Thread by @ClintEhrlich

The conventional wisdom is that we're headed for a second Cold War with Russia.

I disagree. We're flirting with a hot war.

It could involve nukes. Billions of people could die. 🧵

Just like before the Iraq War, there is a pro-war lobby pushing Biden to attack Russia.

They want to put together a "coalition of the willing."

Russia will be attacked if it does not surrender the territory it acquired in 2014. defenseone.com/ideas/2022/01/…

The call for war is not coming from fringe people in lonely corners of the internet.

That op-ed is from Obama's deputy assistant secretary of defense.

She says that we should, "if necessary, prepare for war" with Russia.

Under what conditions would war be necessary?

If Russia does not end its "illegal occupations" of Ukraine and Georgia.

In other words: Russia must respect our version of international law, or we attack.

The problem is that the territory Russia acquired in 2014 includes Crimea.

The Russia hawks claim that is the problem: because we let Putin have Crimea, now he's demanding more.

They want to "roll back" Russia from that territory, "even at risk of direct combat."

This is a far more aggressive strategy than the U.S. pursued during the Cold War.

America primarily employed a "containment" strategy, as I described in my thread on George Kennan.

It never tried to "roll back" Russia inside the Soviet Union itself.

In this instance, direct combat could easily escalate to all-out nuclear war.

From Russia's perspective, this new "coalition of the willing" would be invading its homeland.

The Kremlin views Crimea as just as much a part of Russia as Moscow.

My summary of the Russian government's attitude towards the territory is not an understatement.

It is informed by my time as a Visiting Researcher at MGIMO, where I had the opportunity to speak to Russian officials and strategists.

Let us assume, for the sake of argument, that war advocates are correct that their anti-Russia coalition would have conventional superiority.

The question is whether Russia would use nuclear weapons to defend its territory.

It is the official policy of the Russian federation that certain conventional attacks deserve a nuclear response.

Specifically, Russia has retained the right to use nuclear weapons against conventional attacks that threaten the existence of the state.

In response to a succesful invasion, Russia could interpret this policy as allowing it to use nuclear weapons to defend Crimea.

The technical legal criterion would arguably be met – and, even if it weren't, the policy could be stretched.

It is useful to consider what the U.S. would do under similar circumstances.

Imagine if Russia invaded Texas and defeated the U.S. military conventionally.

Would we give up our land? Or would we use our most powerful weapons to defeat the hostile occupiers?

There are other scenarios, short of a U.S. attack on Crimea, that could also spark nuclear war.

Russia is already threatening to deploy forces to Venezuela and Cuba.

The goal would be to impose costs on the U.S. comparable to the NATO threat in Ukraine. nationalinterest.org/blog/buzz/russ…

This would effectively be a repeat of the Cuban missile crisis.

Most people have no idea how close we came to global nuclear war during the original crisis.

It was a fluke that the planet wasn't destroyed.

This time we might not be so lucky.

When the Soviet Union deployed missiles to Cuba, America imposed a "quarantine" of the island.

It didn't matter that Moscow was responding to U.S. missile deployments in Turkey and Italy.

The Navy was authorized to attack any Soviet ship that refused to be searched.

Depth charges were dropped on a Soviet submarine, B-59, which was armed with nuclear torpedoes.

The captain and the political officer voted to fire one of those weapons in retaliation.

It is widely agreed that, if this command had been executed, global nuclear war was likely.

The two votes would ordinarily have been enough to launch the attack.

But, by chance, the Soviet flotilla's chief of staff, Vasili Arkhipov, was also onboard B-59.

He voted against launching the nuclear attack. The others listened, and extinction was averted.

To a sane person, this incident shows how close the world came to complete annihalation during the Cold War.

It was pure luck that prevented the conflict from going hot and crossing the nuclear threshold.

If we recreate the same scenario, we may easily get a different result.

On its own, that evidence would be enough to make us take the risk of nuclear war seriously.

But there is an added dimension of urgency due to probability theory and computer models.

They discredit the main argument "debunking" nuclear war.

What is that argument?

It's that, because prior crises with Russia didn't trigger nuclear war, this one won't either.

Here is a typical, obnoxious example of this line of thinking. "We didn't die then, so we'll be safe now."

It neglects the fact that global nuclear war is an existential risk.

A danger that threatens the existence of all humans cannot be evaluated using standard intuitions about probability.

futureoflife.org/background/exi….

Specifically, one has to engage in anthropic reasoning.

What does that mean?

That you have to adjust your probability estimates, based on the limits to what you can observe while alive.

To analyze this puzzle at the level of civilizations, we have to use complex terminology. (e.g., "the Self-Sampling Assumption.")

But we can make the problem intuitive if we describe it at the level of an individual.

Imagine you have a friend who engages in reckless activities.

For fun, he runs across freeways and plays Russian roulette.

He tells you, "I haven't died a single time, so I'm confident I'll be safe."

The problem with this line of reasoning is obvious.

Your crazy friend *can't* observe any instances where his dangerous gambles kill him.

His pool of past experiences is filtered: it's a biased sample, since it excludes death as an outcome.

We are facing the same problem as a species.

Our collective experience excludes scenarios that produce human extinction.

By definition, we never get a "second chance" with them. The first time is the last time. We all die.

Does nuclear war qualify as this kind of risk? Yes.

There was a period in the late 1980s when scientists began to doubt whether it would actually cause human extinction.

It was theorized that nuclear winter might actually be more like "nuclear autumn" – bad, but survivable.

However, the best available computer models now indicate that nuclear winter would be *even worse* than originally estimated.

The sun would be blotted out for years upon years.

Most humans would starve to death.

Given that a nuclear exchange could eliminate our species, the absence of such an experience in our past is not a surprise.

People who invoke the absence of such an exchange to lobby for war with Russia are irresponsible.

Like your crazy friend who plays Russian roulette.