The podcast episode discusses various topics, including the collapse of FTX, the role of Sam Bankman-Fried, and the creation of Architect, a new company founded by Brett Harrison.
The episode begins with Brett Harrison, the founder of Architect, discussing his experience working at FTX US. He explains that he was hired by Sam Bankman-Fried as the president of FTX US in 2021, and his role was to establish the US business and build out the team. However, he soon realized that Sam was not as involved in the day-to-day operations of the company as he claimed to be.
Brett mentions that Sam was more focused on his public image and media appearances, which he felt was diluting his time and attention away from the company. He also notes that Sam was not a details-oriented person and relied on his team to handle the operational aspects of the company.
The conversation then shifts to the collapse of FTX and the role of Sam Bankman-Fried. Brett explains that he was not aware of the extent of Sam's mismanagement and deception, but he did notice some red flags, such as the company's inability to hire more developers and the lack of transparency in the company's financials.
Brett also discusses his experience with Sam's leadership style, which he describes as "incompetent" and "conflict-avoidant." He notes that Sam would often avoid direct confrontation with his employees and instead rely on his team to handle conflicts.
The episode also touches on the topic of regulation and the role of regulators in the crypto industry. Brett explains that regulators saw FTX as a company that was willing to work collaboratively with them, but he notes that the company's collapse has set back the development of regulatory frameworks for the industry.
Brett also discusses the future of the people who worked at FTX, noting that many of them are talented and will go on to do great things. He also mentions that he expects to see a "FTX mafia" emerge, with many of the company's alumni going on to start their own companies.
The conversation then shifts to Brett's new company, Architect, which provides a unified infrastructure for accessing the digital asset ecosystem. Brett explains that the goal of Architect is to make it easier for individuals and institutions to access all corners of the digital asset ecosystem, and to provide a more efficient and transparent way of trading and investing in crypto.
Brett also discusses the future of crypto and the role of Architect in the industry. He notes that crypto is a viable asset class that needs better infrastructure to support its growth, and that Architect is working to provide that infrastructure.
The episode concludes with Brett discussing the importance of crypto trading and the potential for decentralized trading of stocks and other equities. He notes that existing settlement systems in traditional markets are slow and outdated, and that a decentralized settlement layer could make a lot of sense.
Overall, the podcast episode provides a detailed and insightful look at the collapse of FTX, the role of Sam Bankman-Fried, and the creation of Architect. It also touches on the future of crypto and the role of Architect in the industry.
Key points from the episode:
Takeaways from the episode:
I flew out to Chicago to interview Brett Harrison, who is the former President of FTX US President and founder of Architect.
In his first longform interview since the fall of FTX, he speak in great detail about his entire tenure there and about SBF’s dysfunctional leadership. He talks about how the inner circle of Gary Wang, Nishad Singh, and SBF mismanaged the company, controlled the codebase, got distracted by media, and even threatened him for his letter of resignation.
In what was my favorite part of the interview, we also discuss his insights about the financial system from his decades of experience in the world's largest HFT firms.
And we talk about Brett's new startup, Architect, as well as the general state of crypto post-FTX.
Watch on YouTube. Listen on Apple Podcasts, Spotify, or any other podcast platform. Read the full transcript here. Follow me on Twitter for updates on future episodes.
Timestamps
(0:00:00) - Passive investing & HFT hacks
(0:08:30) - Is Finance Zero-Sum?
(0:18:38) - Interstellar Markets & Periodic Auctions
(0:23:10) - Hiring & Programming at Jane Street
(0:32:09) - Quant Culture
(0:42:10) - FTX - Meeting Sam, Joining FTX US
(0:58:20) - FTX - Accomplishments, Beginnings of Trouble
(1:08:11) - FTX - SBF's Dysfunctional Leadership
(1:26:53) - FTX - Alameda
(1:33:50) - FTX - Leaving FTX, SBF"s Threats
(1:45:45) - FTX - Collapse
(1:53:10) - FTX - Lessons
(2:04:34) - FTX - Regulators, & FTX Mafia
(2:15:42) - Architect.xyz
(2:30:10) - Institutional Interest & Uses of Crypto
Dwarkesh Patel
You are probably going to be fired for this letter that you wrote. Sam is going to destroy your professional reputation. Like, where do you think you're going to be able to work after FTX? He was threatening me.
When I knew Sam when he was 21, 22 years old, he was like a happy, healthy-looking kid. When I got to FTX, I saw someone who was very different than that person I remember. And it felt like he was spending virtually no time helping the company move forward. It was so much about image and brand and PR.
Media was primed for the archetype that was Sam. Doesn't matter how little time he spent with the company. Doesn't matter how he treated employees internally.
Brett Harrison
Architect makes it really easy to access kind of all corners of the digital asset ecosystem.
Dwarkesh Patel
Okay, today I have the pleasure of speaking with Brad Harrison, who is now the founder of Architect, which provides traders with infrastructure for accessing digital markets. Before that, he was the president of FTX US. And before that, he was the head of ETF technology at Citadel. And he has a large amount of experience in leadership positions in finance and tech. So this is going to be a very interesting conversation. Thanks for coming on the Lunar Society, Brett.
Brett Harrison
Yeah, thanks for coming out to Chicago.
Dwarkesh Patel
My pleasure. Is the growth of ETFs a good thing for the health of markets? There's one view that as there's more passive investing, you're diluting the power of smart money. In fact, what these active investors are doing with their fees is subsidizing the price discovery that makes markets efficient. With passive investing, you're free riding off of that. You were head of ETF technology at Citadel, so you're the perfect person to ask this. Is it bad that there's so much passive investing?
Brett Harrison
I think on net, it's good. Most investors in the market shouldn't be trying to pick individual stock names. The best thing people can do is invest in diversified instruments. And it is far, far less expensive to invest in indices now than it ever was in history because of the advent of ETFs.
Dwarkesh Patel
So maybe it's good for individual investors to put their money in passive investments, but what about the health of the market as a whole? Is it hampered by how much money goes into passive investments?
Brett Harrison
It's hard to tell what it would look like if there was less money in passive investment now. I do think one of the potential downsides is creating extra correlated activity between instruments purely by virtue of them being included in index products. When Tesla gets added to the S&P 500, Tesla doesn't suddenly become a different company whose market value is fundamentally changing, yet it's going to start moving very differently in terms of its beta correlation between other instruments in the S&P 500, purely as a function of all the passive investing that moves these instruments in the same direction. So that's the sense in which I think it could be detrimental.
Dwarkesh Patel
But naively, you would assume that the efficient market hypothesis would say that if people know that Tesla's stock price would irrationally climb when it's included in the S&P 500, then people would short it and there should be no impact from this irrelevant information. Why isn't that the case?
Brett Harrison
It probably mostly is. I think sometimes there can be liquidity differences that cause at least temporary dislocations in stocks. The simplest example is an ADR, an American depository receipt that's fungible for some underlying foreign stock. These two things should be almost the same value at all times, net of currency conversion and conversion ratios. But if one of the markets is highly illiquid or difficult to access, then there will be dislocations in price. That's the job of the Jane Streets of the world to arbitrage away the price over time. Long run, you wouldn't expect these things to be dislocated for that long. I'm sure there are people who understand the fundamentals of individual names in the S&P 500, and when there's American news and the entire S&P falls, they are maybe buying S&P and selling individual names, expecting that relative value spread to come in over time.
Dwarkesh Patel
Speaking of these firms, you don't have to tell me specifics, but how similar are the strategies for market making or trading that Jane Street versus Citadel and these firms? Is it the same sorts of strategies, or are they pretty different?
Brett Harrison
I think there are a lot more differences than people appreciate from the outside. Different companies establish different niches and areas. Jane Street established its early niche in ETFs at a kind of mid-frequency level—not ultra fast, but not long-term, year-long discretionary macro. Whereas a Citadel Securities kind of firm built their niche more on lower latency options market making. It could be all over the place. Some firms are trying to optimize for really short-term microstructure alphas, trying to predict where the order book is going to move over the course of milliseconds to seconds. Other firms care more about the relative convergence of instruments over hours to days. There are sophisticated quantitative trading firms doing longer-term trades, days to weeks to months.
A lot of the infrastructure can be similar. Either way, you need to be able to connect to exchanges, download market data, establish simulation platforms, build tools for traders to grasp what's going on in the market, and especially visualize their own proprietary models and alphas. But beyond that, the actual strategies and the ways they make money can be very different.
Dwarkesh Patel
Famously in other kinds of development, there are very famous hacks and algorithms. In gaming and graphics, John Carmack has the famous fast inverse square root for doing graphics calculations and normalizing vectors faster. You were not only a developer in finance—I know the exact term for that—but you led teams of hundreds of people doing that kind of development. Are there famous examples like this in finance, the equivalent of fast inverse square root, but for the kinds of calculations you guys do?
