Matthew McAteer

Black Swan events and the new Cold War

Deterrence ladders, contingency plans, RAND corporation documents, and personally grappling the now much more real threat of nuclear annihilation.

 | UPDATED

UPDATE 05/27/2022: It seems we need to add in the risk that would stem from an accidential conflict in the Western Pacific, too. Added to that section

UPDATE 09/26/2022: I look back on this post on Stanislav Petrov day, a day commemorating a Soiet Lieutenant who successfully stopped a false alarm from escalating into WWIII back in 1983. This is also a few days after Putin announced the ‘Partial Mobilization’, and reiterated the nuclear threat first made back in Febuary. A lot of my assessments in this post still stand, though I’ve added a note on how the “reasons for cautious optimism” have updated. Still, It’s probably a bad idea to over think this. It’s a safe assumption that if world leaders are threatening the use of nuclear weapons much more than they were last year, then the chances of a nuclear war have indeed gone way up.

I apologize in advance for the scattered nature of this post. I’ve done as best as I can to organize it into a coherent structure. Still, some of the word-vomit, stream-of-thought, too-much-information style might still seep through. The only silver lining is that this might better capture the mood of the moment at the time of writing. If this whole situation is overblown, then this post may at least be of some historical curiosity (and if it’s not overblown, then nobody will care for obvious reasons).

With the long-term in mind, I’m also trying to use this as an opportunity to lay out at least part of my reasoning. The goal is that this will make it easier to find steelman arguments for why the situation will develop one way or another. If I encounter compelling enough pieces of information or arguments, I will share those.

Table of Contents

(Hyperlinks to to each section)

Part 1: Initial reactions and responses

One of the ways I’ve tried adding structure to this post is by dividing it into sections. The first captures a lot of the in-the-moment reactions to hearing the news. The next sections capture more of my analysis of the situation after a lot more reading and thinking (not to mention diving into old RAND Corporation & MITRE reports).

This separation is meant to distinguish the initial emotional reaction from the more rational analysis. It’s also because I intensely dislike reasoning about situations in the midst of high emotions or hysteria in those around me.

Hearing the news that rocked me more than COVID did

The year so far had been pretty great. I’m financially better off than I’ve ever been in my life. I have a wonderful network of friends, and spent a lot of the early part of this year prepping for many of their weddings and baby showers. I had been working with a fantastic AI research organization in the Bay Area (one that reminds me of the best parts of OpenAI’s early days). I got a brand new car that I actually feel excited about owning. I also got into a pretty great exercise routine and was beginning to get my 6-pack back.

Generally, I was feeling like years of hard work and “hustle” were starting to more visibly pay off, and not just because I had been on a 90-day streak of meditating every day.

…And then I saw the news about Russia’s invasion of Ukraine, followed a few days later by the news of Russia putting its nuclear forces on a higher alert status.

My Immediate Mental state

I don’t think I’ve ever seen a news story that truly made me fear for my life like this.

I can remember my parents taking me out of school on 9/11, and me getting home to see the scene on the news. At that age (I was just starting 3rd Grade), I didn’t fully grasp until much later why a plane-crash was so scary (though I also realize that I was literally a child back then, so a I was likely spared a lot of the gorier details for that reason).

During college I saw the live news of the Boston Marathon Bombing in 2013. While I was scared, I was also safe in Providence, RI. Even my father who lived in Watertown was travelling at the time.

Fast forward to 2020 I saw the news of COVID-19. While it was pretty scary, I was still young, I had an earthquake preparedness kit that would also be handy in this case (see my speculation on the next global pandemic all the way back in 2012), and I was able to confirm that my family members were safely isolating.

The prospect of a new Cold War hits much differently. I felt like I was suddenly faced with a higher-than-comfortable chance of not just dying, not just dying painfully, but my family and friends and extended social network also meeting that same fate. Intellectually, I knew that the world’s nuclear arsenals had decreased from tens of thousands down to just about a few thousand for the US and Russia, a few hundred for China, and less for the other nuclear powers. I also knew that among these were plenty of megaton-level strategic nukes, ones that would not just destroy a city, but the majority of the surrounding suburbs as well. That’s also not getting into the existence of MIRVs, which could be used to target multiple cities (such as those in the San Francisco Bay area, which was in at the end of February).

Oh, and there were also all the various descriptions of what being hit by a nuclear attack would be like (such as this 2018 video by Kursgesagt: What if we nuke a city?).

In the moment, I had no idea how much of the strong emotional response was rational, or how much it had been primed by watching movies like “The Day After” or “Threads” years ago.

My prepping instincts kicking in

One of the first things I did was look at my stockpile of disaster-prepping materials. Seeing that it was not only partly depleted, but hadn’t really been ever designed for the event of a strategic nuke, I went on something of a shopping spree.

