Book Review: Going Infinite

SBF Is Bad

I have recently finished Michael Lewis's Going Infinite. Michael Lewis, a financial journalist, embedded with FTX in order to report on their business success, and so was present for their downfall. As the least involved person who was there for all of it, Lewis therefore has a unique perspective on the FTX situation.¹²

I've been told that, in interviews promoting the book, Michael Lewis is unduly sympathetic to Sam Bankman-Fried, but I have to say I can't imagine what the unsympathetic version of this book is like. Sam Bankman-Fried comes off as utterly incompetent and completely indifferent to other people's feelings. That said, the early chapters in particular have a certain air of not having been revised after FTX turned out to be a fraud. A lot of things that are cute and quirky from the world's youngest billionaire are just assholish from a thief. 

For further context on the book, I highly recommend this book review by Zvi Mowshowitz, a former Jane Street employee.

A friend of mine caught one mistake: Tara MacAuley is inexplicably described as being a mathematician even though she is a pharmacist with no formal mathematics training. While I haven't caught any other major errors, readers should expect the book to likely contain other errors of similar magnitude, and discount this post accordingly.  

You Probably Would Have Quit Alameda

As far as I can tell, "what would you have done if you'd worked at Alameda Research?" is the effective altruist equivalent of "what would you have done if you were a German in the 1940s?" Instead of the messy moral dilemmas we face in our everyday lives, it's a moral dilemma with exactly one correct answer-- and one that a lot of people like you got wrong. It's entrancing to ask yourself if you would have measured up. Because of who effective altruists are as people, the usual answer is to self-flagellate about how you are biased, you would have fallen victim to groupthink, you wouldn't have been able to get it together in the moment of your greatest trial, you must try harder... 

Guys. Within six months of Alameda Research being founded, the entire management team and half the employees quit. Some of the employees that didn't quit were new employees, or otherwise unaware of what was going on. The bar you have to meet here is being in the top 50% of the population for Likelihood Of Quitting Dubious Organizations.

The people Lewis discusses in more detail-- Gary Wang, Nishad Singh, Caroline Ellison-- all had personal reasons for not quitting Alameda. Wang was pathologically shy and barely spoke to anyone other than Bankman-Fried. Singh was a people pleaser who hated dealing with conflict. Ellison had complicated emotional entanglements with Bankman-Fried and had just quit her job. It seems likely to me that many of the other people who didn't quit had similar reasons for not quitting. If, instead of being a people-pleaser or pathologically shy, you are a notably fighty and disagreeable person who loves taking stands on principle and will occasionally take one even when there are no principles to be found, you would probably have quit too, and you do not need to self-flagellate about it.³

It's Not Philosophy, It's Incompetence

Here are a few of the sentences in Going Infinite that I quoted in my notes and followed with the word BAD:

[Bankman-Fried] told his fellow managers that in his estimation there was an 80 percent chance that [the lost cryptocurrency] would eventually turn up. Thus they should count themselves as still having 80 percent of it. 

Gary, though unintentionally, had done something similar with the computer code, which was indecipherable to everyone but him. “Gary was the only person who understood how to code it, and Gary didn’t talk to anyone,” said Nishad.

Twenty-four different people thought that they were reporting directly to Sam, for example. The group included Sam’s father, Joe, and Sam’s childhood friend Matt Nass, whose game, Storybook Brawl, Sam had for some reason just bought. This group did not include the chief financial officer, because FTX did not have a chief financial officer. There was no chief risk officer or head of human resources, because they had none of that, either.

Six months later, [Sam Bankman-Fried] wouldn’t even know if he still owned $100 million worth of Twitter, or if the shares had been sold to Elon Musk.

“It’s unclear if we even have to have an actual board of directors,” said Sam, “but we get suspicious glances if we don’t have one, so we have something with three people on it.” When he said this to me, right after his Twitter meeting, he admitted he couldn’t recall the names of the other two people. “I knew who they were three months ago,” he said. “It might have changed. The main job requirement is they don’t mind DocuSigning at three a.m. DocuSigning is the main job.”