Brett Harrison
Yeah, they're all over the place. There are tons of hacks and tricks. For example, here's a famous one—not famous, I think I read it in a paper and a bunch of other developers from other companies told me about it. It's not something I saw at places I worked. But if you're sending a message to, say, Nasdaq to buy stock and you want to get there as fast as possible, what is a message to Nasdaq? It's a TCP/IP wrapped message with a particular proprietary protocol that Nasdaq implements. Let's say your goal is to trade Apple, but you're not sure what price and at what time, and you're waiting for some signal to buy Apple as fast as possible. What you can do is pre-construct the entire TCP/IP message: first put the TCP header on there, then the IP header, then the kind of outer protocol that Nasdaq specifies and the inner protocol, except for the byte slot where you put in the price. Then pre-load that message into the network card's sending buffer so that once you're ready to send, you can just pop in the price and send it off, incurring as little latency as possible.
Dwarkesh Patel
That's awesome.
Brett Harrison
I think the analogy to video games is a good one. Just like in video game graphics, the end goal isn't to produce the most theoretically perfect simulation of environmental graphics; it's to have something that looks good enough and is fast enough for the user. That's also true in HFT quantitative finance, where the goal is to get to the approximately right trade as fast as you can. It's not to have the perfect theoretical model of underlying price dynamics.
Dwarkesh Patel
That is so fascinating. But this raises an interesting question. If you have some algorithm like this that gets you a few nanoseconds faster to the Nasdaq exchange, and that's why you have an edge—or you've leased microwave towers to get from New Jersey to Chicago faster, or you've bought an expensive server in the same place Nasdaq is housed—what fundamentally is the advantage to society as a whole from getting that information faster? Is this just a zero-sum game of who can incorporate that signal faster? Why is it good for society that so many resources and so much brainpower are spent on these kinds of hacks and optimizations?
Brett Harrison
If you start from the premise that having liquid, tight, efficient markets is important for the world, and you ask how to design a system that optimizes for that, I think you want smart, sophisticated technologists competing at the margins. Of course, the more they compete, the smaller the margins become, to the point where the little extra activity people are doing to get slightly better doesn't seem to be greatly affecting the whole system as much as it was in the earlier days when things were slower and tick sizes were wider. It's difficult to imagine designing a market where you say, "Everyone should innovate up until this point and then stop competing and stay in stasis." Maybe you can create certain regulatory or market structures to try to prevent that, but I think on average, you want people competing at the margins, even if they seem minuscule.
At the same time, I think it's not zero sum for society for technologists to be creating super fast, ultra low latency, very sophisticated algorithms. We have a lot of geopolitical instability in the world. Who knows if our microwave network that we built out in the US could have greater use cases than just for quantitative finance? Quantitative finance subsidized the creation of these towers.
Dwarkesh Patel
That's a contingent potential benefit. People tell a story about NASA, where they subsidized a lot of the science that ended up becoming products. That's an interesting account of the benefits of finance—that whatever tricks they come up with might be useful elsewhere. But that's not a story about how it's directly useful to have nanosecond-level latency for filing your Apple stock. Why is that useful directly?
Brett Harrison
If there is some kind of news that happens in one part of the world and that should affect the current price of stock in a different part of the world, if you care about efficient markets, you want the gap between source of truth events and ultimate price discovery to be as small as possible. If you want to question whether getting a few extra milliseconds or microseconds or nanoseconds is worth it, you're then putting some kind of value judgment on what the optimal time is to get from event to price discovery, saying a second is too slow but a millisecond is too fast, or a millisecond is too slow but a microsecond is too fast. I just don't think we're in a position to do that. I think we always want as close to instantaneous price discovery as possible.
Dwarkesh Patel
I'm only asking more about this because it's really interesting. There are some level of resources where we would say it's not worth it. Let's say five trillion dollars a year was spent on getting it down from two nanoseconds to one nanosecond. I know that's not a realistic number, but there's some margin where for some weird reason, society is spending too many resources on it. Or would you say we haven't reached that margin yet, where the amount of brainpower and resources spent on getting these tight spreads isn't socially useful?
Brett Harrison
I don't know how large a percentage of GDP prop trading is. I suspect it's not that large, so I don't think we're close to that theoretical limit where I would start to feel it's a waste. But I also think there's a reason why they're willing to spend the money on this kind of technology—they're obviously profiting from doing so. It has to come from somewhere. The market is subsidizing the creation of this technology, which means there's still ability for value capture, which means there's still a service being provided in exchange for some kind of profit. I think we wouldn't spend $5 trillion on a microwave network because there isn't $5 trillion of extra value to be created in doing so.
Dwarkesh Patel
Got it. Has being a market maker changed your views about civilizational tail risk? You're worried about personally getting run over by some weird event in adverse selection. Does that change how you think about societies getting run over by a similar thing? Or are the mental models isolated?
Brett Harrison
Working in high-speed finance teaches you to understand how to more correctly estimate the probability of rare events. In that sense, working in finance makes me think more about the likelihood of civilization-ending problems. But it doesn't suggest different solutions. There's a very big difference between being in a financial setting where your positions are numbers you can put in a spreadsheet and you can model what happens if every single position goes against you three times the wrong way, and what instruments you would have to buy or sell to hedge that portfolio. That's a closed system you can actually model and do something about.
Having a trader mentality on future pandemics doesn't help you much. It might slightly change your ability to estimate the probability of such events, but the actual solutions are a combination of collective action problems and being able to model the particular type of unknown unknown about whatever the event is. Those kinds of solutions should be left to the experts in those particular fields, not up to traders. I don't think having the trader mentality around rare events in normal civilization outside of finance helps you much. Maybe in some ways it's led people to think more hubristically that they can do something about it.
Dwarkesh Patel
Gee, who could it be talking about? [Laughs] That's really interesting. You would say that. Famously, these market making firms really care about having their employees be well calibrated and good at thinking about risk. I'm surprised you think the transfer between thinking about that in a financial context and thinking about that in other contexts is that low.
Brett Harrison
Again, I think it helps you at estimating probability of rare events, but it does not translate super well to what action you take in the face of knowing those rare events.
Dwarkesh Patel
Were your circles or people in finance earlier to recognize the dangers of COVID?
Brett Harrison
That's a good question. I think people in my circles were quicker to take action in the face of knowing about COVID. A lot of people who stuck around in cities, not knowing where this was going to head long term—if you have the fortune of having the financial flexibility to be able to do something, a lot of people in financial circles immediately recognized there was this big risk, this unknown, and they didn't want to get adverse selected against in terms of being able to get out of the locus of bad pandemic activity. People were fleeing cities faster than other people.
Dwarkesh Patel
That seems to point in the opposite direction of them not being able to estimate and deal with geopolitical risk.
Brett Harrison
Well, there you have an actual event that has occurred, and then in the face of the event, what do you do right now? That's different from what do we do about the potential for AI to destroy civilization in the next hundred years, or what do we do about the next potential biological weapon or the next pandemic that could occur.
Dwarkesh Patel
Speaking of COVID, you were head of semi-systemic technology at Citadel when COVID hit, right?
Brett Harrison
Yes, exactly.
Dwarkesh Patel
How did these HFT firms react to COVID? What was it like during COVID? Obviously the market moved a lot, but on the inside, was it good, bad?
Brett Harrison
All the companies—Citadel Securities, but really all the ones in this finance sphere—were extremely resilient. A lot of them found that their pre-existing ideas that for the team to succeed, everyone needed to be in the exact same place, and it was very important from an IP perspective to make sure people weren't taking work home with them, completely went out the window. People had to completely adjust to the idea that actual trading teams used to having eye contact with each other at all times needed to adjust to this pandemic world. They largely did. At least from a profitability perspective, it was some of the best years of HFT firms' P&Ls in recent history.
Dwarkesh Patel
Matching engines already have to deal with the fact that you can have orders coming from Illinois and orders coming from Japan, and given light speed, they're not going to arrive at the same time. You still have to work around that. Is there any hope of a single market and matching engine once humanity goes interplanetary or interstellar? Could we ever have a market between us and Alpha Centauri or even us and Mars? Or is the lag too much for that to be possible?
Brett Harrison
Without making any changes to a matching engine, there is nothing that says when an order comes in, it can't be older than X time. What it does mean is that if the sender is sending a market order from halfway across the world, by the time the order reaches the exchange, they might end up with a very different price than the one they were expecting when they sent it. There's probably a lot of adverse selection sending a market order from halfway across the world versus in a co-location facility. So you can technologically run an interstellar exchange, but it might not be good for that person living on the moon.
Dwarkesh Patel
Is there any way to make it more fair?
Brett Harrison
There's a real-world analog of that: automated market makers on slow blockchains. If you're used to working on Nasdaq, which processes a single message in somewhere between tens and hundreds of nanoseconds per order, a blockchain like Ethereum processes 15 to 50 messages per second—significantly slower by orders of magnitude. Yet they've been able to establish pretty mature financial marketplaces by saying that rather than sending orders with prices on them and canceling them when the prices aren't good anymore, there will be an automated function that moves the prices at the matching engine. Whenever your order reaches the exchange, it will always be a predetermined fair price based on the prevailing liquidity at the time. One can imagine building a Nasdaq for an interstellar market is kind of similar to building Uniswap now on Ethereum in terms of order magnitude and speed. There are other things you can do too, like establishing periodic auctions instead of continuous matching, which could help mitigate some of these issues.
Dwarkesh Patel
That's actually something else I want to ask about. What is your opinion of periodic, frequent batch auction systems? Should we have more of that instead of continuous matching?