While I was doing this, I was also paying attention to my social media feed. This situation made it incredibly easy to get wrapped up in doomscrolling, but I sill felt like it was worth it to check on what the EA community, rationalist community, and everyone adjacent to those communities was up to.

While I usually follow Balaji Srinivasan for his crypto takes rather than his nuclear deterrence takes, I did agree that getting out of a major NATO city would at least offer me a bit more breathing room to think.

I had played around enough with NukeMap to know that in my current location, assuming it was hit with something between 800 kt (like the Topol SS-25 currently in Russia’s arsenal) and 2.42 Mt (like the R-12 SS-4’s used by the Soviets during the Cuban Missile crisis) I was most likely screwed. Even if I were to survive the blasts, I would need to find a place within the next hour to shelter from fallout for two weeks (and there’s a surprising dearth of houses with basements in the Bay Area).

and

So I decided to go and stay in an AirBnB in Sebastopol for a week.

Why Sebastopol? I knew that I wanted to get away from a large west-coast city, but I knew that wouldn’t be the only target if things went south. I knew that various military bases and airports would be targets in the worst-case scenario. Only issue is that scraping maps for maps of military bases was probably a time-consuming task.

In the end, the heuristic I used was that if something is close to one of the no-fly zones outlined in the FAA’s handy maps for quadcopter drone users, then it’s probably best to stay away from them if the missiles do end up flying.

Example of such map. The no-fly zones extend over a lot of national parks, but a lot of them cover airports and military bases.

Sebastopol isn’t as ideal of a location as a mineshaft or cave, but in the event that the situation got hotter I would be able to make it to the Modoc Plateau in northeast California in 5 hours instead of 7 1/2 hours. Sebastopol also makes it much easier to work remotely than in either Modoc or a mineshaft.

During this week, I was able to get a little more mental clarity, though admittedly this was also a very low bar. Feeling embarrased and ashamed at this state of panic I was in, I decided to make a list of possible future scenarios and envision what my appropriate reactions to them should be (and whether those should warrant evacuation). The result was the following document (heavily based on Robert Wilbin’s list posted on Twitter):

What became my phone lock screen for several weeks

As I looked at this list I had created, as all these scenarios played out in my head, I thought more about how I wish I could be certain of my family’s safety, I scheduled a trip to Thailand to see my father with only a week’s notice.

I applied for a visa, scheduled my flight, packed my bag (with a few potassium iodide pills as well), and went to the airport with everything being approved by the Thai travel authorities only 13 hours before the flight.

And so here I find myself writing this post, in my Father’s Bankgkok apartment in the middle of the night because I’m still suffering Jet Lag.

Part 2: Recalibrating and assessing the situation (now that I’m temporarily far from NATO)

Despite being in Thailand, I doubt that would offer much protection if the Ukraine situation escalated to nuclear war. Still, I was hoping that I would be able to get a better sense of the situation from afar.

Mentally estimating the likelihood of someone refusing orders to launch a missile

Much of the EA community is aware of the story of Stanislav Petrov. Stanislav was a lieutenant colonel of the soviet Air Defense Forces who in 1983 correctly identified a false alarm of an incoming nuclear strike as exactly that, a false alarm. Several friends of mine have used examples like this to express optimism that humans in the loop would ultimately refuse the order to initiate a nuclear launch.

The question is, can we always count on level heads to prevail? After all, the Hawaiian nuclear attack false alarm was caused by a human resolutely following instructions based on what they saw from their machine. There have also been plenty of incidents such as Bombers accidentally being loaded with live nukes. These are also just the cases that have been either declassified, or were screw-ups on such a collosal scale that the mistake was impossible to hide.

I looked to the prediction markets. They were adjusting accordingly, though all they knew was that the probability was going up. This was clearly a bayesian estimate that was based on probabilities going up or down, but not based on any frequentist precedent. These estimates also have truly enormous ranges of estimates from less than 1% to occiasional 50% estimates.

I looked for more specific estimates. In one particularly dour report, the Boston Consulting Group gave a 10% chance of a civilization-ending nuclear war over the next 12 months.

Made even more existentially terrifying by that report bringing up Brandon Carter’s Doomsday Argument, a probabilistic argument that claims to predict the future number of members in the human species given an estimate of the total number of humans born so far.

And then there were all the descriptions of previous wargames.

And the dynamics of nuclear deterrence

Of course all this prompted me to turn to the classic work on nuclear war outcomes: Herman Kahn’s 44-step escalation ladder from the RAND corporation. This is made more complicated by the fact that, in Kahn’s theory, actors do not necessarily move sequentially or steadily towards escalation; conventional warfare or even limited nuclear strikes might just as likely lead to de-escalation in specific scenarios.

