lethal, but only for days to several weeks. 

From Nuclear War Survival Skills, 1987 edition

Let's try the same calculation we used with kinetic damage, and see if an attack aimed at optimizing fallout for killing everyone could succeed. Using Nukemap again, I'll go with the fallout contour for 100 rads per hour. 400 rads is thought too be enough to kill 50% of people, so 100 rads per hour is likely to kill most all people not in some kind of shelter. We need to switch to using a groundburst detonation rather than an airburst detonation, because groundbursts create far more fallout. A 1mt ground burst would create an area of about 8,000 km²  of >100 rads per hour. Okay, multiple that by 14,000 warheads, and we get 112 million km². That's a lot! It's still less than the  510.1 million km² of earth's land mass, but it's a lot more than the ~10.2 million km² of urban space. Presumably this is enough to cover every human habitation, so in principle, it might be possible to kill everyone with radiation from existing nuclear weapons.


The bright red and slightly less bright red indicate fallout contour for 1,000 rads and 100 rads per hour, covering 1,140 km² and 7,080 km² respectively, from a 1,000 kiloton ground burst. Nukemap settings

In practice, it would be almost impossible to kill every human via radiation with the existing nuclear arsenals, even if they were targeted explicitly for this purpose. The first reason is that fallout patterns are very uneven. After a ground burst, fallout is carried by the wind. Some areas will be hit bad and some areas will be hardly affected by fallout. Even if most human population centers were covered, a few areas would almost certainly escape.

Two other things make extinction by radiation unlikely. Many countries, especially in the southern hemisphere, are unlikely to be affected by fallout much at all. Since most of these countries are likely to be neutral in a conflict, and not near combatant countries, they should be relatively safe from fallout. While fallout might travel hundreds of kms, it still won't reach places separated by greater distances. Fallout that reaches the upper atmosphere will eventually fall back down, but usually after the period of lethal radioactivity.  The other mitigating factor is that in typical nuclear war plans, ground bursts are usually restricted to hardened targets, and air bursts are favored for population and industry centers. This is because air bursts maximize the size of the destructive pressure wave. Air burst detonations result in little lethal fallout reaching the ground, so populations not downwind of military targets would likely be safe from the worst of the radiological effects in a war scenario.

The final protection from extinction by radiation is simply large amounts of mass between people and the radiation source, in other words, fallout shelters. After several weeks, the radionuclides in fallout from ground burst detonations will have decayed to the point where humans can survive outside of shelters. Many fallout shelters exist in the world, and many more could be made easily in a day or two with a shovel, some ground, and some boards.  Even if lethally radioactive fallout from ground bursts covered all population centers, many humans would still survive in shelters.

The risks of extinction from nuclear-weapon-induced-radiation wouldn't be complete without discussing two factors: nuclear power plants and radiological weapons. I'm only going to cover these briefly, but they both don't change the conclusions much.

Nuclear power plants could be targeted by nuclear weapons to create large amounts of fallout with a longer half-life but less energy per unit time. The main concern here is that nuclear power plants and spent fuel sites contain a much greater *mass* of radioactive material than nuclear missiles can carry. The danger comes primarily from spreading the already very radiative spent or unspent nuclear fuel. The risk this poses requires a longer analysis, but the short version is that while nuking a nuclear power plant or stored fuel site would indeed create some pretty long-lived fallout it would still be concentrated in a relatively small area. Fortunately, even a nuclear detonation wouldn't spread the nuclear fuel more than several hundred km at most. Having regions of countries covered in spent nuclear fuel would be awful, but it doesn't much raise the risk of extinction.

Radiological weapons are nuclear weapons designed to maximize the spread of lethal fallout rather than destructive yield. The particular concern from the extinction perspective is that they can be designed to create fallout that continues to emit levels of radiation that can make an area uninhabitable for months to years. These kind of radiological weapons kill more slowly, but they still kill. In principle, radiological weapons could be used to kill everyone on earth. However, in practice, the same constraints that apply to standard nuclear weapons apply to weapons optimized for long-lasting fallout, as well as some additional constraints. 

Radiological weapons wouldn't produce more fallout than standard warheads, they would just produce fallout with different characteristics. As a result the amount of radiological weapons required to cover every part of earth's surface would be massively expensive (likely as expensive as the largest existing nuclear arsenals), and serve no military purpose. Their inefficiency in destruction and death compared to standard nuclear weapons is probably why radiological weapons have never been developed or deployed in large numbers. This makes them an ongoing theoretical concern, but not an existential risk in the immediate future. A concerning development is Russia's claim to have developed a large-yield (100mt) submersible nuclear weapon with the suggestion that it could be used as a radiological weapon, but even if this is true, it's unlikely to be deployed in large numbers.

3) Climate alteration

The bulk of the risk of human extinction from nuclear weapons come from risks of catastrophic climate change, nuclear winter, due to secondary effects from nuclear detonations. However, even in most full-scale nuclear exchange scenarios, the resulting climate effects are unlikely to cause human extinction.