In conclusion: Alameda and FTX were simply not very good at things. They lacked a number of skills which I would consider to be very basic for any large organization, such as "knowing how much money you have," "making sure your software is properly documented," "making sure people are accountable," and "having a remotely normal organizational structure." In fact, the only organizational chart ever created for FTX was created by their collective therapist (?!) to help him keep track of who was who. 

A lot of people have takes like "the FTX collapse was because of Sam Bankman-Fried's weird opinions on the St. Petersburg paradox." That is giving Sam Bankman-Fried far too much credit. FTX and Alameda were in no way run well enough for their collapse to have been related to any sort of plan or calculated gamble on Bankman-Fried's part. In order for his fall to be related to his weird opinions about infinite bets, Sam Bankman-Fried would first have had to know how much money he was betting, instead of taking the same approach to finances that I do to the location of my bike helmet.⁴

Things almost never go badly because people have weird opinions on obscure thought experiments. Things regularly go badly because people fail to document their code or follow generally accepted accounting principles. Obscure thought experiments are fun to think about, and code documentation and generally accepted accounting principles are boring, but we should not confuse interestingness for importance. 

In my opinion, shitty accounting should be understood to be an enormous red flag of organizational incompetence. This is table stakes.  

Have Less Confidence In Young Brilliant Generalist Contrarians Who Defy All Conventional Wisdom

Another quote on my "bad!" list:

“It’s a moderately bad sign if you are having someone do the same thing they’ve done before,” [SBF] said. “It’s adverse selection of a weird sort. Because: why are they coming to you?"

FTX and Alameda employed exactly the kind of people you'd expect from organizations run by someone with that hiring policy: brilliant people in their twenties with absolutely no domain experience.

To be fair, no one really has much domain experience in crypto trading, so it makes sense to hire the smartest people you can and hope they learn on the job. But more than three of the first twenty employees at your hedge fund should have some experience trading. And there's no reason for the head of PR for a billion-dollar company to be a person who had never before worked in public relations. Alameda and FTX would likely have been less grotesquely incompetent if they had ever hired anyone who had any idea what they were doing. 

There have been a lot of jokes and arguments about Sam Bankman-Fried's conclusion that literature in general and Shakespeare in specific was bullshit. But ultimately, as much as it hurts my snobbish little heart to say so, it doesn't matter if someone can't appreciate Shakespeare. It really does matter if someone skims a couple of self-help books about management, concludes that it's all bullshit, and decides to yolo being the CEO of a billion-dollar firm.⁵

I am a brilliant young generalist with no particular domain expertise, so I'm hurting my own team here. But having a high IQ does not substitute for knowing what you're doing. Sometimes, when stodgy old grownups tell you to do something, you need to do that. Sometimes you need to do that even if it doesn't make very much sense to you. 

“There’s a functional religion around the CFO,” said Sam. “I’ll ask them, ‘Why do I need one?’ Some people cannot articulate a single thing the CFO is supposed to do. They’ll say ‘keep track of the money,’ or ‘make projections.’ I’m like, What the fuck do you think I do all day? You think I don’t know how much money we have?"

 Yes, Sam. I do think you don't know how much money you have. 

You Can Infer Things About People’s Behavior In Important Situations From Their Behavior In Unimportant Situations

There's a meme in effective altruist circles that it's a waste of time to be too concerned with frugality. Sometimes, of course, that's true: washing and reusing your Ziploc bags to save the cost of buying them new isn't exactly a good use of time for a partner in a BigLaw law firm. But I think that this meme can make one of the problems with FTX's culture harder to see.   

FTX was careless with money, in that it ordered in enormous amounts of takeout food for its employees every day most of which was thrown out. FTX was careless with money, in that it dropped hundreds of millions of dollars on real estate without having any instructions to give the architect, and in that it paid Riot Games a hundred million dollars for a sponsorship for no apparent reason other than Bankman-Fried liking League of Legends. FTX was careless with money, in that it and its associated hedge fund lost billions and no one seemed to have, like, noticed.  