Herman Kahn’s 44-step escalation ladder from the RAND corporation, along with a more legible version

As you can see, not only does the situation with US and Russia over Ukraine appear to have skipped a few steps, but the steps that did take place happened over widely varying time-scales (from days and weeks to years). This escalation ladder is also so old that it does not take into account where things like cyber-warfare and hacking would fit in.

Questioning the Concept of Nuclear Winter (and realizing why the alternative outcome is equally terrifying)

Here’s a TED talk from 2018 that explains how we would be killed even if we’re not in the country being nuked. In short, even a “local” nuclear war between India and Pakistan could throw up tons of ash and smoke into the atmosphere. This is a part of the atmosphere where rainclouds don’t normally form, so the smoke would be free to stay there, and reduce sunlight penetration all over the world. This is the reiteration of the concept of Nuclear Winter, which was one of the motivations behind the treaties credited with ending the cold war. After all, if the result of a nuclear winter

That being said, there’s a few groups and thinkers that have cast doubt on the concept since 1993.

For example, a lot of the assumption around particles in the atmosphere is based on volcanoes. Aerosols can spread very far, but particulates can eventually fall to earth. In a nuclear war, the biggest sources of smoke would not be the initial bombs, but the resulting forest fires. If we were to look at how quickly the smoke from the Australian Forest fires cleared, it seems less plausible that the forest fire smoke alone would be able to block the sun for years. However, it wouldn’t just be the forsts that are burning, but human towns and cities with all their plastics and petrochemicals. The fear here would be that the results of such chemicals burning would persist in the atmosphere much longer. In fact, this was also a fear during the Gulf War, when Saddam Hussein set fire to oil fields. Against expectations, the smoke cleared quickly once the fires were put out.

Does this make me feel better about the prospect of a thermonuclear war? Not really, especially considering that the alternative might be a nuclear summer. The fires from the US and Russia being nuked would probalbly set Global Warming forward by almost a century. Even if the smoke cleared quickly, the radioactive particles in the atmosphere would probably make the ozone hole much more global. (This is not even getting into how much topsoil would be destroyed worldwide). Agriculture also wouldn’t be spared, as our modern factory farming is so dependent on fuel supply lines and large complex pieces of equipment that it would quickly disintegrate without the sun being blocked.

I don’t think most people realize just how sensitive plants are to things like ionizing radiation. A lot of the plants growing in the Chernobyl Exclusion zone show extremely elevated stress markers (not counting the huge swaths that turned red and died right after the explosion).

See this example of a breed of Potato grown in Peru. Due to the high altitude, the Potato contains dyes that double as extra UV protection. The overwhelming majority of plants we rely on for agriculture do not have this kind of resilience (heck, tomato reproductive organs usually burst when exposed to temperatures around 50 C for extended periods of time)

Agriculture could very well suddenly become centered on organisms like seaweed, insects, and mushrooms that could still be farmable in low-light conditions. There are already several think thanks that have brainstormed similar proposals, such as using termite stomach enzymes in paper mills to turn leftover wood and paper into human-digestible sugars. Unfortunately, I have not seen any evidence that these emergency plans have been put into place. Getting enough food beyond a starvation diet would still be enough of a challenge that hundreds of millions at best would face starvation.

Even without a nuclear winter, the disruption to fuel sources would still be devastating. Rebuilding petrochemical infrastructure would be a daunting task, partly because human civilization has already harvested a lot of the easy-to-access oil and gas.

Estimating the odds that someone is bluffing

When it comes to any international event, or any kind of interactions between nations, it usually helps to think through the following:

  1. What is the full context?
  2. Cui bono? Who benefits?
  3. Occam’s razor. The explanation with the fewest assumptions (NOT to be confused with the simplest explanation) is often the correct one.
  4. (try to) disregard press narratives.
  5. What are the involved parties’ priorities and comparative advantages?

There’s been plenty of discussion on Twitter about the “context” of the situation. Both from clear experts (those who have lived in russia and Ukraine, and othes that have had more “skin in the game” to quote Nassim Nicolas Taleb), passable experts (those who have studied the history in an acaemic setting, but might make have a huge gap in their analysis), and everyone else. What makes this kind of analysis of the context harder is that this is as hostile of an information environment as it gets. Every party seems to be intent on psyopping everyone else. In terms of reliable context, I’d say resources on the Russo-Georgian War and Second Chechen War offer some good examples about why many are pessimistic that a piece of land being part of a peace agreement would last in long-term peace (which it did not). I’m also not going to go into comparisons of one complex environment to a completely different one. WWI and WWII analogies are not really helpful here (not just because of the obvious absense of nukes). Arguably, the nuclear standoffs of 1962 and 1983 aren’t even fully informative here either Both crises can be used to justify either resolute toughness in the face of threats or compromise, or even something in between. The only real lesson one can take from those nuclear standoffs is that it was dumb and unfortunate that they happened at all. Jumping to these comparisons is sloppy thinking and deludes you into having more context than you actually do.