Reasons for this:

a) Under scenarios where a severe nuclear winter occurs as described by Robock et al, some human populations would likely survive.
b) The Robock group’s models are probably overestimating the risk
c) Nuclear war planners are aware of nuclear winter risks and can incorporate these risks into their targeting plans

Before diving into each subject, it’s worth understanding the background of nuclear winter research. In the 1980s a group of atmospheric scientists proposed the hypothesis that a nuclear war would result in massive firestorms in burning cities, which would loft particles high into the atmosphere and cause catastrophic cooling that would last for years. Many found it alarming that such an effect could be possible and go unnoticed for decades while the risk existed. Some scientists also thought the proposed effect was too strong, or unlikely to occur at all. Until a few years ago, if you looked only at peer reviewed literature you would only find papers forecasting severe nuclear winter effects in the event of a nuclear war. Understandably, many people assumed that this was the scientific consensus. Unfortunately, this misrepresented the scientific community’s state of uncertainty about the risks of nuclear war. There have only ever been a small numbers of papers published about this topic (<15 probably), mostly from one group of researchers, despite the topic being one of existential importance.

I’m very glad Robock, Toon, and others have spent much of their careers studying nuclear winter effects, and their models are useful in estimating potential climate change caused by nuclear war. However, I’ve become less convinced over time the Robock model is largely correct. See section B below for why I’ve changed my mind. However, I’m quite uncertain about the probability of strong cooling effects from nuclear war, and am still quite concerned about the potential for severe cooling, even if the risk of extinction from such events is small.

A:  Under scenarios where a severe nuclear winter occurs as described by Robock et al, some human populations would likely survive.

The latest and most detailed model of potential cooling effects from a fullscale nuclear exchange comes from, Robock et al.,  “Nuclear winter revisited with a modern climate model and current nuclear arsenals: Still catastrophic consequences” found here.

The effects from this model are severe. In the 150Tg case, after a year, summer temperatures in the Northern hemisphere are 10-30 degrees C cooler. The effects are less severe at the equator (5 degrees C), but basically all places in the world are affected. The most likely outcome is that most people starve to death. Many would freeze too, but starvation is likely the greatest risk. Even in this model, it appears that in equatorial regions, some farming would still be possible, enough for some populations to survive. After a 10-15 years, agriculture in most of the world would be possible at reduced capacity.

Surface air temperature changes for the 150 Tg case averaged for June, July, and August of the year of smoke injection and the next year. Robock et al., 2007

Carl Shulman asked one of the authors of this paper, Luke Oman, his probability that the 150Tg nuclear winter scenario discussed in the paper would result in human extinction, the answer he gave was “in the range of 1 in 10,000 to 1 in 100,000.” This strikes me as quite plausible, though one expert opinion is no substitute for a deep analysis. The Q&A with Oman contains his reasoning for this assessment.

Two different analyses are required to calculate the chances of human extinction from nuclear winter. The first is the analysis of the climate change that could result from a nuclear war, and the second is the adaptive capacity of human groups to these climate changes. I have not seen an in depth analysis of the latter, but I believe such an assessment would be worthwhile.  

My own guess is that humans are capable of surviving far more severe climate shifts than those projected in nuclear winter scenarios. Humans are more robust than most any other mammal to drastic changes in temperature, as evidenced by our global range, even in pre-historic times. While a loss of most agriculture would likely kill most people on earth, modern technology would enable some populations to survive. Great stores of food currently exist in the world, and it is l likely that some of these would be seized and protected by small groups, providing enough food to last for years. While even such populations with food stores wouldn’t have enough to survive for 10-15 years, such food stores would give groups time to adapt to new food sources. The organization ALLFED has explored a number of alternative food sources that could keep populations alive in the event of a nuclear war or other large solar disruption, and I expect great necessity to drive the discovery of even more in the event of such a disaster.

B: The Robock group’s models are probably overestimating the risk

The nuclear winter model at its simplest: Nuclear detonations → Fires in cities → Firestorms in cities → Lofted black carbon into the upper atmosphere → black carbon persists in upper atmosphere, reflecting sunlight and causes massive cooling

Each step is required in order for the effect to occur. If nuclear war causes massive fires in cities but does not lead to firestorms that loft particles, then no long term cooling is going to occur. Some of these steps are easier to model than others. Based on my reading of the literature, the greatest uncertainties involve the dynamics of cities burning after a nuclear attack, and whether the conditions would produce firestorms sufficient to loft large numbers of particles high enough in the atmosphere to persist for years.

We’re finally beginning to see some healthy debate about some of these questions in the scientific literature. Alan Robock’s group published a paper in 2007 that found significant cooling effects even from a relatively limited regional war. A group from Los Alamos, Reisner et al, published a paper in 2018 that reexamined some of the assumptions that went into Robock et al’s model, and concluded that global cooling was unlikely in such a scenario. Robock et al. responded, and Reisner et al responded to the response. Both authors bring up good points, but I find Reisner’s position more compelling. This back and forth is worth reading for those who want to investigate deeper. Unfortunately Reisner’s group has not published an analysis on potential cooling effects from a modern full scale nuclear exchange, rather than a limited regional exchange. Even so, it’s not hard to extrapolate that Reisner’s model would result in far less cooling than Robock’s model in the equivalent situation.

C: Nuclear war planners are aware of nuclear winter risks and can incorporate these risks into their targeting plans

A very simple way to reduce risks from nuclear winter is to refrain from targeting cities with nuclear weapons. The proposed mechanism behind nuclear winter results from cities burning, not ground bursts on military targets. I’ve spoken with some of the officials in the US defense establishment responsible for nuclear war planning, and they’re well aware of the potential risks from nuclear winter. Of course, being aware of the risks does not guarantee they will have reasoned about the risks well, or have engaged in good risk management practices. However, the fact that this risk is well publicized makes it more likely that nuclear war planners will take steps to minimize blowback risk from climate effects.