Brett Harrison
In theory, they help mitigate the advantages of high frequency trading. If there's going to be an auction every 30 seconds and it's by price priority, not time priority, it doesn't matter if you send that order at the beginning of the 30-second period or the end—the price determines whether you get filled, not your particular latency to the exchange. In practice, the couple of exchanges around the world that used to have those have switched away from them. The Taiwan Stock Exchange used to have a periodic auction system, and they thought the liquidity and price discovery wasn't good. It was complained about a lot, and they eventually moved to a continuous matching system. So in practice, it doesn't quite work as well, but it's hard to tell.
Dwarkesh Patel
In your long experience of building financial infrastructure, what country do you feel has the best infrastructure and set up for good markets?
Brett Harrison
I would say the United States, except what's happened is US companies like Nasdaq have licensed their exchange matching engine technology to other exchanges around the world. The Nasdaq OMX technology powers a number of exchanges in Europe and some in Asia. So it's hard to say the technology is American—I'm not sure who wrote a lot of the stuff underneath the Nasdaq technology. But I do think the US markets are some of the most efficient, low latency, and expansive in terms of products.
Dwarkesh Patel
How do adverse selection in trading and hiring differ?
Brett Harrison
In hiring, there are many more opportunities for positive selection versus the negative selection you usually encounter in finance. The other thing is that most financial markets in the US are liquid markets. The hiring market is highly inefficient. The pipeline of orders from Harvard, MIT, Princeton, Yale to Jane Street and Citadel Securities is a very liquid pipeline, but there are many universities and colleges throughout the country and the world that have extremely talented individuals whose resumes will never end up on your doorstep. You might get a resume from a graduating senior from college with no internship experience, and your trader mindset might think this is terrible adverse selection. But it could be that if that person is willing to put themselves out there and apply to your company from a relatively unknown university, that might be the signal that they are the best person in that entire region. That might be a positive selection. So I think it's not exactly the same adverse selection dynamics as in the traditional trading world.
Dwarkesh Patel
Especially if you have mission-oriented companies, they have a particularly good way of getting rid of adverse selection, right?
Brett Harrison
Yeah, exactly. Companies with really strong brands—that's one of the things we saw at Jane Street. I heard stories from the old days of Jane Street that the first resumes from Harvard were terrible. People couldn't do basic math. They were the worst candidates compared to other people they were able to find. Then they established this brand and recruiting pipeline and reputation for having very difficult interviews, paying people really well, and having an amazing work environment. All of a sudden, all the people getting through the pipeline from Harvard were really, really great. It wasn't that the quality of students at Harvard changed—there's probably a bell curve there like everywhere else. It was the positive selection resulting from the branding efforts and the mission-driven focus of the company that brought that positively selected pipeline to them.
Dwarkesh Patel
Should Jane Street replace OCaml with Rust?
Brett Harrison
No, because there's too much infrastructure already in OCaml.
Dwarkesh Patel
Starting from scratch, if they could snap their fingers and suddenly replace all their OCaml infrastructure with Rust at zero cost, would it be worth it?
Brett Harrison
In that case, I would say yes, because you get a lot of the static typing and compile-time safety in Rust that you get from OCaml, but the base level program you can write in Rust is much, much faster than one you can write in OCaml. Because of the way OCaml is designed, with automatic garbage collection, the worst thing you can do in high-speed finance is any memory allocation that results in garbage collection. You have to write very careful OCaml that almost ends up looking like C in order to stay in functional programming land but not create tons of memory on the heap or stack that gets collected later.
Dwarkesh Patel
You've been playing around with Rust a lot recently, right? What is your impression of the language? Have you been enjoying it?
Brett Harrison
It's great. It's come a very long way in the last three to five years. I think crypto has something to do with that—it seems to be one of the languages of choice for people writing blockchains and smart contracts. There's been an enormous amount of open source contribution to Rust. Compared to when I last looked at it a couple years ago, it's a lot easier to write really good, sophisticated programs now in Rust and get all the type safety and speed, which is very comparable to C++ on the speed side.
Dwarkesh Patel
When I'm writing programs, they're not large code bases with many people contributing. When I use Rust, it's a huge pain—why did they have to put an arc inside of a box instead of an option? But I can totally understand if you have something where billions of dollars are at stake.
Brett Harrison
There's definitely a learning curve. For basic scripting, you want to use something like Python. But if you're writing low-latency distributed infrastructure that has to never fail, Rust is a pretty good choice.
Dwarkesh Patel
Speaking of Jane Street, why does a company pay interns like $16,000 or $17,000 a month for their summer internships? Is the opportunity cost for one of these smart students that high in the summer?
Brett Harrison
The short answer is yes. The long answer is that the starting salary for the top people in not just finance but tech is in this low to mid six figures now. You can debate whether you think that's the appropriate starting salary for a person with no experience coming out of college, but the reality is the talent pool is extremely competitive from the employer side. If you start with that as a reasonable salary plus bonus for an employee, and Jane Street's mentality is that these interns are doing real work, they should be paid like a full-time employee, prorated for the time they're actually here. That ends up checking out to be the right numbers.
Dwarkesh Patel
On net, are interns—forget about the salary—actually contributing, given the time of the traders who are training them?
Brett Harrison
Maybe it sort of breaks even if you consider the time to train them. But it's extremely worthwhile because when those interns come back full time—and Jane Street hires a significant percentage of its incoming classes from their internship program—they're already trained. They're ready to go day one. They're almost immediately useful because they had that three-month period where they got trained, and only the ones that really liked it and were good come back. Rather than wait for them to come on site, train people, and maybe half of them aren't good or don't fit with the culture and you don't know what to do with them, the internship program provides a really good place to get on-the-job training and then select only the best from both sides.
Dwarkesh Patel
Is there a free rider problem with internships? If a company like Jane Street puts in the effort to train somebody for three months, they might get some of them to work for them, but they've also trained some people who might work for the competition. Is there some sort of free rider problem there?
There for sure is, which is why the companies have to work as hard as possible to make their experience as good as possible, which is good for the interns. When you go to Jane Street, not only do you learn a lot, but they pay you really well, and you get to visit one of the foreign offices for like a week or something. They also have all these really fun programs where they bring famous speakers to come to the office and speak to the whole intern class, and they have parties and all sorts of stuff. It adds to the experience of thinking, "Okay, this is the place I want to work. I don't want to take my training and go to the competitor. I want to come to you."
Brett Harrison
Yeah.
Dwarkesh Patel
I clearly got in the wrong business with podcasts. Why did you pursue finance dev instead of trading? What appealed more to you?
Brett Harrison
In college I studied computer science and math, and I really liked programming, but I think I didn't quite know what a career in programming looked like. The conventional wisdom, at least in 2009 when I was applying for internships, was that I was going to sit in a cubicle and stare at a screen for 16 hours a day, and I was going to be miserable. It wasn't going to be a very social job, and I consider myself a pretty social person. I had a lot of friends who had various internships in quantitative finance, mostly from the trading side. So when I eventually went to Jane Street as an intern, I had a hybrid summer of doing some dev stuff and some trading stuff. To me, I thought the traders were much closer to the real action of the company, and I wanted to be a part of that. When I joined Jane Street, I was hired as a trader on the ADR desk. I realized very soon into that that, one, the developers have just as much if not more impact on the success of the company, and two, I just enjoyed it a lot more. It was much more in my alley and my training, so I ended up going that route instead.
Dwarkesh Patel
I want to ask about the culture at these sorts of places like Citadel or Jane Street. You spend some time in Silicon Valley and around the traditional startup scene as well. What is the main difference between Silicon Valley tech culture and quant New York culture? Or I guess Chicago.
Brett Harrison
I have a ton of personal experience in the Silicon Valley culture or the tech culture, since I've only really worked at finance places my whole life. But the sense I get is that the New York, Chicago quant finance dev culture is one about extreme pragmatism. You know what the outcome is: be the most profitable at the strategy, and you try to draw a straight line between what you're doing and that profitability as fast as you can. Compared to that, I think the Silicon Valley culture is much more about creativity and doing things that are new that no one else has done before. A healthy amount of cross-pollination would be good for both.
A lot of trading firms are doing the exact same things that all the other trading firms have done, and a healthy injection of creativity to think slightly outside the box—like getting slightly faster to go to NASDAQ or something, which is probably fine but not that creative—would be good for those quant firms. At the same time, the sheer approach to pragmatically getting something done and out there and sold and making money would help a lot of Silicon Valley firms that hang out in this creative land for too long and don't end up getting a product to market.
Dwarkesh Patel
Yeah. It seems like there should be one founder from both those cultures in every single startup. It's similar to what you were saying earlier with SBF and visionaries versus pragmatists in that context. How conspicuous—you were just mentioning that these traders are making mid-six-figure salaries to begin with, let alone where they rise over their careers. How conspicuous is their spending and lifestyle? Is it close to Wolf of Wall Street, or is it just Walmart t-shirts? Where are we talking?
Brett Harrison
It's probably a lot closer to Walmart t-shirts than it is to Wolf of Wall Street. Certainly it is now. Even when I started, it was pretty inconspicuous. I don't think it was that way in the previous decade or two before I joined finance. I'm not really sure, but I get the sense that the current culture around inconspicuous consumption is a function of millennial consumption habits, where people are focusing a lot more on experiences than having shiny material objects. I think that's had a large effect on the high-earning tech and finance culture that exists today.