On the subject of cui bono, It’s clear that Putin himself benefits from the increased anxiety about whether nukes would be used, regardless of if they actually will be.

When it comes to occam’s razor, the simplest explanation is that the threat of nukes is far different from intent to actually use it, and such intent would be met with responses by other nations. It’s also worth remembering that when it comes to escalation, there are many more ladders to be crossed before reaching the nuclear rung.

In terms of press narrative, this is a strange mix between touting how justified the war is against an “evil” and “incompetent” Russia, mixed in with articles highlighting the catastrophe that would follow a nuclear war. Regardless of which narrative is the main one, the important point is that such articles are consistently ranked on the “most popular” lists on most US news sites I’ve seen. This subject definitely gets eyeballs, but I also need to seek out alternative information sources (especially if many of the news articles are repetitive or citing the exact same sources over and over).

In terms of priorities, there is a lot that can be said about Putin’s political leanings, but it should be uncontroversial that his top priority is staying in power. Staying in power is partly done by making sure his inner circle of courtiers are happy. Everyone involved here also has some sense of self-preservation (and I sincerely doubt the rumors of Putin’s ill health are true, he’s just gotten older like everyone else in the gerontocracy).

True, there are plenty of questions about the state of mind Putin is in since isolating so much during the COVID pandemic. After all, he already made several tactical blunders ranging from underestimating the strength of the sense of national identify of the Ukranians, to the readiness of Russian forces, to the willingness of NATO to supply weapons, to the willingness of Europeans to risk having their gas cut off. On the other hand, it’s possible that Putin is already aware of this on some level, and is looking to use this to play into his own version of Richard Nixon’s Madman Theory of deterrence.

…it’s times like this I really wish I had studied modal logic more in school. Then again, I don’t know whether applying this to the game theoretic analysis of a situation where you’ll hear parties saying things like “I knew that you knew that I knew that you knew that I knew that you knew all along” is more or less of a nightmare.

The main takeaways from most of this analysis is that the parties involved are (for now) bluffing. There are probably other frameworks one could resort to to figure out the likelihood of bluffing (though unfortunately I am far from an expert in them). For example, I’ve heard a lot of professional poker player friends of mine talk with confidence about how they believe Putin is bluffing. Professional poker players rarely go all in to the point of ruination beyond what they staked on the game, but then again professional poker players are also fully-in-control singletons that do not have the communication lags or procedural momentum of nation states.

One of the biggest fears right now is not that a nuclear war will start because one or more parties wanted it or intended for it to happen, but because of a miscalculation, miscommunication, misunderstanding, technical glitch, or some other mistake. After all, since I made the tentative emergency plans seen earlier in this post, Russia has seemed much more calm about Sweden and Finland’s seemingly inevitable entry into NATO, though this also more than doubles the length of the border between Russia and NATO and thus greatly reduces the buffer for mistakes.

So what, should the west just give up on Ukraine?

Responding to nuclear threats with “Okay, fine, we’ll give you what you wanted” is a terrible idea.

I’m not just saying this because I’ve spent the past few months talking with friends and coworkers that have lost friends and family and hometowns during the invation (or even having just evacuated themselves). Even if I stay away from the news, their descriptions of what’s happening there don’t go away so easily.

Unfortunately, this situation is not just about Ukraine, but also about not encouraging other nuke-holders (or those who want them) to try the same approach to getting what they want in the future. It would likely also increase the number of countries trying to get their hands on them. Part of the reason South Africa and Ukraine gave up their nukes in the first place (beyond the cost of maintaining them) was the taboos that had discouraged their use in non-test cases since 1945.

A cursory cost-benefit analysis of the Ukraine situation might cause someone to question whether the conflict is worth risking even a 1% chance of global thermonuclear war. However, when framed as an iterated prisoner’s dilemma, backing out in the face of nuclear weapons suddenly becomes a non-option.

Giving into demands so that Putin doesn’t launch nukes isn’t even kicking the can down the road. Its kicking the can an embarrasingly short distance, and doing so in a way that you will shatter the bones in your foot trying to kick it a second time.

In a game of chicken with two people driving their vehicles towards each other, where the first one to swerve loses, the best strategy is to make it clear to the other driver that you’ve torn out the steering wheel. My biggest fear is a miscalculation wherein both sides already think themselves as the one that’s proverbially “torn out the steering wheel”.