It’s hard to know to what extent this has been done. Nuclear war plans are classified, and as far as we know current US nuclear war plans do target cities under some circumstances but not under others. However, the defense establishment has access to classified information and models that we civilians do not have, in addition to all the public material. I’m confident that some nuclear war planners have thought deeply about the risks of climate change from nuclear war, even though I don’t know their conclusions or bureaucratic constraints. All else being equal, the knowledge of these risks makes military planners less likely to accidentally cause human extinction.


This post discussed the three plausible mechanisms of human extinction caused by nuclear weapons. The fact that one of these mechanisms, nuclear winter, wasn’t characterized until the 1980s, is a good reminder of the possibility of unknown unknowns. While nuclear tests provided information about the effects of these weapons, the test environments were significantly different than war environments. Large model uncertainties remain. Given that the greatest existential threat from nuclear war appears to be from climate impacts, it would be great to see more researchers study the climate effects from nuclear war and the resilience capacity of different human groups.

There appear to be several interventions possible for reducing existential risk from nuclear war. At the policy level, a commitment from the largest nuclear powers to refrain from targeting the majority of cities would reduce risk of accidental omnicide. Improving the maximum resilience capacity of human populations best positioned to survive a nuclear winter would also make humanity less vulnerable to nuclear winter, and could also protect against other existential threats.

Further reading

Toby Ord conducts a quantitative estimate of extinction risk from nuclear war in:
The Precipice: Existential Risk and the Future of Humanity

Nuclear War as a Global Catastrophic Risk

Nuclear winter and human extinction: Q&A with Luke Oman (by Carl Shulman)


Nuclear Winter Responses to Nuclear War Between the United States and Russia in the Whole Atmosphere Community Climate Model Version 4 and the Goddard Institute for Space Studies ModelE


Climate Impact of a Regional Nuclear Weapons Exchange: An Improved Assessment Based On Detailed Source Calculations


Comment on “Climate Impact of a Regional Nuclear Weapon Exchange: An Improved Assessment Based on Detailed Source Calculations” by Reisner et al.


Reply to Comment by Robock et al. on “Climate Impact of a Regional Nuclear Weapon Exchange: An Improved Assessment Based on Detailed Source Calculations”


Comparing Economic and Crop Models: The Case of Climatic and Agricultural Impacts of Nuclear War


New Comment
47 comments, sorted by Click to highlight new comments since:

When I've previously considered human extinction caused by nuclear wars, I've known that the immediate blasts wouldn't kill everyone. However, what are the effects of a lower overall population with fewer habitable areas and less access to resources? That's doubly true since the areas that will have more people survive will almost definitionally be developing countries that are now suddenly cutoff from imports. I believe that humans as a species would likely survive, but I also suspect that it would be the end of modern civilization. Adding to that, I've seen hypotheses before that the remaining resources left underground but easily accessibly by non-modern technology would not be enough to "reboot" civilization, especially fossil fuels. Overall, despite the very likely non-extinction of the human species, it would much more likely be an extinction of the human race as a space-faring species.

I really don't think fossil fuel depletion is very likely to permanently curtail humanity's potential in a nuclear or other collapse situation. I've seen this point argued a bunch though, so I think it's worth taking the hypothesis seriously. I'd love to see an in depth analysis of this question.

I appreciate how Toby Ord considers "knock-on effects" in his modelling of existential risks, as presented in "The Precipice". A catastrophe doesn't necessarily have to cause extinction for it to be considered an existential threat. The reason being knock-on effects, which would undoubtedly impair our preparedness for what comes next.

Is this isomorphic to the framing of existential risk as one category? (Which is not something I came up with, I just distilled the idea.) It still seems to me like the idea that first- and second-order effects are qualitatively different is just a mistake born of the "risk from xxx" framing.

Yes, it looks isomorphic. Thanks for sharing your write-up. You've captured this idea well.

Also consider whether a "second-try/rebooted" human civilization would be more or less probable to eventually succeed at space-faring / AGI construction. 

Speculating about this is a fun exercise. I argue that the answer is less probable.

The survivors might have a more substantial commitment to life affirmation given that the fragility of life is so fresh in their minds following Armageddon. I argue that this would have a minimal effect. We know that the dinosaurs went extinct, and we know that the average lifespan for a mammalian species is about one million years. We know that we have fought world wars, and we know that life is precious and unreasonably rare in the universe. Yet, in aggregate, we still don't care that much about safeguarding the future of life. Existential risk reduction doesn't factor heavily into our decision making and how we allocate our resources.

In terms of technological progress, I think a post-nuclear war civilisation would likely quickly bootstrap itself, given the abundance of already extracted raw materials and knowledge left behind. However, on the whole, Armageddon would be a severe hindrance to technological progress. Political, social and economic stability appears advantageous to the creation of knowledge.

One thing that seems important to note: nuclear warfare need not occur in a vacuum. If countries possessing nuclear weapons are trading all-out strikes, as in your model, they probably are in a state of (World?) war already, and either have fought with other weapons prior to the nuclear exchange, or plan to continue to do so after it. This may include use of non-nuclear weapons with high collateral damage, like chemical or biological agents, or saturation bombardment targeting high-population areas. I wonder if that skews the assessment of damage in any meaningful way.

