100Joined Jul 2020


One addition (that is not legal advice!): under Section 546 of the Bankruptcy Code, the Trustee has up to two years from the bankruptcy filing to initiate a clawback under Section 547. These clawbacks may be filed en masse a few months or weeks before that deadline. It may be some time before this is resolved.

I'm a little skeptical of the "new value" point. The rationale behind the exception is that, "because new value is given, a contemporaneous exchange does not diminish the debtor's estate." In re JWJ Contracting Co., Inc., 371 F.3d 1079, 1081 (9th Cir. 2004). A grantee might provide value in a philanthropic sense, by doing charitable work, but that work wouldn't restore the financial value to FTX, which is what the bankruptcy court cares about.

(This is not legal advice; I am not a lawyer; I very much agree with Jason's caution against overconfidence in either direction.)

Thanks for writing this Max.

I believe the risk is less than 1%. Here are some quick thoughts on why, adopting your framework.

(I am not an expert on Russia, Ukraine, or nuclear war; these are just rough comments)

Odds of Russia using a nuclear weapon in Ukraine in the next year

I put this around 10%. I have the highest uncertainty here, as it's hard to understand decision making processes in the Russian leadership from abroad, Russian nuclear doctrine is (purposefully) opaque, and escalation dynamics between a nuclear and a non-nuclear state are not well understood.

Yes, Putin's nuclear signaling has grown more aggressive recently, and Russian strategists do not seem to believe that limited nuclear use would necessarily lead to uncontrolled escalation, so they may be willing to countenance the use of smaller nuclear weapons against military targets.

But I'm lower than you because: 

(1) While Russia's military doctrine relies on heavily on nuclear weapons, it involves more coercion through threats than actual use, so nuclear signaling from Russian leaders shouldn't surprise us. 

(2) It isn't clear what military objective using nuclear weapons would achieve. Russia would have to use several to truly cripple the Ukrainian military, and nuking a big city like Kyiv would serve no military purpose while alienating the major countries that are still friendly to Russia.

(3) Russia's response to the Kerch Strait Bridge attack was very muted. The bridge is a key supply line for Russian forces in the south and a major part of Putin's legacy. Russia has not turned the rhetoric up in response to the attack, instead accusing Kyiv of terrorism, not crossing a red line.

(4) A generalization of (3)—I don't think we're in "a self-perpetuating escalation spiral." Neither side has openly escalated the war outside Ukraine's borders; Russia hasn't used chemical weapons despite fears it would; and the U.S. seems to be working to restrain some Ukrainian escalation (e.g. the recent story about U.S. officials opposing Ukraine's alleged attempt to assassinate Dugin). 

(5) From public reporting, U.S. intelligence doesn't seem to think Russia is preparing for nuclear use. The U.S. has had very good insight into Russian plans since months before the war started and would probably be able to tell if Russia was moving weapons from storage sites, preparing warheads, etc.

Odds of a U.S. conventional military response directly against Russian targets

I think there is maybe a 25% chance of this. In 2016, the U.S. ran a wargame premised on a Russian invasion of the Baltics followed by a nuclear attack on responding NATO forces. The National Security Committee Principals Committee (Joint Chiefs + some Cabinet members) recommended a nuclear attack on Belarus in response, aiming to avoid escalation but preserve deterrence. The NSC Deputies (the next rung down) recommended a conventional attack on Russian forces and an attempt to rally the world against Russia for breaking the nuclear taboo.  Some of those deputies are now principals, Biden himself is generally fairly dovish, and that wargame involved a scenario where Russia had used a nuclear weapon on a NATO country. I think the odds of a direct military response to a nuclear attack on a non-NATO country are much lower. 

The U.S. has taken significant care not to cross any Russian red lines so far. U.S. decisionmakers universally rejected a no fly zone; the U.S. has held back on a lot of advanced military hardware, like fighter jets, that could enable Ukraine to strike Russian territory; and U.S. leaders have fairly consistently signaled that there needs to be a negotiated solution.

I'd expect the U.S. to respond by giving Ukraine greater capability to launch conventional counter-attacks than it has so far and by trying to get China and India to pressure Russia to end the war. For a direct military response by the U.S. to make sense, it would have to be signaled beforehand (probably, but not necessarily, publicly). It's important to separate commentary by those outside the administration from those within it. As far as I know, no actual decision maker in the US (president or NSC principal) has threatened such a response. 