Dwarkesh Patel
Are they spending that much money on experiences either? How expensive is a flight to Hawaii? Even after you subtract that, where is this money going? Are they just saving it?
Brett Harrison
Maybe it's not just a flight to Hawaii, but it's bringing your 10 friends to Hawaii with you. Or it's getting involved in a charitable organization in a way that someone who is 24 normally wouldn't be able to do, purely by being able to donate a lot.
Dwarkesh Patel
What is the social consequence, for lack of a better word, of having a bunch of young, nerdy people, often male, often single, having this extraordinary level of wealth? What impact does it have? I don't know if "society" is the right word, but what is the broader impact of that class of people?
Brett Harrison
I think we'll have to play this out over the next decade or two to really see where this goes. If I'm going to be an optimist about this, I'd like to think that when it was older, single or married males were hoarding a large amount of wealth, for the most part they kept it to themselves and waited till later in life to do anything with it. They were the kind of people who really saved their entire careers. As opposed to if younger and younger generations are amassing wealth through what they can actually perform with their skills, then I think that hopefully injects more dynamism into the distribution of that wealth later on. Those millennials or Gen Z will go on to found new companies, and they'll be able to seed the companies themselves with their own money and have a much easier time bringing interesting new things to market. Or they'll be able to donate to really interesting causes, or help out their friends and family more easily from a young age. They'll be more selective in the kinds of things they give to or contribute to that don't just involve getting their name on a building of a school or something.
Dwarkesh Patel
Yeah, that's a very optimistic story. I hope that's the way it plays out. Tell me about the psychology of being a quant or a trader or a developer in that space. You're responsible—one wrong keystroke and you've lost millions of dollars, one bug in your code. There are historical cases of this, where entire firms go down because of a bug. What is the day-to-day psychological impact of that kind of responsibility?
Brett Harrison
Maybe the job selects for people who don't crumble under the theoretical stress of that job. But personally, I don't lose sleep overnight over that because within any mature financial institution like a trading firm, there are typically many layers of safeguards in place. There are limits on how many dollars you can trade in a minute, how much you can trade overall or for your desk, how many messages you can send to the exchange. There are limits at the individual trader, desk, and firm level. There are layers of different checks. Often there are actual regulatory rules to comply with, like market access checks, like FINRA rule 15c3-5.
When you're writing new code, it's not a completely blank slate thing where you're connecting directly to an exchange and hoping for the best. Usually you're embedding some piece of code within a very large established framework where the goal is to make something trader-proof. No matter what some trader clicks on or does or configures with their system, there's a limit to how badly they can actually go. Especially in my particular role as a developer, being able to understand the technological stack and verify that these particular safeguards are in place and it is as trader-proof as I think it is, I can sleep at night knowing nothing too bad is going to happen.
The times I actually lose sleep are when a trader in London or Hong Kong calls me in the middle of the night to say, "Hey, can you explain how this thing works? I need your help." Those are the times where I actually lose sleep, but not over being concerned about risk.
Dwarkesh Patel
That's interesting. If you asked the people who work in these firms what the social value they're creating is—separate from the question of what the correct answer to that question is—what would the majority of them say? Would they say, "I'm doing something really valuable," or "I'm indifferent to it, but it's earning me a lot of money"? What is their understanding of the value they're creating?
Brett Harrison
It really depends on the company and how diffuse the culture is. At older firms that have fewer people impacting the culture in any significant way, you might not get a clear answer. For a place like Jane Street, where the firm is really run by 30 or so partners and senior employees who have been there for a really long time and have carried through the core culture up to the present day, and with that large number of people at the top in a very flat environment have been able to propagate that culture and maintain it throughout the company, I think you'll find a much more homogeneous view on their social value. I think they would say that they provide the best pricing and access to markets that are critical for facilitating capital allocation throughout the world, and allow people to very efficiently invest in vehicles that are global in nature.
Dwarkesh Patel
That seems very abstract. While it is probably correct and very valuable for society, it might not seem that tangible to somebody working in that space. Is there some technique these firms have of making visceral the impact these traders have? Do they bring out some child who benefited from efficient markets or something?
Brett Harrison
Probably not children. I think it's more anecdotes about the pension fund behind a state government that needed to get exposure to some diversified asset class and came to one of these companies and said, "We want to move a $5 billion portfolio. Can you help us do it in an efficient way?" It ends up saving them significant numbers of percents or basis points over what would happen if they went to the market. You can say there's a direct connection between the price that someone like Jane Street gives them and the amount of money that they ultimately get to save and pass on to the people in their state who are part of their pension plan. There's a direct connection there.
Dwarkesh Patel
Okay, let's start by addressing the elephant in the room, which is FTX. Let's begin at Jane Street, which is where you met SBF. Can you tell us the origin story of how you first met him and what your first impressions were?
Brett Harrison
Absolutely. I was at Jane Street from 2010 to 2018. Sam was at Jane Street for a couple of years in the middle of that, I think 2013 to 2017. One of the things I did at Jane Street was I started this program called OCaml Bootcamp. It was a yearly course for the new trader hires to spend four weeks with me learning programming in OCaml, which was the esoteric programming language that we used at Jane Street, along with a lot of our other proprietary systems. Sam was in one of the first cohorts of students, so I got to meet him through that experience.
Dwarkesh Patel
What was your impression of him?
Brett Harrison
He was a smart kid. He was nice. He got along well with other people in his class. He was definitely above average, but not completely standing out at the top. Then again, the bar was extremely high at Jane Street, so I think that's already a compliment. People liked him a lot and thought he had a lot of promise, but he was a young guy like everyone else.
Dwarkesh Patel
Did that perception change over time while you were at Jane Street?
Brett Harrison
It slowly started to. Sam was on one of the largest trading desks at Jane Street, with 50 or 60 people on it. He had several managers. One of my roles at Jane Street was to work with all of the different trading desks on the designs of their particular strategies and systems. I would frequently go over to his desk and talk with his managers about stuff. I started pulling him into conversations more and more, specifically to talk about some of the lower latency ETF stuff we were doing and some of the original OTC automation things we were working on. He started actually contributing more to the actual design and thought behind some of these systems. I thought he was precocious and had a lot of really good intuitions about markets.
Dwarkesh Patel
What exactly was your role at Jane Street at this time? And what was his?
Brett Harrison
At this time, I was sort of leading the group of software developers building the technology that was closest to actual trading. You can think of it as HFT. At any trading firm, there are lots of different developers—people who work on stuff relating specifically to the trading technology, people who work on core systems, networking, internal automation tools, tools for developers. We were on the part of the spectrum that was closest to actual trading. My job was to go over to the different trading desks within the company, talk to them about their specific strategy for the products they traded, understand their priorities about what venues they want to connect to, what different systems they want to create, what different parameter changes they need in their automated trading systems, what kind of research tools will help them do their job better, what user interface would make it easier for them to understand what's going on in the market, and all of that.
Dwarkesh Patel
Did SBF at this point have any sort of reputation of being either uncooperative or cooperative, or anything ethical or professional that was noteworthy?
Brett Harrison
I don't think there was much that stood out. He was, again, pretty precocious at that particular time period. One anecdote that sort of drew me closer to him: Jane Street's offices were in 250 Vesey in New York City, and there's a big food cart on the second floor. I once went down to meet with a development person from a nonprofit that works in animal welfare, something that my wife and I had donated to for a long time. I met with this guy and he said, "You're the second person I met from Jane Street today," which was wild because Jane Street was only a couple hundred people. This was a pretty niche organization. I said, "That's crazy. Who did you meet?" And they said, "Sam Bankman-Fried." I said, "Sam, I just came down from talking with him upstairs." I went back and we sort of realized we had this shared interest in helping animal welfare causes. We were both vegans, and we bonded over that. That's how we became friendly.
Dwarkesh Patel
So it seems his interest in effective altruism was genuine at this point and early on. There was a history of this.
Brett Harrison
Yeah. It wasn't like EA was super popular at Jane Street. I feel like that's a bit of recent sampling bias among the younger crew of Jane Streeters. There were a couple people there who were fairly vocal about donating the majority of their yearly salary and bonus to charitable causes. Sam was one of them, and he started to become known for that.
Dwarkesh Patel
Fast forward to when he's no longer at Jane Street, you're no longer at Jane Street, and you're at Citadel. He started FTX. Actually, before we go there, were you in contact with him up until the point where you started talking about a potential role?
Brett Harrison
Off and on. When I first left Jane Street and he left—sorry, we both left Jane Street around the same time, him before me—he had told everyone at Jane Street that he was leaving to join the Center for Effective Altruism full time. I guess he did that, but I'm not sure if it actually happened because he very soon after started this trading firm and tried to pull lots of Jane Street people to join him, which didn't make people super happy. It was funny. We had a phone call and he told me it wasn't really going super well. He said it was really great in the beginning—they made a lot of money, they had this arbitrage trade—and then a few things kind of went by the wayside.