Yeah, the point that risks from nuclear war would be coupled with risks from great power conflict is a good one. I expect this to be more of a problem in the future, but there could be some risks at present from secret bioweapon systems or other kinds of WMDs. 

My mainline expectation is that in a nuclear war scenario, chemical, biological, and conventional weapon effects would be dwarfed by the effects from nuclear weapons. This is based on my understanding of the major powers deterrence strategy, but might be wrong if there are secret weapons I'm not aware of. The logic of deterrence makes this a little less likely, fortunately. The whole point of a deterrent is lost if you keep it a secret. Of course it's possible that it's kept secret from the public but not from other countries, but this seems harder to keep secret, especially since it relies on one's potential enemies to keep it secret.


My mainline expectation is that in a nuclear war scenario, chemical, biological, and conventional weapon effects would be dwarfed by the effects from nuclear weapons.

I would classify biological weapons as more dangerous than nuclear, but that's a different topic. Besides, biological and nuclear warfare don't mix well - without commercial air travel and trade biological agents don't spread well.

This will not be a full review—it's more of a drive-by comment which I think is relevant to the review process.

However, the defense establishment has access to classified information and models that we civilians do not have, in addition to all the public material. I’m confident that nuclear war planners have thought deeply about the risks of climate change from nuclear war, even though I don’t know their conclusions or bureaucratic constraints.

I am extremely skeptical of and am not at all confident in this conclusion. Ellsberg's The Doomsday Machine describes a horribly incentivized military establishment which pursued bloodthirsty and senseless policies, deceiving their superiors (including several presidents), breaking authentication protocols, refusing to adopt plans which didn't senselessly destroy China in a conflict in the Soviet Union, sub-delegation of nuclear launch authority to theater commanders and their subordinates (no, it's not operationally true that the US president has to authorize an attack!), lack of controls against false alarms, and constant presidential threats of first-use. The USAF would manipulate presidential officials in order to secure funding, via tactics such as inflating threat estimates or ignoring evidence that the Soviet Union had less nuclear might than initially thought.

And Ellsberg stated that he didn't think much had changed since his tenure in the 50s-70s. While individual planners might be aware of the nuclear winter risks, the overall US military establishment seems insane to me around nuclear policy—and what of those in other nuclear powers? 

However, The Doomsday Machine is my only exposure to these considerations, and perhaps I'm missing a broader perspective. If so, I think that case should be more clearly spelled out, because as far as I can tell, nuclear policy seems like yet another depravedly inadequate facet of our current civilization. 

I'm fairly confident that at least some nuclear war planners have thought deeply about the risks of climate change from nuclear war because I've talked to a researcher at RAND who basically told me as much, plus the group at Los Alamos who published papers about it, both of which seem like strong evidence that some nuclear war planners have taken it seriously. Reisner et al., "Climate Impact of a Regional Nuclear Weapons Exchange: An Improved Assessment Based On Detailed Source Calculations" is mostly Los Alamos scientists I believe.

Just because some of these researchers & nuclear war planners have thought deeply about it doesn't mean nuclear policy will end up being sane and factoring in the risks. But I think it provides some evidence in that direction.

I've edited the original to add "some" so it reads "I'm confident that some nuclear war planners have..."

It wouldn't surprise me if some nuclear war planners had dismissed these risks while others had thought them important.

The post claims:

I have investigated this issue in depth and concluded that even a full scale nuclear exchange is unlikely (<1%) to cause human extinction.

This review aims to assess whether having read the post I can conclude the same.

The review is split into 3 parts:

  • Epistemic spot check
  • Examining the argument
  • Outside the argument

Epistemic spot check

Claim: There are 14,000 nuclear warheads in the world.

Assessment: True

Claim: Average warhead yield <1 Mt, probably closer to 100kt

Assessment: Probably true, possibly misleading. Values I found were:

The original claim read to me that 100kT was probably pretty close and 1Mt was a big factor of safety (~x10) but whereas the safety factor was actually less than that (~x3). However that’s the advantage of having a safety factor – even if it’s a bit misleading there still is a safety factor in the calculations.

I found the lack of links slightly frustrating here – it would have been nice to see where the OP got the numbers from.

Examining the argument

The argument itself can be summarized as:

  1. Kinetic destruction can’t be big enough
  2. Radiation could theoretically be enough but in practice wouldn’t be
  3. Nuclear winter not sufficient to cause extinction

One assumption in the arguments for 1 & 2 is that the important factor is the average warhead yield and that e.g. a 10Mt warhead doesn’t have an outsized effect. This seems likely and a comment suggests that going over 500kt doesn’t make as much difference as might be thought and that is why warheads are the size that they are.

Arguments 1 & 2 seem very solid. We have done enough tests that our understanding of kinetic destruction is likely to be fairly good so I don’t have much concerns there. Similarly, radiation is well understood and dispersal patterns seem kinda predictable in principle and even if these are wrong the total amount of radiation doesn't change, just the where it is.

Climate change is less easy to model, especially given that the scenario is a long way out of our actual experience. To trust the conclusion we have to trust the models. I haven’t looked into this – different researchers have come to different conclusions and we hope that the true value is between the worst and best case found (or at least not too much worse than the worst case). 