Credit goes to Gary Larson and The Far Side. Wasn’t quite sure how to convey these feelings, so I decided to opt on the side of levity for once in what is probably my darkest blog post ever. It is interesting that the original publication date was 07/10/1985, about 2 years after the tensions of 1983.

Reasons for Cautious optimism

There is no end to the pessimistic predictions about how the war might turn out. A majority of the optimism I’ve seen has come from Under 40s that have never experienced the heighted tensions of the Cold War.

Still, I continued to look for steelman arguments for why the the conflict might not escalate to a nuclear war.

Beyond a few scattered optimistic paragraphs in otherwise pessimistic analyses and articles, I came across a helpful summary on Tyler Cowen’s Marginal Revolution: Optimism about the threat of nuclear war

I’ve posted the excerpt in its entirety below. The text was turn was the text of an email that had been sent to Tyler Cowen by Trey Howard. Given this game of telephone, I hope Tyler Cowen and Trey Howard don’t mind me posting it here (if you so, please let me know and I’ll remove it).

From an email from Trey Howard, I won’t impose further double-indent on it:

“I recently came across the pessimistic Edward Luce column you retweeted, and wanted to offer some trends that I think point in the opposite direction. I offer these as someone who was much more worried about nuclear war in the first 2 weeks of the war, before the factors below became apparent.

1. Putin has been willing to revise his objectives. The Russian army fell back from Kyiv, did not launch an amphibious assault on Odessa, and has not attempted to storm the Azovstal steelworks. All of these indicate that Putin is receiving some objective information about the poor performance of his military, and is revising his plans accordingly.

2. Putin’s objectives are amorphous. What does it mean to de-nazify Ukraine? What does control of “the Donbass” mean exactly? These kinds of objectives are susceptible to BS-ing for the domestic audience. They are not like “Kill Zelensky” or “Capture Kyiv”. They permit Putin an off-ramp at any time he wants to declare victory.

3. NATO is unwilling to intervene directly. If anything, I have heard less chatter about no fly zones since the first two weeks of the war.

4. Putin has not escalated to chemical weapons, despite having an opportunity to use them effectively on the Azovstal works.

5. NATO has limited the supply of weapons to short range weapons that a) do not require a complex supply chain of trainers and contractors close to Ukraine or b) are unlikely to cause mass casualties in Russia itself (airplanes, tactical ballistic missiles). This seems likely to continue for the foreseeable future. For all the breathless talk of “heavy weapons” being shipped to Ukraine, it is hard for me to imagine that Russia sees T-72 tanks, towed howitzers, or M113 personnel carriers from the 1970s as tilting the balance. They have thousands of comparable weapons in storage.

6. Russia has not attempted to interdict the flow of weapons inside NATO countries. Not even “plausibly deniable” things like train derailments or warehouse fires. This is all the more remarkable when you consider that the GRU committed attacks in NATO countries in the years before the war started.

7. Putin is not threatened at home. If anything, support for the invasion seems to have increased. The Russian economy has not collapsed as some predicted, and this will bolster support for him.

8. Russia continues to make payments on its foreign debt. To me this indicates a long-term outlook and is not the kind of thing one would do if contemplating murder-suicide at a national level.

9. Russia has not increased the readiness of its strategic nuclear forces (like putting SSBNs to sea).

10. Russia is actively recruiting foreign mercenaries and seems likely to order a general mobilization soon. Some people see this as a sign of escalation, but I think it is more likely that Putin realizes that he needs more bodies to garrison captured territory. Additional conscripts will eventually allow some of the BTGs in action to rotate away from the front lines. It will increase his perception that time is on his side. More troops will make it less likely that Ukraine can inflict a decisive defeat on Russian forces in the Donbass (which might really precipitate tactical nuclear weapon use).

11. Russia is taking over administration of infrastructure in captured territory, and is preparing residents to switch to the ruble. These are long-term thinking measures consistent with a power planning to occupy and administer new territory (which they would not want to irradiate).

12. Putin thinks that the political winds are on his side. Viktor Orban being re-elected, Le Pen performing better than her prior showing, and the coming midterms in the USA all point to populations becoming impatient at the high inflation and constant drumbeat of scary news coming out of Ukraine. Of course, the biggest break for him would be Trump 2024…

I disregard all public statements from Russia (whether from state TV, Putin himself, or lesser officials). There is never going to be a situation where the Russians say “relax, we aren’t going to use the nukes”. They want to keep us guessing. I look at the trends above instead.

Many of these trends are bad news for Ukraine and the west in general, but they are factors that make nuclear war less likely. As you said on a recent podcast “things are never as bad or as good as you might think.""

I am still terrified of nuclear war, but there are at least some reasons for arguing why the conflict might not escalate (or if it does, that it might escalate slower than expected).