They had taken out these huge loans to get their initial capital for Alameda, and there was a big fracture within the company. Half the company split. People left. He didn't tell me much about that at the time. He said he was probably going to do something else. When I asked him, he said, "I think I'm going to work on political prediction markets." I thought, "Okay, that doesn't sound super exciting to me. I'm going to continue on with what I was doing, which was moving to Chicago and taking a new role." But fast forward, I guess that idea—maybe he wasn't telling me the whole truth at the time—became FTX, and he had supposedly resuscitated Alameda in the process.
Dwarkesh Patel
Do you have some sense of what it was that went sideways?
Brett Harrison
I pieced together some details over the years because he told me a little bit more after I first joined. I heard a little bit more later from other people and then saw some reporting post-FTX collapse. I think there were two things. One was the infrastructure they had built was really poor in the beginning—a lot of Python scripts slapped together. A couple of times they had sent tokens to the wrong wallet and ended up losing millions in the process. They had some big non-arbitrage directional bet on some token, maybe ETH or something, and it went against them, so they lost a lot of their trading capital. The other thing was that after some of their technical problems, there were internal disagreements, supposedly—this is what Sam told me—about how to move forward with tech. Half the crew wanted to rewrite everything from scratch in a different programming language. The other half said, "Okay, we can make some small incremental changes from here and fix things up." Sam and Gary and Nishad were more in that latter crew. The former crew kind of broke off and started their own thing. That's what originally happened.
Dwarkesh Patel
Were you aware of the extent of this at the time, or is this something pieced together?
Brett Harrison
Not at all. Sam told me a little bit about it, but this was over the course of years where I had two different roles—one at Headlands, one at Citadel Securities. Sam was starting Alameda. We spoke maybe once a year briefly on the phone. All this stuff was happening in the background and I had no clue. The first time I even heard about FTX was when one of my colleagues from Citadel Securities told me, "Hey, did you ever work with this Sam Bankman-Freed guy?" I said, "Yeah, a little bit. Why?" And they said, "He's a billionaire and he has this Hong Kong crypto exchange." I said, "What? Since when?" I started to see him pop up in articles. There was a Vox article about him and a few other things, especially related to his political donations. That's when I got back in touch with him and we started talking a little bit.
Dwarkesh Patel
When was it that he called you to say there were potentially troubles and he was considering starting a political prediction market?
Brett Harrison
That was in 2018, right after I left Jane Street.
Dwarkesh Patel
So now you've moved on to Citadel, and you're still in touch at this point?
Brett Harrison
Very briefly, a text every now and then.
Dwarkesh Patel
Then at some point you become president of FTX US. Do you want to talk about how he approached you about that and what was going on at the time?
Brett Harrison
It was interesting. At the time, I was running what was called the Semi-Systematic Trading Technology at Citadel Securities. This was the group of technologists working on systems for ADRs, ETFs, options, and OTC market equities. It was around 100 software engineers that rolled up to me. That was going well. But Sam and I started talking. It was March of 2021. He was telling me a couple things going on in FTX, and then he said, "If you're interested in coming over to FTX, we would still love to have you." I thought, "Still? We never talked about doing this before." But sure, let's entertain this. We started talking. He had me meet him and Gary and Nishad over video call. I was in Chicago, they were in Hong Kong at the time. These calls were taking place late at night my time. Very quickly, an offer came together. I thought this was a really cool opportunity to jump into a field and take a role that was very different from stuff I'd done in the past. I signed up.
Dwarkesh Patel
Where was FTX at this point in terms of its business development?
Brett Harrison
FTX was doing quite well. It was basically finished its first, its second year of operation. It was maybe the fourth or fifth largest exchange in the world by volume, if you include crypto spot and crypto derivatives. It was also one of the primary destinations for institutions—proprietary trading firms, hedge funds—to trade crypto and derivatives, especially because of how it was designed. It was doing really well. FTX US was virtually non-existent. They had started the exchange, I think, in either December 2020 or January 2021, but it had de minimis volume compared to the other exchanges around the world, especially in the U.
S. Sam talked to me a lot about the aspirations for the U.S. business. One was to grow the spot exchange, of course. Two was to find a regulated path for bringing some of these offshore products like Bitcoin and Ether futures and options onshore in a regulated way. On top of that, Sam had also told me about longer-term desires to be a single app or marketplace for everything, not just crypto—launching a stocks trading platform as well. I think that was one of the reasons he wanted to bring me on, because I had all this experience inside regulated broker dealers and knew roughly what it took to get that started.
Dwarkesh Patel
The initial offer was specifically for president of FTX US, right?
Brett Harrison
Yeah. Sam wasn't someone who loved thinking hard about titles. Even what my original title was going to be was a point of contention. I'm not sure it was clear exactly what my role was going to be. I think Sam wanted me to write software for FTX US. But to me, I thought there was this bigger opportunity to work with Sam to lead this other startup, FTX US, and build it up and follow in FTX's footsteps and its success. That was the part that was most exciting to me. This is what I've been doing for years—managing large teams of people, thinking about strategy, getting people together, occasionally doing some software development myself. That was the primary reason for wanting to join.
Dwarkesh Patel
What was the relationship between FTX and FTX US at that time? Were they subsidiaries? Were they separate entities?
Brett Harrison
They were separate entities. They weren't subsidiaries. There was technology sharing between them. The FTX US exchange technology was licensed from FTX. You can think of FTX US as FTX stripping away most of the interesting parts of FTX, because it was just a dozen or two spot tokens. When I joined, there were very few people within FTX US, maybe two or three dedicated people. Over the course of the next year or so, my job that I sort of fit for myself was to open up some offices, hire a bunch of people, establish separate compliance, legal, operational, and support teams, and start to build out these regulated entities.
Dwarkesh Patel
Was Chicago the initial base of that operation?
Brett Harrison
Yeah, for selfish reasons. My family is here, and I wasn't going anywhere. But also, I thought Chicago was a great place for FTX US because if our main goal was to establish regulated derivatives, Chicago is really the place where that happens. We have the CME. We have many of the top proprietary trading firms. A lot of the Futures Commission merchants and various brokers are all here. Historically, the floor of the Chicago Board of Trade and the Chicago Mercantile Exchange are here. It felt like a good place to be.
Dwarkesh Patel
Before you joined, did you get a chance to ask him about the relationship between FTX and Alameda?
Brett Harrison
I did. It was definitely of interest to me. The primary reason was that I wasn't interested in doing prop again. I worked at Jane Street, I worked at Headlands Tech, I was at Citadel Securities. If I wanted to continue doing prop trading, I would have stayed at one of those places. I wanted to do this exchange business. What Sam told me was the same thing he said publicly everywhere: that Alameda was basically running itself, all of Sam's time was on FTX, they were walled off from the FTX people, and their access to the exchange was just like any other market maker. There were public API feeds, there were benefits for market makers that trade enough volume, but it wasn't like Alameda had any special privileges in that sense. I thought they were just basically separate.
Dwarkesh Patel
Did you ask to audit their financials or this relationship before you joined?
Brett Harrison
No. I don't know about you, but I've never gotten an offer to a company and said, "Well, before I sign, please show me your audited financials." That's just not a thing that happens.
Dwarkesh Patel
Fair enough. So you join FTX, and you mentioned some of the stuff you were working on—the operational, legal, getting the organization set up. Feel free to talk in more detail about what things came up during your tenure and what accomplishments you're proud of.
Yeah, I guess on a professional and personal front. So I guess on a professional front, I'm most proud of establishing our team and making significant headway to a lot of our goals to establish these regulated businesses. For example, LedgerX—we acquired LedgerX and we had this application to the CFTC to enable real-time direct-to-customer 24-7 margining and cross-collateralization. It was an extremely innovative proposal, and it felt like we were making real progress towards establishing new and very exciting regimes for CFTC-regulated derivatives in the U.S. I also established a broker-dealer in the U.S. for the purposes of letting people trade stocks, similar to Robinhood. I wrote like 90% of all the code for that stocks platform myself. I was very proud of that accomplishment. And then just on a personal front, it was great to get embedded into the crypto industry. I was very excited by everything that I saw. It was great to make all the connections through FTX with the different people in the crypto ecosystem and become friends with these people. That has certainly influenced where I am today. So I'm proud of all of that.
Brett Harrison
How did you manage the management of—I don't know how big the team was at its peak—and it sounds like you were heavily involved in the actual engineering. How were you able to manage both roles at the same time?
Dwarkesh Patel
Yeah, so we were between 75 and 100 total people in the U.S. And it was challenging. It was one of my biggest complaints, which I'm sure we'll get into. Yes, I can write code. But I feel like my comparative advantage is helping leverage teams of people to get them to work towards the common goal of building out large distributed systems that are complex and multivariate in nature. The best use of my time was not me programming between 10 p.m. and 2 a.m. every night while trying to keep on board with what all the personnel were doing. So I really wanted to grow the U.S. team significantly to at least be more than a handful of developers. That was one of the initial points of contention.
Brett Harrison
Okay. Speak more about that. So he was opposed to growing the team?