One thing that I found frustrating in the post is that the <1% risk of human extinction from full scale nuclear war is given but the post is not explicit which arguments give such high confidence. The post rightly refers to a number of steps which need to happen to make catastrophic climate change happen:

  1. Given a nuclear exchange how probable are firestorms?
  2. Given firestorms how probable is black carbon to be lifted into the atmosphere?
  3. Given lofting, how probable is it to stay aloft for sufficient time?
  4. Given long term cooling, how probable is human extinction?

But it is not explicit as to which step(s) are most improbable.

The OP points towards an exchange between Robock and Reisner (could a small scale conflict cause a nuclear winter?) where it finds Reisner’s position more compelling. However, finding the position more compelling does not suggest to me that this part of the logic chain can be doing much heavy lifting.

(similarly, I don’t think military planners being aware of the danger of nuclear winter can be doing much heavy lifting to get us up to <1% probability).

So the main part of the logic to conclude high confidence must be that even catastrophic climate change would not cause human extinction.

The OP quotes Luke Oman as saying he believed that it would be ~1 in 10,000 to 1 in 100,000. It also links to him giving his reasoning. This is based mainly on the Toba supervolcano and other similar events failing to wipe out humanity’s ancestors and that the southern hemisphere is likely to be less badly hit than the north in a nuclear exchange. (I would have liked this to have been summarized in the OP as it is probably the strongest argument for non-extinction). 

The interview doesn’t go into enough detail for me to assess how strong Luke’s arguments are and this makes it hard for me to get the same level of confidence that Luke has – how close an analogy to full scale nuclear war is Toba likely to be? 

The post also suggests food stores and human ingenuity as reasons to think that humanity would avoid extinction.

Given how key this part of the logic chain seems to be I would have liked this section to be more detailed.

Outside the argument

I would have also liked the post to consider uncertainty outside the argument. 

A few comments point out other possible mechanisms which might cause additional risk:

Giving <1% probability that full scale nuclear war will cause human extinction implies sufficient confidence that we have thought of all the possible causes of extinction and/or that other possible causes will have similarly low probability of extinction.

This isn’t discussed in the OP which seems like a weakness.


I think that the post (plus the link to the Luke Oman interview) give sufficient evidence – a full scale nuclear exchange is unlikely (<1%) to cause human extinction.

If I wanted to hone in my probability further the post helpfully supplies further reading which is a great feature.

The main subject where I would have liked to see more detail is what I see as the crux of the matter with climate change – given extremely large temperature changes how likely is humanity to go extinct? The arguments in the post itself wouldn’t have been enough for me – it was only reading the Luke Oman interview that convinced me.

Overall I think the post is strong and think it is a good candidate for the 2020 review.

Good point re Australia hosting missile tracking capabilities, I agree that it might be targeted given that. I'm less worried about carrier groups and such things being hit. I don't disagree that some of these might be hit, and this may result in some fallout in the southern hemisphere, it doesn't seem like enough to move the dial. The ocean has a lot of area.

I didn't cover sterilization or birth defects from either the initial fallout or from ingested radionuclides later on. These are both problems, but I would be quite surprised if they killed a large percentage of people's children or grandchildren in places not near places hit with a nuclear attack. If you have a source for a good assessment of this, I'd be quite interested. The best I have seen is in On Thermonuclear War and we know a lot more about radiation than we did back then. 

In a future post I want to discuss how long term effects from various technologies, disasters, weapon systems, etc. could combine to lower the habitability of earth. I certainly think nuclear war could play a big role in this, but not cause extinction without other significant factors.

I don't think I agree that nuclear war planners would use nuclear winter effects offensively. A lot of effort in nuclear war planning goes into counterforce targeting, which is explicitly to reduce the effectiveness of the enemy's retaliatory (or reserve / 3rd wave) nuclear strike. This suggests some priority to limiting damage to one's own side. My impression is that the "mutual suicide pact" approach to deterrence is less popular than it used to be.

"Something that I hadn't considered before: would it be possible to move people into target areas (before attacks) or radiated areas (afterwards) by using conventional and/or area denial weapons?"

I don't think so. Generally if you want to increase casualties you would want to have people concentrated as much as possible, so move people into already large cities. However, people during wartime (and pandemics) usually tend to move out from such places, this is shown both by historical experience and to me seems to be the logical way to act (as cities are targeted due to critical infrastructure they contain and most services cities offer become severely limited).

Even if the countryside were targeted deliberately for this effect, conventional weapons cannot be used efficiently for this kind of area denial, for such nuclear weapons seem to be the most effective, maybe alongside with chemical weapons, but those have the same limitation (fallout directed by weather conditions, wind in particular) with far less power.

Promoted to curated: I think this post makes an important point, and is also a good illustration of a type of thinking I care a lot about. I think it's a pretty common misconception, even among a good chunk of people working on existential risk, that nuclear war has the potential to directly eradicate humanity. This seems important to clear up, and I think the way the post does it is pretty good. I also think the general principle of "stare into the darkness and distinguish between bad and very-bad outcomes" is very important, and one that I can imagine being fruitfully applied to many other domains.

This post feels quite important from a global priorities standpoint. Nuclear war mitigation might have been one of the top priorities for humanity (and to be clear it's still plausibly quite important). But given that the longtermist community has limited resources, it matters a lot whether something falls in the top 5-10 priorities. 

A lot of people ask "Why is there so much focus on AI in the longtermist community? What about other x-risks like nuclear?". And I think it's an important, counterintuitive answer that nuclear war probably isn't an x-risk. 