Now slightly less optimism (Edit 5/27)

Update 5/27: The risk now also includes potential warfare in Western Pacific.

Adding to the list of conerns is yet more sabre-rattling over the Taiwan strait, this time over a possible visit by Nancy Pelosi.

Still, even if we’re considering the possibility of a conflict int he Western Pacific, I’m not sure that such a conflict would catch us completely off guard. After all, due to the weather around Taiwan, it would be hard to stage an invasion outside the months of April or October. Given that China is also far less food-independent or energy-independent than either the US or Russia, I doubt their plan to invade is imminent (this remaining calendar year, at least). The situation in Ukraine is still far more concerning.

Do the assumptions still hold with the partial mobilization? (Edit 9/26)

With the announcement of the partial mobilization, these thoughts returned to the fold.

1. Is Putin still willing to revise his objectives?

The objectives stated in the partial mobilization speech seem very different from those when I first posted this. So yes, I do think this is still the case.

2. Are Putin’s objectives still amorphous?

As before, it does seem like there are plenty of ways for Putin to BS his domestic audience. There hasn’t been qutie as much variation since the beginning, and it seems like Putin is getting ready to annex several regions in Ukraine, but I think the overall objectives are still amorphous to be changed if needed.

3. Is NATO still unwilling to intervene directly?

I haven’t heard anything about “no-fly” zones since the beginning of the war. There’s only been warnings about what will happen IFF nuclear weapons are used in Ukraine, but that’s still a far cry from boots on the ground.

4. Has Putin still not escalated to chemical weapons?

There were unconfirmed rumors of the use of chemical weapons in Ukraine a few months ago, with both sides pointing the finger. Nothing seemed to come of these rumors, especially not any confirmation. As such, it still seems to be the case that chemical weapons have not been used (which appear to have a lower penalty than even small nuclear or radiological weapons).

5. Has NATO still limited the supply of weapons to short range weapons?

Seems like there’s still limitations on the long-range weapons, though Ukraine also has a new supply of abandoned Russian weapons.

6. Has Russia still not attempted to interdict the flow of weapons inside NATO countries?

Aside from some questionable cyberattacks on gas pipelines, we haven’t seen anything directly targeting weapons in NATO countries yet.

7. Is Putin still not threatened at home?

Support for Putin is lower after some embarrasing military defeats in Ukraine, but given the “mysterious” deaths of some Russian Oligarchs, it seems like Putin isn’t quite backed up against the wall yet. Overall support for the war within Russia appears to be high (even if a lot of the “support” is still passive and there are heavy sampling biases in these polls). In terms of criticism within Russia, most of it has been directed towards military leaders other than Putin, so he still has plenty of scapegoats left.

8. Has russia continued to make payments on its foreign debt?

Russia defaulted on its dollar-denominated foreign debt, but the talks of debt settlemet with china don’t seem like what you’d expect from someone planning a nationwide murder-suicide.

9. Has Russia still not increased the readiness of its strategic nuclear forces?

If the recent testimony of Air Force Brig. Gen. Pat Ryder is anything to go by, this still holds true. That being said, when it comes to increasing the readiness of tactical nuclear forces instead (which could be fired from regular tanks or missile launchers), there would likely be far less warning.

10. Is Russia still actively recruiting foreign mercenaries and does it still seem likely to order a mobilization?

Proven correct about the mobilization, though it’s also been called a “partial” mobilization. This means there’s at least some kind of plan for the next few months. What happens AFTER this stage is more unknown, as it is heavily dependent on just how embarrasing a defeat Russia suffers in the Donbas region.

11. Is Russia still taking over administration of infrastructure in captured territory?

Regardless of whether the referrendums are legitimate (I don’t think forcing people to vote at gunpoint counts), this still holds true. On the other hand, it also seems like it could be a tactic to extend the russian nuclear umbrella to territory that it doesn’t even fully control.

12. Does Putin still think that the political winds are on his side?

Midterms in the US are looking more in favor of the dems, but Italy is poised to elect a right-wing coalition. While everyone seems to have competing definitions of what ‘facism’ is, this still might convince Putin that things will look more politically favorable in the future.

A few of the assumptions still hold, but some are looking weaker or more uncertain.

Manifold Markets’ “Will a nuclear weapon be launched in combat by the end of 2023?” has been hovering around a 10% chance. Metaculus’ “No Non-Test Nuclear Detonations by 2024” also gives roughly 90% odds of no nuclear detonations by 2024 (i.e., also 10% odds of the reverse that there will be a non-test detonation). Metalculus recently posted the much more specific question of Nuclear Detonation in Ukraine by 2023, for which the probability has been hovering between 2% and 8% since the question’s creation (at the time of writing).

