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 

I agree that nuclear war - and even nuclear winter - would be very unlikely to directly cause human extinction. My loose impression is that other EAs who have looked into this agree as well.

However, I'm not sure if it's good to publicize work on existential risk from nuclear war under this headline, and with this scope. Here is why:

• You only discuss whether nuclear war would somewhat directly cause human extinction - i.e. by either immediately killing everyone, or causing everyone to starve within, say, the next 20 years. However, you don't discuss whether nuclear war could cause a trajectory change of human civilization that make it more vulnerable to future existential risk. For example, if nuclear war would cause an irrecoverable loss of post-industrial levels of technology that would arguably constitute an existential catastrophe itself (by basically removing the chance of close-to-optimal futures) and also make humanity more vulnerable to natural extinction risk (e.g. they can no longer do asteroid deflection). FWIW, I think the example I just gave is fairly unlikely as well; my point here just is that your post doesn't tell us anything about such considerations. It would be entirely consistent with all evidence you present to think that nuclear war is a major indirect existential risk (in the sense just discussed).
• For this reason, I in particular disagree "that the greatest existential threat from nuclear war appears to be from climate impacts" (as you say in the conclusion). I think that in fact the possibly greatest existential threat from nuclear war is a negative trajectory change precipitated by 'the collapse of civilization', though we don't really know how likely that is or whether this would in fact be negative on extremely long timescales.
• (Note I'm not intending for this to just be a special case of the true but somewhat vacuous general claim that, for all we know, literally any event could cause a negative or positive trajectory change. The point is that the unprecedented damage caused by large-scale nuclear war seems unusually likely to cause a trajectory change in either direction.)
• [Less important:] I'm somewhat less optimistic than you about point 3C, i.e. nuclear war planners being aware of nuclear winter. I agree they are aware of the risk. However, I'm not sure if they have incentives to care. They might not care if they view "large-scale nuclear war that causes all our major cities to be destroyed" or "nuclear war that leads to a total defeat by an adversary" as essentially the worst possible outcomes, which seems at least plausible to me. Certainly I think they won't care about the risk as much as a typical longtermist - from an impartial perspective, even a, say, 1% risk of nuclear winter would be very concerning, whereas it could plausibly be a minor consideration when planning nuclear war from a more parochial perspective. Perhaps even more importantly, even if they did care as much as a longtermist, it's not clear to me if the strategic dynamics allow them to adjust their policies. For example, a nuclear war planner may well think that only a 'countervalue' strategy of targeting adversaries' population centers has a sufficient deterrence effect.

So overall I think our epistemic situation is: We know that one type of existential risk from nuclear war is very small, but we don't really have a good idea for how large total existential risk from nuclear war is. It's of course fine, and often a good idea for tractability or presentation reasons, to focus on only one aspect of a problem. But given this epistemic situation, I think the cost of spreading a message that can easily be rounded off to "nuclear war isn't that dangerous [from a longtermist perspective]" are high, particularly since perceptions that nuclear war would be extremely bad may be partly causally responsible for the fact that we haven't yet seen one.

Note I'm not claiming that this post by itself has large negative consequences. No nuclear power is going to chance their policies because of an EA Forum post. But I'd be concerned if there was a growing body of EA work with a messaging like this. For future public work I'd feel better if the summary was more like "nuclear war wouldn't kill every last human within a few decades, but is still extremely concerning from both a long-termist and present-generation perspective" + some constructive implications (e.g. perhaps focus more on how to make post-collapse recovery more likely or to go well).

It seems to me that this post fills an important void. Several commenters have said that they fear that letting people know that nuclear war wouldn't mean the end of the world would encourage them to take nuclear war less seriously.

First, let me say that I've spent forty years working to try to reduce the danger of nuclear war and (as I've gotten older) to eliminate nuclear weapons. I've received outside funding during only six of those years. Not asking for sympathy, I just want you to realize that I am quite serious about this stuff.

Partly because of that seriousness, I think it is essential to foreground reality in all arguments against nuclear weapons. In fact, I think it is the key to elimination.

I've spent some time studying attitudes toward nuclear war and I think landfish is exactly right that people often have unrealistic ideas about it. Mostly they exaggerate its consequences. I believe these exaggerations are far more harmful to attempts to reduce or eliminate nuclear weapons than any realistic explanation of the reality might be.

Let me explain.

For the most part, ordinary people, when they don't use the exact words "nuclear war" refer to nuclear war as "the end of the world" or "apocalypse" or "Armageddon." These terms are colorful, exaggerated, and unrealistic. They are, in some ways, fascinating anthropological evidence. Two of these terms are religious terms (apocalypse and Armageddon). Why would it make sense to refer to a twenty-first century military phenomenon (nuclear war), in religious or mythical terms?

Why would millions of people use an apocalyptic story dating back thousands of years to refer to nuclear war? The short answer is that I think in some way apocalypse is a narrative that they are familiar with. They take a strange and uncomfortable phenomenon, like nuclear war, and they make it familiar and less threatening by covering it with a story they know.