Like Jeff and Bucky, I still think it's worth someone following up and investigating the phenomenon here in more detail. It's disappointing that humanity hasn't studied this problem in as much depth as we could have.

Thanks for writing this up!  I think having more well-researched and well-written dives into things like this are great

A bunch of scattered thoughts and replies to these:

Overall I agree with the central idea (Nuclear War is unlikely to cause human extinction), but I disagree enough with the reasoning to want to hammer it into better shape.

Writing this I feel like in a "editing/constructive feedback" mood, but I'd welcome you to throw it all out if its not what you're going for.  To the feedback!

This seems to only consider current known nuclear weapons arsenals.  It seems worth including probabilities that different kinds of weapons are built before such a war.  In particular, longer-lived species of bombs (e.g. salt bombs, cobalt bombs, etc)

I think I want to separate "kill everyone with acute radiation right away" and "kill everyone with radiation in all of the food/water", and the latter seems less addressed by the energy/half life argument.  I think the weapon design space here is pretty huge, so ruling these out seems hard to me.  (Though I do think if we're assigning probabilities, they should get lower probabilities than conventional weapons)

In general I would prefer approximate numbers or at least likelihood ratios for what you think this evidence balances out to, and what likelihood of odds you would put on different outcomes.

(For example: "what is the likelihood ratio of the 3.C evidence that nuclear war planners are familiar with ideas like nuclear winter" -- I don't think these are strictly required, but they really help me contextualize and integrate this information)

In particular, Toby Ord gives a bunch of excellent quantitative analysis of X-risks, including nuclear war risk, in The Precipice.

(In fact, if your main point of the post was to present a different model from that one, adding numbers would greatly help in comparing and contrasting and integrating them)

Finally, I think my mental models of case 3) are basically the same as any event that is a significant change to the biosphere -- and it seems reasoning about this gets harder given your premise.

A hypothetical: if there are 3 major climate events in the next 100 years (of which one is a bellicose nuclear exchange), and humanity goes extinct due to climate related symptoms, does the nuclear war "cause" the human extinction in a way you're trying to capture?

Maybe what I want is for the premise to be more precise: define a time limit (extinct within X years) and maybe factor what it means to "cause" (for example, it seems like this suggests that an economic collapse triggered by an nuclear war, which triggers other things, that eventually leads to extinction, is not as clearly "caused by nuclear war")

Also maybe define a bit what "full-scale" means?  I assume that it means total war (as opposed to limited war), but good to clear up in any case.

That's all that came to mind for now.  Thanks again for sharing!~

Thanks Alex! Yeah, I agree with you that adding approximate numbers or likelihood ratios would improve this, as would comparing my credences with Toby Ord's. I might do a followup post with some of this if I get time. Originally I was going to find a co-author and go in more depth on some of these things, especially the nuclear winter literature, but I keep starting and not finishing posts and I figured it was finally time to just put up what I had.

It would be good to separate "kill everyone with acute radiation right away" and "kill everyone with radiation in all of the food/water". I discussed risk of the first one and basically didn't at all cover risk of the second one, but I'd like to see a better assessment of the second. I've never found good sources for these kinds of long term effects of radiation from food and water after a nuclear war, despite spending probably 4-8 hours searching. On Thermonuclear War discusses this in depth but it's very out of date (1960), and I haven't found anything like a comprehensive analysis of this anywhere else. Lots of speculation here and there but nothing that looks rigorous.

Thanks for your feedback!


One thing to consider that doesn't seem to be covered is that nuclear war planning is focused on a diplomatic outcome, not genocide. The fastest way to a diplomatic resolution is through economic means. As a result, it seems reasonable to conclude any nuclear attack will target the productive cities and regions of the target country necessary to cripple their economic system. The US Homeland Security agency has a map of these locations produced some 30 years ago by the military that is still largely valid. The locations most likely targeted coincide with but do not exactly correspond to population centers. In the US, they include Atlanta(trade and shipping) Kansas City(production). Cities that produce more customers and consumers than producers are of more value alive, as their now unmet needs will pressure favorable decisions. Targeting of military facilities will also be limited to focusing as much allocatable firepower on suppressing the retaliatory strike potential of the enemy. At the final negotiating table, ones position is as much dictated by ones own productive assets as they are by the opponents losses. A nation whose productive capacity remains intact is in a favorable position militarily as they can afford to subsidize and fabricate more resources for any extended war.

So, given that any weapons used will be likely divided between destroying production and protecting production, the vast majority of people, especially those in rural environments and smaller cities, will likely not be at direct risk. And since the vast majority of the industrial regions are made of concrete and steel and other non-flammable materials, it also seems less likely there would be any substantial firestorms.

When modeling the total destructive potential of any arsenal, i also would presume that any combatant is likely to minimize their usage of warheads so that only a safe level of redudancy is employed, and a large stockpile remains on hand as a strategic deterrant to third parties not part of the initial exchange. If america, say, were to be in a war with either russia or china, they would likely retain enough warheads for a MAD scenario involving the other nation, and russia would retain warheads for europe, and china for india.

The same may not be true of "salt the earth" type genocidal wars as are possible in nuclear exchanges among the smaller countries where hostility is more about fundemental differences in ideology and race..