Metaculus has longer track record on this topic than Manifold, so I would put more trust there.

In a 2019 post Alex Tabarrok pointed out that expert surveys (not markets) suggested the annualized probability of a nuclear war was on the order of ~1% (which was described as uncomfortably high). There’s also another old report giving a 0.35% chance of full scale nuclear war.

What’s important is that the markets are predicting a probability ~10X higher.

Even if we’re also considering the problem that with a prediciton market some of these outcomes would make it hard to collect on your winnings, a previous discussion on the EA forums also gave a higher probabilities (this time the base rate was calculated based on the historical frequency of mishaps with nuclear weapons)

What if Nuclear ruin doesn’t come?

Even if nuclear war doesn’t come, this whole situation has made me realize just how many other risks we face this decade, from AI Risk, to a 21st century Carrington Event, to even just the immediate destabilizing threats posed by things like lithosphere depletion and Lake Mead drying up (and the energy crisis that would result from that).

In the end, this whole post was just about mentally dealing with an expected value tree that has a few nodes with unknown-but-seemingly-small probabilities, but wit big ol’ negative infinities slapped on their values. I’m reminded of this quote from the abstract of A Paradox of Tiny probabilities and enormous values (v2.1, June 2021)

We show that every theory of the valuue of uncertain prospects must have one of three unpalatable properties. Reckless theories recommend risking arbitrarily great gains at arbitrarily long odds for hte sake of enormous potential; timid theories permit passing up arbitrarily great gains to prevnt a tiny increase in risk; non-transitive theories deny the principle that, if A is beter than B, and B is better than C, then A must be better than C. While non-transitivity has been much discussed, we draw out the costs and benefits of recklessness and timidity when it comes to axiology, decion theory, and moral uncertainty.

While this paper was dealing with extreme probabilities in general, it seems uncomfortably relevant now, save for the fact that the probabilities we’re dealing with in real life are far more uncertain.

Part 3: Figuring out my long-term plans now

I’m not sure how much the Boston Consulting Group’s “10 percent” probability has changed since their March 4th Report. I’m not sure how much the tensions in the South China Sea factored into that. I’m also not sure where the true probability lies due to the probably enormous error bars on all of these predicitons.

I do know that there are various cancers that have a roughly 60% five-year survival probability. If you take the 10% chance of death per year, and compound it over 5 years as 0.95=0.590490.9^{5}=0.59049, you get around the same 60% survival rate (Yes, I’m aware that cancer death rates are not uniform over that 5-year period.). When it comes to figuring out my life plans, my closest analogy is to that of someone planning around having something like colorectal cancer. The obvious difference is that unlike with cancer, this probability is being assigned to most of my family, friends, co-workers, and even strangers I meet both in person and online as well.

Five-year cancer survival in England, all adults, updated 05/25/2021. Source: Public Health England, Cancer survival for England https://www.nuffieldtrust.org.uk/resource/cancer-survival-rates#background

This framework doesn’t exactly give a lot of certainty. I think the only real responses I can have to a situation like this are.

  1. Stop stressing about how my retirement account is going to be doing 10+ years down the line, and make sure my health, family, and finances are all in order in the short-term first.
  2. Do some kind of prepping (this is the equivalent of doing chemo and changing one’s diet in the case of the cancer analogy). Even without a nuclear warhead being detonated, I have no doubt that the coming decade will be chaotic for a varety of reasons (at the very least counting natural disasters like Forest Fires).
  3. Appreciate life and the time I have. Take care of those bucket list items because I don’t know how much time I have left (this advice would apply even without the renewed Cold War).
  4. Tell my friends and family how much they mean to me at every chance I get. Aftr all, you can’t always be sure how long they will be in your life for.