But there is an important point about apocalypse stories -- they are all in the hands of God. If nuclear war is really the apocalypse then it is out of our control. Only God controls the apocalypse. Only God determines where the Last Battle at the End of Days will be fought (a hill called Armageddon).

Exaggerating nuclear war to mean the end of the world makes it horrible, no doubt, but it also makes it (perhaps comfortingly) beyond our control. It gives us a pass not to worry about the problem.

At least that is the way it seems to me.

I think facing the realities of nuclear war is essential.

I think you missed an opportunity to quantify targeting in US military nuclear planning. Now while the details of specific weapons allocations on specific targets is classified, the process of determining force allocation on legal targets is not.   The US has made it clear that we do not target cities since the early 1960's. since that time, the rules that govern the use of nukes in the us military have grown to be very well defined.

Under US military law, weapon choice must match the scale of the threat. In other words, you don't use a nuke on terrorists in Afghanistan. The weapon cannot be used to target civilians, civilian infrastructure, historical or cultural areas of importance or monuments and museums, The weapons cannot be used against the environment like destroying a huge dam or targeting a nuclear power plant. Additionally, a weapon must be chosen that minimizes civilian casualties while still delivering the outcome that is the justified military objective. So, if a target can be taken out with conventional weapons to minimize collateral damage and civilian deaths, then conventional weapons must only be used.

In areas where civilians are regularly associated with, the military must give advance notice of their planned attack so civilians can vacate the area. Further, any use of such weapons cannot be in retaliation or revenge, you cannot take out a city in response to an attack on one of your cities, no tit for tat. You are also not allowed to start a war of aggression such as invading another country in an effort to subjugate the land and its people.

These rules hold true even if the President orders the use of nukes in violation of these requirements under military law. Such an order would be illegal and the Military would not follow an illegal order.

^[1]

Further guidelines exist under the US Nuclear Declaratory Policy, which further restricts their use in more general terms, stating nukes will never be used against a non-nuclear adversary. However, such statements are left intentionally vague. The US maintains a policy of “calculated ambiguity”. While the US seeks to signal its intent clearly enough to highlight “red lines,” or situations in which the United States may consider employing nuclear weapons. Yet, it also refrains from telegraphing the type and size of an attack an opponent should expect, should it choose not to be deterred.

This ambiguity seems to be in conflict with military law at times. The threats and innuendos that exist between politicians and states are filled with more rhetoric than policy and law. Something the media likes to fan the flames under. No matter what is being said in the media, nothing said overrides the military law on the use of weapons. Politicians talk a lot of smack at the end of the day, keeping a grounded perspective is very useful and necessarily warranted.

So, when is it legal for the US to use nukes?…

Targeting an adversary's nukes and nuclear support and production facilities not near population centers and some military bases that are also outside of civilian centers. In certain battlefield scenarios that were envisioned during the cold war when Russia could have invaded western Europe with overwhelming force using thousands of tanks. A situation that no longer exists or is even possible. They can be used to destroy or damage deep hardened underground military targets where conventional weapons would not do, such as weapon production facilities in North Korea.

Smaller nukes, like the new W76–2, a 5kT warhead that was recently added to our strategic SLBM fleet or the lowest yield on the B61–12 at only 1/3 a kT, blur the lines between nukes and conventional weapons. Such weapons could be legally used as airbursts over large military facilities in civilian population centers with minimal civilian casualties. The concern here is that it becomes too easy to use nukes and that line can be crossed with small nukes and not larger nukes. To any adversary, any attack on them using any nukes, no matter how small, changes their playbook dramatically.

Anything else, with few exceptions, would be illegal under US military law.

This is why the Military considers nukes largely useless which is the primary reason they have not been used since WWII. Their role today is 99.99% deterrence and .01% functionality.

Generally speaking, any effort to improve the utility of nukes, such as developing lower yield warheads, is controversial. While it seems like something you would try to do, make a weapon more useful. In the case of nukes, that change in utility detracts from their primary role of deterrence. Nukes must be viewed from their strategic role more so than any potential tactical usefulness to prevent the temptation of their use in future combat scenarios.

Thanks for the post. The updated nuclear winter modeling is Coupe 2019 - it has similar results to the 2007 work. The new work was funded by Open Phil. I agree that extinction is quite unlikely, but I think there are a few routes to get there. Many people mention the lower climate impacts in places like New Zealand, but they do not consider the possibility of refugees overwhelming New Zealand. You may very well be right that some food storage will be protected, but I don't think it's guaranteed that people could or would forcibly repel the desperate people trying to get food. Furthermore, if there is an eventual collapse of anthropological civilization (cooperation outside tribes), we may have to go back to hunting and gathering, and there is evidence that that transition may not go well (The Secret of Our Success book). And the current hunter gatherers generally don't have much food storage, so they would likely die out.

But then I agree with Max Daniel that even if extinction is unlikely, I think worse trajectory changes have significant probability, which is supported by the poll of GCR researchers here.