Dwarkesh Patel
Sam would frequently talk publicly about how proud he was that all of FTX was built by two developers. And all of these crazy organizations that hire thousands of developers and can't get anything done, they should learn from me about how a small lean team can be much more effective. There's some truth to that. I do think the conventional wisdom now is a lot of big tech companies overhired for software engineers. Not only was it an expense on the balance sheet, but it was also expensive in terms of slowing down the operational efficiency of the organization. Having a small lean team can help you get to your first or your end product a lot more quickly.
That's great for a startup. But once you're a north of $10 billion valuation company, promising the world to customers and investors, two software developers doesn't really cut it anymore. At some point you have to grow up and face the reality that it's time to actually grow an organization into a real managed enterprise with teams of software engineers specializing in certain tasks. There was always pushback. People would tell me, "Look, we're not trying to be like Jane Street or Citadel in terms of our number of software engineers. We want to stay lean. That's our comparative advantage." And most importantly, they didn't want two separate development teams—one in the U.
S., one in the Bahamas. They wanted to keep the nexus of software development underneath Nishad and Gary in the Bahamas, which I just thought wasn't going to be sustainable long term. If you run a broker dealer in the U.S., you need to have staff that is specifically allocated towards broker dealer activities. It can't be that if FINRA comes and says, "Well, who's working on the broker-dealer?" you say, "Well, it's this Gary guy who lives in the Bahamas who sometimes is awake at 4 a.m. and spends 20 minutes a day thinking about stocks." That can't fly.
Brett Harrison
Right, right. And has no images of him on the internet. Yeah. Okay. So were Nishad and Gary contributing code to the FTX US code base?
Dwarkesh Patel
Remember, the FTX US side of things was a strict subset of FTX, so in that sense it kind of flowed into FTX US, with the exception of the FTX US derivatives. The LedgerX stuff was actually a completely separate team because that was through an acquisition.
Brett Harrison
When you're talking about the code of the matching engine or things like that, was the code shared between FTX and FTX US?
Dwarkesh Patel
Yes.
Brett Harrison
Okay. Who was in charge of ultimately approving the pull requests of the FTX US code base?
Dwarkesh Patel
It was all Gary and Nishad.
Brett Harrison
Okay, got it. And so the code you were contributing was also going to the universal global code base?
Dwarkesh Patel
Yeah.
Brett Harrison
Got it. Did you have access to the entire code base or just the FTX US side?
Dwarkesh Patel
Yeah, again, it was one shared repo. There was an enormous amount of code. One of the big problems, another problem that I raised while I was there, was that 90-plus percent of all the code of FTX was written by these two people. And it was very hard to follow. I don't know if you've ever seen a large Python code base before. Whenever there were issues that arose—like a particular problem with an account on the exchange—the only answer was, "Call the shot. Call Gary." Which I also knew to be unsustainable from an organizational perspective. One of the guiding principles at Jane Street, for example, was mentor your junior dev so that you can hand off all your responsibilities to them. In the process of handing off responsibilities, you make the code better, more automated, more robust, problems more easily debuggable in real time. If you hoard everything to yourself in your own brain, you end up with a code base that is only understandable by that one person. A lot of people talked about this internally: if Gary got hit by a bus and couldn't come to work anymore, FTX is done.
Brett Harrison
Exactly, exactly. So what do you think was the motivation behind this? Was it just that he wanted to avoid a Google growing to 100,000 people kind of thing? Or was there something else going on? Why this sort of concentration?
Dwarkesh Patel
Well, clearly there was something else going on. An open question now, only thinking about this in hindsight, was how much of this very cloistered organizational decision around the development team was a function of the various things they were doing that they were hiding from the rest of the company? Or was it really this sort of paranoia about growing too large too quickly and losing control of the organization? And two, an almost cult-like belief in this small team being the but-for cause of all past, present, and future success.
Brett Harrison
What was the discretion that you had at FTX US? It sounds like you weren't even given the capacity to hire more engineers if you wanted to. What were the things you did control?
Dwarkesh Patel
Yeah. Hiring, for example. I pushed for many months that we should hire more people. Eventually, I got permission for us to interview people. But then those would ultimately have to get finally approved by the people in the Bahamas. And they would frequently say no to people who I thought were good candidates. Finally, we hired one person. This person was doing well. He was here in Chicago, and they invited him to go spend like a month in the Bahamas to kind of hang out with them and ramp up on the system. This person comes back to Chicago and says, "You know what, I really want to move to the Bahamas. They really convinced me to do it." It was so frustrating.
Brett Harrison
Poaching from your own company.
Dwarkesh Patel
Exactly. It was such a constant battle, and at some point I kind of gave up on this idea that I was going to be able to actually grow a separate developer team. So the bottom line is, on the day-to-day operational stuff, especially the decisions within some of the things I was responsible for, like the stocks trading platform that I was working on, I had a fair amount of discretion. People certainly looked up to me for management and advice and direction. But ultimately the discretion ended up with this small group in the Bahamas who not only had final say on decisions, but would often make decisions and not communicate with the senior people on the U.S. side. We would just sort of find out things were happening.
Brett Harrison
Is there a specific example or set of examples that comes to mind?
Dwarkesh Patel
Sure. The biggest example for me was, this was sort of post my effective resignation, but some of these strategic acquisitions that were being done in the U.S. during the summer of 2022. I would find out from the news or it would sort of be mentioned on a Signal chat or something that this was happening. There was no opportunity to actually wade into the discussion about how this was going to greatly affect the U.S. business, greatly affect our priorities. It wasn't clear if this was a good decision or a bad decision. It was just a unilateral decision that was made. Like, "Oh, we're acquiring this company" or "We have the option to acquire this company."
Brett Harrison
Are there decisions that were made from the Bahamas that stick out to you as being unwise? That you tried to speak out against? I mean, you mentioned some of them, like not hiring enough people and not getting more developers. But are there other things like that that stick out to you as bad decisions?
Dwarkesh Patel
A lot of the spending. On everything from lavish real estate to all of these partnerships to very, very large venture deals. These were the kinds of things in the company where people asked, "When does this stop? To what end? Are we doing a lot of these things?" Some of those resulted in direct confrontations. Like, "Why are we doing yet another deal with a sports person or a celebrity? This is ridiculous. This is not doing anything for the company. And we're completely distracting from the role that we thought we all had, which is to build a really great core product for people trading crypto and crypto derivatives."
Brett Harrison
Yeah. And did you bring this up directly with SBF?
Dwarkesh Patel
Yeah, multiple times.
Brett Harrison
And how would he respond?
Dwarkesh Patel
Sometimes he was nice about it. He would say, "Yeah, I see where you're coming from. I do think what we've done so far has been really valuable and we probably should do some more of it, but maybe at some point we should stop." A lot of this sort of hedging language that was ultimately non-confrontational, non-committal. He was a very non-confrontational person, very conflict avoidant person within the company. And then at worst, he would just—there were other times where I brought up specific things that I thought he was doing wrong. There was one really unfortunate time where I visited the Bahamas for the first time in, I think it was November of '21.
I'm the kind of person who, if I see something wrong at a company, it doesn't matter what company I've worked at or how junior or senior I've been, I like to go to the person most senior in charge and tell them, "This thing seems wrong to me." I feel like it's one of my superpowers. I'm just not afraid of saying when something seems wrong to me. Sometimes I'm totally wrong and don't understand the full picture. And sometimes it results in something better happening and people will thank me for having been honest. So I said to Sam, "I think you're doing way too much PR and media. First of all, it's really diluting you and the FTX brand to constantly be doing TV interviews and podcasts and flying to banking and private equity conferences.
It was so much time spent on this stuff. And also, it was completely taking away from the management of the company. People would sometimes send Sam Slack or Signal messages and not get responses for weeks at a time. It felt like he was spending virtually no time helping the company move forward. It was so much about image and brand and PR." And he was really angry at hearing this criticism directly.
Brett Harrison
How did he react?
Dwarkesh Patel
He was sort of emotional. He was worked up. He told me, "I completely disagree with you. I think you're completely wrong. I think the stuff that I've done for PR is maybe the greatest thing that's happened to this company. I should do more of it." I didn't think it was physically possible to do more of it. And I realized at that moment that this was not really going to work super well long term. If we're not in a relationship where I can give my direct superior real honest, constructive criticism that I thought was for the good of the company, then this wasn't really going to work.
Brett Harrison
He actually did my podcast about eight months ago or something. While I was very grateful he did it, even at the time I'm like, "I don't know if I would have agreed to this if I was in charge of a $30 billion empire." Yeah. I mean, sometimes reporters would say to me, "Can you get me in touch with Sam?" And I would say, "Why? I'm not really his keeper. You can contact him yourself." They'd say, "Oh, because we want to come to the Bahamas and do a special on him." And I would say, "Okay, you're going to be like the sixth one this month. There's no exclusives here." So I guess to steel-man his point, he did get a lot of good PR at the time, right?
Potentially—well, not potentially, like actually—too much, and in a way that really created the "king of crypto" sort of image. So, I mean, was he right about the impact of the PR at the time? Or maybe let me ask the question a different way. How did he create this image? I mean, people were saying that he's the JP Morgan of crypto, like he could do no wrong. Even things that in retrospect seem like clear mistakes, like only having a few developers on the team, were universally praised. Huge empire run by a few developers. How was this image created?