Additional Thoughts

A few thoughts that seemed difficult to categorize in the above

  • Crypto in its current form is definitely a scam (I sold all my crypto back in February), but it may very well be the form that can survive a thermonuclear war like the original design for Arpanet. With that in mind, I am even more bullish than I was before on my good friend Sonia Joseph’s Alexandria project. I don’t think this is something that only crypto can solve. If you had to set up a system quickly, you might be better off creating geographically distributed fararaday-cage-enclosed racks of machines with the entirety of Wikipedia on them (plus instructions for Fertilizer production, making steel and concrete, harvesting antibiotics, and getting petrochemicals production and electric generators up and running again), which could be updated with a different consensus algorithm than blockchain (like RAFT and/or PAXOS).
  • I’m doing a lot of ruminating, ruminating too much is probably a sign that one 1) doesn’t trust their instincts, and 2) one believes that if they think about something enough they can control it.
  • There are plenty of resources out there on disaster prep. Unfortunately a lot of them are either marketting expensive products that probably won’t be your first concern in a disaster (or may be impractical), or they may be written by people who are essentially LARPing about surviving a disaster. I am leaning much more towards the adivce of people that have actually survived such catastrophes like the testimony “Words from a Bosnian Survivalist” (Warning, may require a strong stomach to read). That being said, I’m not really sure who has the requisite experience for prepping for a nuclear attack. Such a scenario combined prepping for a tornado (from the force of the blast and resulting winds), prepping for an earthquake (from the chances of buildings collapsing), prepping for a forest fire (from the resulting fires), prepping for a radiological disaster (the fallout), AND prepping for societal disruptions (for obvious reasons) all at the same time.
  • Despite spending high school at a wildlife hospital where I was routinely in the necropsy ward, despite being constantly reminded of my mortality through my academic career in the longevity space, this whole situation has served as a much better mememto mori than anything else I’ve ever experienced. Just about every physical sensation is suddenty much more intense, both good and bad sensations, in part due to being thankful for being a concious being that can express such sensations.
  • I had grown up christian, and became an atheist at around age 8 (long story). Last year, I decided to try reading the entire bible back to back. On the one hand, a lot of the Old Testament seems to gloss over the more interesting parts in favor of a bunch of text on geneology. On the other hand, with the world seemingly in a game-theoretic death spiral, I feel like I can suddenly appreciate much more Jesus’ message of not just love and forgiveness, but self-obliterating love and forgiveness. I doubt that I’m the first person to have these thoughts. I’m sure there are plenty of other sects of Christianity that think there are much more important sections of the Bible to focus on. I’m also not exactly returning to Christianity or suddenly re-believing in the commonly-depicted Christian afterlife (I think the standard practice in the SF Bay Area is to be a devout adherent of a syncretic religion that you just made up). Still, I do suddenly have a much greater appreciation for this message than before.
  • On the one hand I’m gald I’ve kept a dream journal that’s let me better partally remember the 1/3 of my life that I spend asleed. The downside is that I’m increasingly reminded of how many of my recent dreams are not lucid dreams like I’ve trained for, but more emotional end-of-the-world dreams. On the one hand, being aware of this has allowed me to have fewer dreams like this just from force of will. While I’m aware of just how powerful mimetic contagion can have on dreams, I’m still unhappy with having to clear it out of my head.

Two screenshots from my dream-journal app. On the left is a sample of dreams from Late November 2021, the others coming in March 2022 after the invasion.

  • Interacting with friends again after having the above dreams has been an interesting experience. I’ll be hanging out with a friend, the day after dreaming about trying and failing to save their life during a bombing, and I’ll be trying to strike a balance between letting them know how much I appreciate them and breaking down into tears over how glad I am that they’re still alive. (in my experience, most people tend to react with confusion when you respond with the latter).
  • After looking through the above thoughts and emotional responses, I cannot help but be reminded of this quote by “Louise Erdrich”:

Life will break you. Nobody can protect you from that, and living alone won’t either, for solitude will also break you with its yearning. You have to love. You hvae to feel. It is the reason you are here on earth. You are here to risk your heart. You are here to be swallowed up. And when it happens that you are broken, or betrayed, or left, or hurt, or death brushes near, let yourself sit by an apple tree and listen to the apples falling all around you in heaps, wasting their sweetness. Tell yourself you tasted as many as you could.

— Louise Erdrich

Cited as:

@article{mcateer2022coldwar,
    title = "Black Swan events and the new Cold War",
    author = "McAteer, Matthew",
    journal = "matthewmcateer.me",
    year = "2022",
    url = "https://matthewmcateer.me/blog/new-cold-war-thoughts/"
}

If you notice mistakes and errors in this post, don’t hesitate to contact me at [contact at matthewmcateer dot me] and I will be very happy to correct them right away! Alternatily, you can follow me on Twitter and reach out to me there.

See you in the next post 😄

I write about AI, Biotech, and a bunch of other topics. Subscribe to get new posts by email!


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

YOU MIGHT ALSO LIKE

Handy GPT-3 PromptsHandy GPT-3 Prompts

September 18, 2022

Or, best tips for offloading your creative work and social media presence to your AI replacement.

Mimicking DeepMind's Chinchilla with GPT-3Mimicking DeepMind's Chinchilla with GPT-3

May 12, 2022

Short piece on using GPT-3 to reconstruct behaviors of one of DeepMind's new models

Matthew McAteer

@MatthewMcAteer0

Biologist-turned-ML-Engineer in San Francisco. I blog about machine learning, biotech, distributed systems, dogs, and more.

At least this isn't a full-screen popup

That'd be more annoying. Anyways, subscribe to my newsletter to get new posts by email! I write about AI, Biotech, and a bunch of other topics.


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.