As Robock points out, even without firestorms (simultaneous burning of a large area), some smoke can go into the upper troposphere and be lofted into the stratosphere. Indeed, this has been demonstrated for wildfires (which are not firestorms).

There are four different spellings of 'Reisner' (which is correct) in this paragraph:

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 Riesner et al responded to the response. Both authors bring up good points, but I find Rieser’s position more compelling. This back and forth is worth reading for those who want to investigate deeper. Unfortunately Reiser’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 Reiser’s model would result in far less cooling than Robock’s model in the equivalent situation.

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 former, but I believe such an assessment would be worthwhile.  

Do you mean "I have not seen an in depth analysis of the latter"? I.e. humans' adaptive capacity?

This all seems very plausible to me, I've always been sceptical of the idea that nuclear war (or climate change, or a pandemic) could kill all humans on the planet. There's a lot of us, we're very widely distributed and we're very adaptable, we'd probably just end up with a new world order dominated by Australia, New Zealand, Africa and Latin America. It would be great to have more research on nuclear winter, especially since it would overlap with climate modelling and potentially with geoengineering projects designed to deliberately cool the planet, so understanding this better would be great for the future.

Nuclear war would still be very bad, and  I think we should probably be concerned with non-existential risks, since after a near miss it would take a very long time for things to return to where they are today.

The idea that we should lobby the military to be selective in its nuclear targeting is interesting, I'm not sure how tractable that is but the pessimist in me does suspect that militaries are the only branch of government incentivised to think long term.

the amount of radiological weapons 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.

Do you mean that radiological weapons serve no military purpose, or that the amount required to cover every part of the Earth's surface  would serve no military purpose (relative to a more moderate amount of them)?

If the former, that claim surprises me, given that: 

• a large part of the military purpose of nuclear weapons in general is deterrence
• it sounds like radiological weapons would in some cases cause more deaths per kiloton than standard nuclear weapons
  • So other nations may in some cases be more deterred by a country with some radiological weapons than a country with only standard nuclear weapons

If the latter, then it seems like your analysis is consistent with the view that there may still end up being a quite large amount of radiological weapons? If so, that seems potentially important, as it could perhaps affect indirect extinction risks or other existential risks from nuclear war (e.g., by perhaps increasing the expected number of deaths from a nuclear exchange, and changing where and how those deaths occur).

(Note: I don't really know much about radiological weapons, have no particular reason to believe governments are likely to develop them, and would indeed be surprised if they developed enough to cover the Earth's surface.)

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.

...
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 Riesner et al responded to the response. Both authors bring up good points, but I find Rieser’s position more compelling. This back and forth is worth reading for those who want to investigate deeper.

I've always found it a bit weird that so few researchers have work on such an important question. It's good to hear the more researchers are now engaging with the nuclear winter modeling. Besides genuine scientific disagreements about the modeling, I wasn't surprised to find that Wikipedia also notes there are some doubts about the emotional and political bias of the researchers involved:

As MIT meteorologist Kerry Emanuel similarly wrote a review in Nature that the winter concept is "notorious for its lack of scientific integrity" due to the unrealistic estimates selected for the quantity of fuel likely to burn, the imprecise global circulation models used, and ends by stating that the evidence of other models, point to substantial scavenging of the smoke by rain.[179] Emanuel also made an "interesting point" about questioning proponent's objectivity when it came to strong emotional or political issues that they hold.[11]

I think that funding another group of climate modellers to conduct nuclear winter simulations independently of the Robock group would provide a valuable second perspective on this. Alternatively, an adversarial collaboration between the Robock group and some nuclear winter opponents could also produce valuable results.

Your perspective is refreshing. I agree that nuclear winter is overblown. Nonetheless achieving the same countervalue strategy is possible without directly attacking cities. The EMP commission reported that up to 90% of the US population would die within a year after an EMP attack with nuclear weapons. The weapons would not create smoke and an EMP would force the country to devote resources to keeping people alive. In other words an EMP attack keeps demand high by not killing anyone but instantly drops the supply which eventually kills almost everyone.

North Korea has created weapons and the missiles to carry them. However, they have not created the technology necessary for a warhead to re-enter the atmosphere and they have not created necessary guidance systems to accurately hit targets. They are also developing a constellation of satellites that in theory could hold nuclear weapons that are detonated when they are above the targeted countries.

I think this asymmetric strategy is likely to be used by the US and Russia too. Even after a successful counterforce first strike enough warheads would survive to achieve countervalue objectives without risking a nuclear winter. In a nuclear war where cities are targeted the initial SLBM EMPs would still kill more people than the direct attacks on cities. So why risk a nuclear winter even if it is a remote possibility?

Effective altruism is about investing my income to do the greatest good. Should I max out my 401k now to give away more money in the future or invest now in a wood gasifier, windmill, biogas generator, solar power system and micro-hydro in order to avoid having to plough after a possible devastating EMP attack?

I agree that nuclear winter is unlikely because an attacker has better options than firestorming cities.