Dwarkesh Patel
I think that media was primed for the archetype that was Sam, this sort of young upstart prodigy in the realm of fintech. We have a lot of these characters in the world of big tech. And I think that he had a particular role to play in the world of finance. By making himself so accessible all the time, he gave people a drug that they were addicted to, which was that constant access. Any time of day or night, someone could text Sam and get him on the phone with them if they were in media. And they loved it. It was like getting access to a direct expert who was also this famous person, who was also this billionaire, who was also this extremely well-connected person, who was also this very insightful person who knew a lot about what was going on in the industry and could give them insight and tips.
I think there was some amount of what I like to call "reputation laundering" going on here. You get the famous celebrity to endorse Sam, which makes this politician think highly of Sam because they also like that celebrity. And then the investors are writing really great positive things online about it. But also the media is enforcing how cool it is that Sam is doing all these other things. It all fed into this flywheel of building up Sam's image over time in a way that didn't necessarily need to match the underlying reality of who he was at the company.
Brett Harrison
And what was the reaction of other employees at FTX to this media hype train, and also the amount of time Sam was spending with the media?
Dwarkesh Patel
On one hand, I think people were growing frustrated within the company because of the lack of direction and some of the power vacuums that resulted from Sam's continual absence. On the other hand, so many people within the company just hero-worshipped Sam. When you hear all the really tragic stories now of all the employees who kept all of their funds and life savings on FTX, they really, really believed in Sam. It doesn't matter how little time he spent with the company. It doesn't matter how he treated employees internally. He was this sort of genius pioneer. And that image couldn't be shaken. I certainly don't blame anybody for it. I interviewed him. I tried to do a lot of research before I interviewed him. And I certainly was totally taken with this. I thought he was the most competent person who had ever graced crypto.
Brett Harrison
But so what was he actually like as a manager and leader? Other than the micromanaging aspect of it, or feel free to speak more on that as well. But in terms of the decisions he would make in terms of business development and prioritizing things, can you describe his management style and leadership?
Dwarkesh Patel
In the beginning, when I joined FTX, my initial impressions were that he had pretty clear intuition and insight into the simple things to do that would work. In many ways, if you think about what FTX did, it wasn't really super complicated. It was just: be operationally good and give your trading customers as predictable of an experience as possible with regards to collateral management, auto liquidation, matching engine behavior, and latency. And so they did it. I would say, aside from the intuition, Sam wasn't a details man. That was usually left up to the people below him to really take care of—to drive a project to completion, to figure out all the details that had to be done.
I think besides that, as a leader, I thought he was fairly incompetent. I thought he was very conflict avoidant. He didn't like to get into direct confrontation with any of his employees. Most of the reasons why people needed to talk to him were because there were significant issues, whether those were personnel or otherwise, and he just blew them off. That was a frequent occurrence within the company. If you went to the Bahamas—I went only a couple of times to actually visit the office—if he was in the office, he was there all day on calls. All day. Whether those were with investors or with media, podcasts, whatever.
It was just consistently doing that. I saw very, very little time where he actually got up and talked to anyone else within the company about anything. So to me, that was the primary impression I got of his leadership: virtually that there was none. Which made me feel a lot like I and others needed to step up and sort of take that role in the absence.
Brett Harrison
Got it. And so who was making these day-to-day decisions in the absence of Sam?
Dwarkesh Patel
On the foreign side, in the Bahamas, Nishad was really the number two person there. He was making a lot of decisions. There were a couple others in the Bahamas who were taking kind of swaths of the company, whether it was investments or marketing or legal, things like that. On the U.S. side, we had a different crew trying to make decisions where we could on U.S. regulated matters. But again, we were always sort of below the decision-making authority that was happening in the Bahamas, especially inside the home where they were all living.
Brett Harrison
So it seems like FTX was a really good product compared to other crypto exchanges. I've heard a lot of traders praise it. Was this built while SBF was still doing media stuff, or was this built before he kind of went on the PR train? How was this product built while the CEO was kind of distracted?
Dwarkesh Patel
I think the core of the product was built before my time. My understanding was, in the transition from Alameda to FTX, where there was no publicity around Alameda, there wasn't any publicity around FTX, it was very much heads-down build mode for several months and just think, think, think about the core product. Having been a trader on these different exchanges around the world that also offer derivatives and knowing all their problems—for example, if you had an Ether futures position and also an Ether spot position on one exchange, you could get liquidated on your Ether futures position even if you had enough Ether spot as collateral because you needed to have that spot crypto within the Ether futures spot collateral wallet, which was different than the Ether spot wallet.
It was this game of shifting assets around to different wallets to make sure you kept meeting your collateral requirements, which was just an operational nightmare. Sam told and worked with Gary and Nishad to build basically a cross-collateralization system where you have just one wallet with all of your assets, all haircutted appropriately based on volatility and liquidity, but then summing up to a single collateral value that represents what you can put on in terms of margin for all your positions. Or having an auto-liquidation system that doesn't just, the second that you're slightly below your margin fraction, send a giant market order into the book and dislocate the order book by 10%.
It would automatically start liquidating small percentages of your portfolio at a time to try to minimize market impact. And then if the position got too underwater, it would auction that position off to backstop liquidity providers, a number of them, who would then take on that position again without having to rip through the book and cause dislocation. It was much more orderly, much more predictable. That had to have come from the initial intuitions that Sam and his colleagues got from being traders on these exchanges and thinking, "How should this work if it were perfect?" So I do think in the beginning, they were really working on that product together.
And then once the success came and Sam got drunk on the celebrity of being so out there and known and having all these newfound connections, things started to go by the wayside.
Brett Harrison
You mentioned that one of the things he was doing was making these sort of exorbitant deals with celebrities, acquisitions, branding. What was your understanding at the time of where the money to do this was coming from?
Dwarkesh Patel
Yeah. So for example, when I joined the company, FTX had just inked that Miami Heat deal. I think it was something like $19 million a year. And I was like, "Well, that sounds like a lot of money, right?" But at the time, you could see the publicly reported volume on FTX. It was something around 15 to 20 billion in notional per day. The fee schedule was also public. So even at the highest volume tiers, the take fee would be something like two basis points per trade. So if you just did $20 billion traded per day times two basis points times 365, because crypto trades every single day, you can get a sense of how much money FTX was making a year. At the time, I think the run rate for FTX was something like close to a billion dollars in income. And you think, "Okay, is $19 million a reasonable percentage of the total income to spend on a very significant, important marketing play?" I don't know. It feels kind of reasonable. How much does Coca-Cola spend per year on marketing as a percentage of their income? It's probably somewhere between 50 and 150%. I don't actually know what it is. It doesn't seem crazy.
Brett Harrison
Yeah. But if you add on top of that the real estate, the other acquisitions...
Dwarkesh Patel
First of all, that stuff came later. And secondly, a lot of that wasn't known to the employees within the company. Most of the venture deals, the value of the real estate, et cetera, was non-public within the company. There were $100-plus million investments into various companies and other investment funds that were never discussed openly, at least to the U.S. people. It wasn't like there was clear internal accounting where people could look at it and say, "Hey, are we really spending all this money on all this stuff?" No, I think Sam very deliberately kept all that stuff within his innermost circle for a reason, because he didn't want the criticism on what he was spending on.
Brett Harrison
And did you have access to or did you ask to see a balance sheet or any of the financial documents?
Dwarkesh Patel
I had zero access to bank account stuff or financials on the FTX.com side. On the U.S., I had some. But remember now, knowing what we now know about even recent guilty pleas from the shot and seeing the complaints from the SEC, CFTC, they were deliberately falsifying information that went into ultimately the audited financials. So in order to have actually suspected anything, one would have to not only disagree with all of the internal conventional wisdom around how the company was doing, but also have to basically distrust audited financials coming back to the company. Combined with having any concerns about income when it seemed like we were generating income faster than any startup in history. So I think it was very difficult for anyone within the company, especially on the U.S. side, to have a clue what was going on.
Brett Harrison
For sure. Let's talk about Alameda. I guess maybe the best point to start this story is also with Jane Street, where Caroline Ellison, who went on to become the CEO of Alameda, was a trader. Did you happen to cross paths with her at Jane Street?
Dwarkesh Patel
It's hard to remember because it was like the early days, but I'm pretty sure she was also one of my boot camp students.
Brett Harrison
It all starts there.
Dwarkesh Patel
Yeah. But besides those early interactions, I barely interacted with Caroline, not in the same way that I had done with Sam, just based on the trading desk he was on. And when I joined the company, the FTX US people communication-wise were walled off from Alameda. So we didn't really cross paths almost at all.
Brett Harrison
What was your understanding of the relationship between Alameda and FTX?
Dwarkesh Patel
It was, "This is a completely separate company. Sam doesn't really do anything for them anymore because he is 100% focused on FTX. It's separately being run by Caroline and Sam Trabucco. They have the same access to the exchange, data feeds, and API as any other market maker on the exchange." And also, especially towards the time that I left, Alameda wasn't even a significant percentage of the exchange volume anymore. They weren't in the top 20 market makers on FTX.com or something like that.
Brett Harrison
Yeah. You mentioned that you were contributing to the code base and you had access to the code base. People have been speculating about whether Gary or Nishad had hard-coded some sort of special limit for Alameda. Did you see any evidence of that in the code base?