I propose that these all reflect the same underlying attitude towards money: the (very non-EA!) refusal to make sure that you're spending money the best way possible instead of just shooting it out of a cannon like T-shirts at a basketball game.

Now, of course, an organization might order in enormous amounts of takeout food for its employees every day, most of which is thrown out, as a deliberately considered tradeoff. If FTX were as profitable as it was believed to be, perhaps this would be an investment in employees' happiness and willingness to work long hours that ultimately paid for itself. But FTX wasn't doing that! It was shooting money out of the money cannon! And I think it's fair for a person to say "acknowledging that this could be the right decision and with a lot of uncertainty, I am still making negative inferences about your financial savvy." 

Further: as I mentioned above, as depicted in this book Sam Bankman-Fried is an asshole. When he said "yes" to doing something, it normally meant "I have a nonzero probability of doing this thing.” His head of PR regularly found herself justifying to a room full of important people why Bankman-Fried was inexplicably not there. He flaked on giving the keynote speech at Time's 100 Most Influential People event; the speech was delivered by a drunk and confused FTX employee instead. He said "yup" in response to everything regardless of whether he agreed or not; the longer the yup, the less attention he had been paying to what you were saying. Many people thought he'd agreed to something when in reality he'd just wanted you to shut up and go away. Much of the time, when someone had specifically set aside time to have a meeting with him, he was playing video games and maybe 10% listening. But it's fine! He can just say "yuuuuuuuuup!"⁶

There is a notable contrast between Bankman-Fried's behavior and the behavior of the effective altruists I know, most of whom are strikingly kind, humble, reliable people who keep their commitments, tell the truth, try to understand other people's point of view, and (importantly) listen to you while you're talking. The effective altruists I know also haven't stolen anybody's money? I feel like these things are related? 

Flaky, irresponsible liars with no respect for other people's time do not magically become good people when it is important. If you suck in normal life, you are probably going to suck when making decisions that affect billions (if not, on some plausible theories, trillions) of people. 

The kind of person someone is affects every aspect of their lives. How people do the little things speaks volumes about their character. If someone sucks, notice it. 

Why Exactly Does Dr. George Lerner Still Have A License?

Psychiatrist George Lerner was the collective therapist for almost the entirety of FTX, causing unimaginable conflict of interest issues. He casually socialized with his patients, which is a serious boundary violation. He then proceeded to talk about information about specific people that he gained in therapy to a financial journalist, for the journalist to put into his New York Times bestselling book.

Lerner appears to be under the impression he can get away with all of this because he was technically a consultant, and not a therapist. As far as I can tell, this just means he practiced medicine in the Bahamas without a locally valid license. 

I don't understand how this man still has a medical license. As far as I can tell, he doesn't even have any complaints against him. I assume that everyone directly affected by FTX is a bit too busy to file a complaint, but I really hope they get around to it sooner rather than later, because this man deserves to lose his license and spend the rest of his life getting no closer to medicine than a minimum-wage job stocking shelves at Walgreens. 

1. Some of my friends are friends with people more closely involved in the FTX situation, and Caroline Ellison and I were Tumblr mutuals for a long time. In this review I am not including any nonpublic information, although some nonpublic information has informed my assessments.

2. I did enjoy the number of concepts that Michael Lewis thought I needed a five-page explanation of, including “effective altruism,” “Mathcamp,” “puzzle hunt,” and “Magic: the Gathering.” How many people reading a Michael Lewis book about FTX don’t know what Magic: the Gathering is?

3. If you're a pathologically shy people pleaser, probably work on that.

4. I keep losing my bike helmet too.

5. Yes, he did that.

6. He did play video games through a meeting with Anna Wintour, accidentally agree to be the guest of honor at the Met Gala, and then flake on her, which is the one time this behavior isn't assholish and is instead hilarious.