Odds of full nuclear war if the U.S. does respond directly

I put this around 8%. It is one thing for Russia to nuke a non-NATO country; it is quite another to nuke NATO itself. Russia's leaders all have long experience dealing with NATO and have no illusions about the consequences of a nuclear response to a conventional NATO attack. Backing down to NATO, which Russians realize is a stronger adversary, is also a lot more politically palatable to Putin than backing down to Ukraine, which on paper is much weaker. 

I read the history slightly more optimistically than you do. Able Archer is often cited as the most dangerous moment, but more recent research suggests that the Soviets were not as worried about an attack as previously thought. The biggest danger from the Cuban Missile Crisis was that Kennedy and Krushchev didn't know everything their own forces were doing, not that either of them wanted nuclear war. Military command and control is significantly better today, the US and Russian militaries have well-established lines of communication, and much of the danger you're positing seems to come from reckless behavior by Putin that I don't think his track record justifies. I agree, however, that the U.S. and Russia are much better at escalation than deescalation.

I think this gets at something important, but:

The “moral intuition” is clearly not generated by reliable intuitions because it abuses: 

a. Incomprehensibly large or small numbers 

b. Known cognitive biases 

c. Wildly unintuitive premises 

This list also applies to prominent arguments for longtermism and existential risk mitigation, right? For example, Greaves & MacAskill think that the charge of fanaticism is one of the most serious problems with strong longtermism, which "tend[s] to involve tiny probabilities of enormous benefits." To the extent that's true, the extreme thought experiments seem to capture something significant. If they reveal a failure of utilitarianism, strong longtermism may fail too.  (I realize there are other, non-longtermist arguments for x-risk reduction.)

Does he have a preferred resolution of the St. Petersburg Paradox

Or, put another way, should a probabilist Benthamite be risk-neutral on social welfare? (On my mind because I just listened to the Cowen-SBF interview where SBF takes the risk-neutral position.)

Thanks for sharing, Teddy. Just read the SBF piece and looking forward to reading more. 

EA is great, but I think it could use more and better external scrutiny. A lot of the criticism it gets right now is either silly (e.g., "this is just a way to make billionaires more powerful") or not especially helpful (e.g., "utilitarianism is obviously false"). Engagement that takes EA seriously but keeps a critical distance and isn't afraid to question the big players is really valuable IMO.

Some reasonable points but overall I did not find this insightful. 

A few examples of Burja's sloppy analysis: (1) a "limited no-fly zone" in western Ukraine would be militarily irrelevant (Russia isn't flying fixed wing aircraft over these areas) and would still be seen by Russia as a major escalation ("how do we know NATO will stop there?"); (2) the idea that China has gotten "everything it wants" is flat wrong, apart from anything else it's clearly not in China's interest to see a remilitarized Europe and Japan or an economically wrecked Russia; (3) Burja offers no support for his claim that the west would care much less about an invasion of Taiwan; and (4) a prediction as surprising as war between France and Turkey (both NATO members!) needs more than the zero evidence given.

Burja's implicit theory seems to be some kind of realism, but he never spelled it out, and he's not consistent about it. This point, for example, doesn't fit in a realist framework: "It's, Russia has invaded our friend, and we want to protect this state or at least help it defend itself."

I think EAs would do better reading more traditional IR experts than people like Burja, even if they position themselves as rationalist-adjacent.

"National security law schools:

  • In DC: Georgetown, George Washington, George Mason
  • Others: University of Maryland, University of Virginia"

One law student's view FWIW: these choices may have some national security career benefit beyond other similarly ranked law schools, but in general, I strongly recommend going to a higher ranked school, and if that isn't possible, picking another option, such as a masters. 

Graduates of the top few law schools—Yale most of all, then Harvard, Columbia, UVA, and Georgetown—dominate the top federal government nat sec jobs held by lawyers. For example, of the last 12 national security advisers (since 1993), 3 Yale JDs, 1 Harvard JD, 1 UVA JD, 1 Berkeley JD. Of the last 15 Deputy NSAs (since 1993): 2 Yale JDs, 1 Georgetown JD, 1 Columbia JD, 1 UVA JD. Of the last 9 (since 1993) Secretaries of State: 1 Yale, 1 Harvard JD, 1 Columbia JD, 1 Boston College JD.*

*Not counting interim/actings.

Can you give more details on the $$ you moved in EU animal welfare advocacy/funding? What kinds of approaches did it move from and to? Any exact numbers and specific organizations you can share?

I poked around your website but couldn't see more on this—apologies if I missed it!