I think what you're saying makes sense, but it's not clear to me how prominent diplomatic objectives of nuclear war are in current nuclear war plans. I wrote on this some here: https://www.lesswrong.com/posts/rn2duwRP2pqvLqGCE/does-the-us-nuclear-policy-still-target-cities If you have sources for this I'd be interested!

Re: neutral countries not getting targeted; I've heard it claimed that some nuclear targeting plans involved hitting even neutral countries, on the assumption that anyone who survived a nuclear war unscathed would become the next major power so better to ensure that everyone goes down. I have no idea of whether this had a credible source, though; do we know anything about whether this might be true?

Huh, I haven't heard this. Or, rather this was definitely the case early in the cold war re China, a good account of which is in Daniel Ellsberg's the Doomsday Machine. The US war plans considered China hostile even though it wasn't closely aligned with the Soviet Union, and planned to nuke its major cities in the event of a US-Soviet war. I would expect nuclear war plans to sometimes include military allies of target countries, but usually neutral countries. Though I'd be very interested to see a source to the contrary!

I appreciate the Bottom Line Up Front writing style. Not only overall, but also in each subsection. Thank you!

"My own guess is that humans are capable of surviving far more severe climate shifts than those projected in nuclear winter scenarios. Humans are more robust than most any other mammal to drastic changes in temperature, as evidenced by our global range, even in pre-historic times"

I think it is worth noting that the speed of climate shifts might play an important role, as a lot of human adaptability seems to rely on gradual cultural evolution. While modern information technology has greatly sped up the potential for cultural evolution, I am unsure if these speedups are robust to a full-scale nuclear war.

Have you considered if human extinction could be brought about with something like cobalt bombs? https://en.wikipedia.org/wiki/Cobalt_bomb

As far as I know these aren't in anyone's arsenal. But if you have a nuke, making a cobalt bomb is trivial.

I asked a question about this a while ago: 


Overall my sense is that cobalt bombs are not practical. See also this discussion by Anders Sandberg and Carl Shulman 8 years ago: 


They seem similarly skeptical about cobalt bombs.

The idea that nuclear war = human extinction is still a good deterrent if most people believe it. We're all probably better off overestimating the dangers of these weapons as it will likely decrease the likelihood of an all our nuclear exchange. One that may not be an extinction event but would still devastate humanity. I can see the viral Facebook news story headlines now. "Scientists Say Total Nuclear War Not As Bad As Previously Thought," followed by "<Global Leader> Putting Nukes on High Alert."

I also have a hypothesis I've been testing by polling my friends. Most think they would survive an apocalyptic global event that kills 1 in 3 people. We're all a little optimistic about our own survivability, which doesn't help us avoid global nuclear catastrophes.

As the other commenter have been saying, excellent post.

There is an additional reason to believe, at least given contemporary capabilities and strategies, that the X-risk of an actual nuclear conflict is small. A few years ago I wrote to Fred Kaplan, the author of the stellar military history book "The Wizards of Armageddon"*, a history of US nuclear war planning from 1945-1990. I asked Kaplan what he judged the present state of nuclear war planning was. He responded to me that his sources informed him that nuclear war plans, in the US and presumably the Russian Federation, had been shaped by the same changes that shaped conventional war strategy from the Gulf War onward. The focus, in both nuclear and conventional war, is blinding and decapitation of the other side - destroying their C3I (command-control-communication-intelligence) infrastructure and killing their national command authority. The idea is to render an enemy unable to communicate with and deploy their nuclear forces, rendering them inert. One of the more likely outcomes of a nuclear conflict is the two nations being leaderless but largely intact, with only a few dozen low yield devices (<1 MT) having been used, and the rest stranded and unusable. The problem is that, as documented by Daniel Ellsberg in "The Doomsday Machine: Confessions of a Nuclear War Planner", many countries likely hedge against this strategy by pre-delegation, telling the commanders of their nuclear forces that they may, if contact is lost with civilian authority during a crisis, use their nuclear weapons at their own discretion. But even that is, for the reasons listed above, unlikely to yield an X-Risk scenario. A really, really, really awful situation to live through, but not an X-Risk.

*which I highly recommend. Kaplan is critical of the nuclear war planners, but I think most of the X-risk people on this forum and in academia would have fit right in at RAND and other strategic think tanks during the Cold War.

Well, genetic survival is different than the species thriving as well we all hope our children will as well.  Nuclear war would be a big setback.  Let's say that even though MAD is a message for the masses instead of an actual fact, it is effectively true for the audience the message was created for.

If a nuclear war happened, say, in 1983, what percent of people would die?

The estimates I trust the most said about 50% of Americans and 55% of Soviets would have died in the first 30 days if the Americans and Soviets hit each other with everything they've got. (The Soviets were more concentrated in cities; automobiles with which to flee the cities quickly with supplies were much scarcer in the USSR.)

Note that since 1983 the two sides have gotten rid of most of their nuclear weapons.

"30 days": most people who are going to die from radiation sickness will die within 30 days of the exposure.

How many would have starved in subsequent months and years because of lack of food is harder to estimate.

I think that this analysis is based on idea that nuclear war will be "conventional nuclear war", like it was envisioned in the middle of 20 century: that is, a nuclear exchange between two super powers via nuclear missiles.  However, unconventional nuclear war is also possible because of changes in strategy and-or technology.

The main technological changes:

  1. Very large and salted weapons. Teller worked on 10 Gigatonn bomb. Khan wrote about stationary very large nuclear weapon which is covered with cobalt and intended to produce large radioactivity around the world - doomsday weapon - as a mean of universal defence.  Russians created now nuclear torpedo with cobalt-salted warhead (Poseidon). Nukes also could be used against nuclear power stations which will create very large amount of radioactivity in the air. To evaporate a nuclear power plant, 1 Mt bomb is need.
  2. Very cheap weapons. If cold fusion works, home made fusion bombs could become a possibility. If any terrorsit can create them, there will be much more nuclear explosions in case of global guerilla. 
  3. Unconventional use of nuclear weapons to affect weather. There is an often discussed idea to use nukes to trigger supervolcano, eg, Yellowstone - and US adversaries may try to target the caldera with multiple warheads. 

Strategy changes:

  1. Nuclear blackmail via Doomsday weapons
  2. Nuclear guerrilla - many small states use nukes often.
  3. Geophysical attacks - attempts to cause natural disaster via nuclear weapons: supervolcanos, asteroids deflection to Earth, forrest fires. 

Yeah, you're right I'm making an assumption that a "nuclear war" refers to a nuclear war scenario with current arsenals or those in the near future.

1) Future nuclear weapons, especially if they're designed to kill everyone, could greatly increase the risk. Poseidon / Status-6 aside, I don't think states are likely to invest in omnicide capabilities, for several reasons. One is that it's a really hard optimization problem, and it's easier to be able to just crush your enemy with standard hydrogen bombs. So why pay far more for something that doesn't provide much additional deterrence capability? The other is that it's staggeringly unethical, in such an obvious way that people in most cultures are going to shy away from it. I can see how someone could justify standard nuclear deterrence. But a doomsday device would be bargaining away literally the entire world and future for one's own country, and even then it's a poor bargain. There are some who would contemplate this, but I think their number is relatively small. The fact that such ideas have so far not caught on in a serious way in any nuclear power that we're aware of is a good sign.

2) Very cheap weapons would be bad, agreed. I still think it would be hard to kill everyone with them, but it would at least increase the risk. And if they were continuously build and used, I could imagine longterm effects building up to greater systemic extinction risk.

3) Yeah, these seem like under-explored risks as well, though they don't appear to have much deterrence value so I don't expect military establishments to take them very seriously.

Thanks for your reply. One thing which is in play here is that Doomsday and geophysical weapons is the last resort of weakest side. If a stronger side has effective anti-missle tech and-or first strike capability, than having nuclear misseles becomes useless. This is a situation for Russia now. This is the reason why they are building Poseidon. 

Giving the mindset, a county like North Korea may invest in Doomsday weapon, but not a western country. Russia and China also could do it.


There is no scientific basis for 2 and 3

There are two scientifically proven ways to cheaper nukes: proliferation via laser isotope separation and the use of reactor plutonium for nukes

The use of nukes for artificial nuclear winter via nuclear explosions in taiga was also discussed (can't find the link now).

Can you clarify why the volcano triggering scheme in 3 would not be effective? It's not obvious. The scheme sounds rather lethal.


Not even the right order of magnitude. Yellowstone magma chamber is 5km beneath the surface. If you had a nuke large enough to set off a supervolcano, you wouldn't need to set off a supervolcano. Not to mention Yellowstone isn't ready to blow anyway.

Thanks! Useful info.

Humankind has long known how to produce very large nukes: the largest bomb ever tested (Tsar Bomba, 50 megatons) was tested in 1961. Why then is the maximum yield of every nuke currently in the US inventory under about .5 megatons? Because explosions bigger than that do not produce more destruction on the ground: after about .5 megatons, as the size (energy) of the explosion increases, it just lifts more and more of the Earth's atmosphere into space (but not enough to cause any danger to life on Earth).

Large bomb of gigaton scale could be useful if one wants disperse large amount of radioactivity over whole surface of the planet. In that case, lifting large amount of exhaust in the upper atmosphere will help radioactive elements to be dispersed over all surface of the Earth. This is needed for doomsday bomb, envisioned by Khan. Such bomb is ultimate defence weapon: no one will dare to attack country if it has one. 

Also, Russian Poseidon nuclear torpedo was said to be equipped with 100 MT bombs, intended to create tsunami. 

Such a weapon would actually be completely useless for deterrence. Imagine a nuclear attack were launched against your country killing half the population (immediately and through fallout). The survivors are then going to activate their doomsday weapon, ensuring they all die as well, just so they can get revenge? I find that incredibly implausible.

Using conventional nuclear weapons as a deterrent posts many of the same challenges due to the threat of retaliation. For example, say some foreign nation attacks NYC with non-nuclear weapons, or perhaps hits it with a single nuclear weapon, and threatens nuclear destruction of the entire US if the US retaliates in any way. Then what happens?

I think the idea that constructing weapons of mass destruction makes us safer is quite a dangerous one and I’m worried that someone might come away from reading a post like this with the impression that there’s nothing to worry about since we’d only manage to kill perhaps 95% or 99% of the world’s population and not 100%. Building and stockpiling thousands of hydrogen bombs should seem just as abhorrent to people as building a doomsday weapon that serves no functional purpose other than to enable us to kill all life on Earth.

The idea of the Doomsday weapon as it was envision by Kahn is that it will be activated automatically and can't be turned off by survivors - and it is well known fact for all players.