Dwarkesh Patel
I mean, definitely not.
Brett Harrison
You mentioned that you visited the Bahamas offices a few times. There's the Alameda hut, there's a meeting room, there's where Sam and the engineers are, there's the Future Fund, and then there's the Alameda hut. Did the physical proximity between the offices—and of course, the fact that the leaders were living together—was that something you inquired about or were concerned with?
Dwarkesh Patel
I never visited the places where they lived in that Albany section of the Bahamas. So I think I didn't fully grasp the extent to which they were all living in this particular arrangement. But I understood that as long as Sam was going to be the 90% owner or something of Alameda, he would want oversight there. So having them close by made sense. But the actual hut setup was such that they had physical separation from minute to minute. It wasn't like Alameda could overhear stuff happening on the exchange or people in the exchange could overhear stuff that was happening at Alameda. So to some extent, I felt like, "Well, at least they're going through the right motions of setting up physical separate buildings." I mean, this is not uncommon within trading firms and investment banks. If you imagine there needs to be wall separation between buy side and sell side at different institutions, the way they do that is they put them on different floors in the same building, right?
And sure, they can meet each other for lunch in the lobby, but they set up some actual physical separation. This is super par for the course when it comes to financial firms that have these businesses that need to be walled off from each other. That didn't seem like a particularly strange thing to me at all.
Is there anything that in retrospect seems to you like a yellow flag or a red flag, even if at the time it's something that might make sense in the grand scheme of things?
Brett Harrison
Yeah. The most obvious thing only in hindsight was that Sam liked to do bonuses for the employees twice a year—once at the end of June, once at the end of December. Semester bonuses. In the previous semesters, he had paid them early, like in May for the first semester or early December for the second one. But he was extremely late in doing the mid-year 2022 bonuses. So much so that people in the company started to freak out because there was a lot of bad news in the press about other companies doing layoffs or folding. It was two to three months late, and people were expecting to get bonuses to pay rent and do whatever. There was very little communication around this, and people were very concerned.
At the time, people said, "Look, Sam's really busy. He's flying to DC every week. He has all this stuff going on. He just hasn't gotten around to it, but don't worry, it's coming." In hindsight, I felt like there was some clear liquidity issue. That was probably the most obvious thing.
I think everything else is all just things that were red flags about the organization, not red flags about potential liquidity issues or fraud. Things like the complete inability to hire more people, especially on the developer side. Not allowing me to establish separate C-level staff on the U.S. side that would have authority that was really separate from the ones in the Bahamas. How completely tightly controlled the dev team was around access to the code base and the inner workings of the exchange, really wanting to keep that nexus of the developer group in the Bahamas next to Gary and Nishad. Those seem like red flags now, but not at the time.
Dwarkesh Patel
Did you notice anything weird during the Terra Luna collapse? Because in the aftermath, people have said that's probably when Alameda defaulted on some loans and maybe dug itself deeper.
Brett Harrison
Really nothing at all. Maybe that's a function of being here in Chicago and just not seeing a group of people freaking out. But nothing seemed wrong. In fact, we started having conversations about paying out mid-year bonuses a couple weeks later after the Terra announcement, and everything seemed very normal. Sam sent out an announcement to the whole company basically saying, "Okay, we're going to be paying out bonuses soon. People should expect they're going to be a little bit lower because we have very similar revenue to last year, but we've also grown in size, and the market is slowing, and we need to be a little bit more conservative." All the signs pointed to things as normal.
Dwarkesh Patel
You said you had a thread sort of boiling down this experience on Twitter. One of the things you pointed out is that you saw symptoms of a mental health issue or addiction issue at the time. Are you referring to the management mishaps and bad decision making, or was there something more that made you come to this conclusion?
Brett Harrison
I think it was more than that. When I knew Sam when he was 21, 22 years old, he was a happy, healthy-looking kid who was very positive, very talkative, got along super well with his cohort of traders. The people on the desk really liked him. When I got to FTX, over the course of my time there, I saw someone who was very different than that person I remembered. I think he was angrier, seemed more depressed, more anxious. He couldn't get through a conversation without shaking his leg.
Dwarkesh Patel
At Jane Street, he wasn't like that?
Brett Harrison
Not something I remember at all. He would snap easily. He would not respond to messages for long periods of time. People had different theories. People would attribute it to the unbelievable stress of being in the position that he was in, the complete lack of sleep, his diet, lack of exercise. People had plenty of thoughts about what could be causing it all. But something had definitely deteriorated mentally and physically about him compared to who I remembered.
Dwarkesh Patel
If you had to guess the most likely cause of that, what would you say?
Brett Harrison
I don't know. I think that's up for a professional with credentials that I don't have. But I do think it was probably a combination of everything. The lack of sleep, the stress he probably was under not just from his role, but from having kept this secret for so many years around whatever was happening with the holes in the exchange and the lying he was doing to his own employees, to investors, to auditors. Maybe that weighed on him. Maybe it had something to do with his medications. Maybe it was just a plain deterioration in mental state or some kind of personality disorder or a different kind of anxiety disorder. I really don't know. Maybe a mixture of everything.
Dwarkesh Patel
Got it. You said you gave him a sort of ultimatum letter where you said, "Unless you change these things, I'm resigning." What were the things you asked to be changed in that letter?
Brett Harrison
The top three things were: one, to communicate more with me in particular. I could probably count on one hand the number of times I had a one-on-one phone call with Sam, which probably seems insane given the position I was supposed to be in. I basically said, "We got to talk every week. It's impossible for me to get anything done if I don't have the authority but I have the responsibility to be able to push this company forward, and we're not talking at all." So that was number one.
Number two was to establish separate C-level management staff on the US side. If Sam was going to be so busy doing what he was doing, at least he needed to delegate that responsibility to a set of professional managers who could actually take care of the day-to-day operations within the company. It felt like things were starting to unravel in the absence of that.
The third was to grow the tech team and move a lot of the authority and management of that team away from Nishad and Gary so that we could actually spread the knowledge and be able to keep up with a lot of the tasks that we were assigning ourselves and trying to build all these new business lines. That pretty much summarizes it.
Dwarkesh Patel
How regularly were you talking?
Brett Harrison
It wasn't regular. I was on chat groups that he was in, so occasionally he would respond to something I said on that group. But one-on-one conversations, I think there were fewer than ten for my entire tenure.
Dwarkesh Patel
Wow. And that was over a year and a half? So about less than one every two months.
Brett Harrison
Yeah.
Dwarkesh Patel
How did he respond to this letter?
Brett Harrison
It took a little while before we got on the phone, and he went through every point and refuted every one. Starting with communication, he said, "I think phone calls are a waste of time. If I promise people regular phone calls, they will use it to waste my time, and it's not an efficient mode of communication." He said, "I think we have the best developer team in the world, and I think anyone who suggests otherwise is completely wrong. If we add a lot more people to the dev team, if we move them to the U.S. and move them away from the Bahamas, we're going to be worse as an organization." He kind of ignored the point about separate leadership. I think he hated the idea of giving other people titles that would reflect greater responsibility within the company.
That conversation ended with us not knowing what the future was going to be, because I basically said, "Look, I'm going to resign if you don't fix this." He said, "We're not fixing anything."
What happened next was he had deputized another person within the company to come here to Chicago and pull me into a side room and say, "You are probably going to be fired for this letter that you wrote. And not only are you going to be fired, but Sam is going to destroy your professional reputation. Where do you think you're going to be able to work after FTX after all this happened?" He was threatening me. Not only that, he said, "If you are going to have any hope of staying, and if you can forget about getting paid bonuses, you need to write Sam an apology letter and show it to me first because I'm going to edit it. I'm going to tell you what to say."
I said, "Absolutely not. This isn't a mafia organization. This is extremely unprofessional." I knew at that point there was absolutely no way I was staying. It was a matter of when, not if. But what I did know was that I'm still a professional. I'm still loyal to the company. I still believed the company itself had incredible potential to continue its road of profitability. And I really liked all my employees here on the U.S. side, and I wasn't going to abandon them. So I thought a three to six month time period is about standard to take the time to unwind responsibilities, to finish the stocks platform I was working on, to get my team in a position where I knew they would be in good standing and they wouldn't be retaliated against after I left. I took the time to do that before officially resigning at the end of the summer and early fall.
Dwarkesh Patel
Did that happen? After you left, did he try to injure your professional reputation?
Brett Harrison
The most acute thing that happened was that I actually offered to stay on longer. I said, "I could stay on for a couple more months and help this transition to whomever you name as the successor president of FTX US." He said, "No, I want you gone more quickly." He was communicating through other people at that point. He said, "I want you gone on September 27th." So that was fine with me.
On September 27th, not only did he announce to the company my resignation, he also announced that he was closing the Chicago and San Francisco offices and that everyone had to move to Miami. Basically, if they didn't move to Miami by a certain date, they were not going to be at the company anymore. The employees were distraught.
What I learned later from several investors and reporters who talked to me was that when they talked to Sam about my leaving, Sam told them that my leaving was a combination of resignation and firing. One of the reasons I had to leave was because I refused to move my family to Miami. So basically, "That guy was constructively fired. He closed down this office that I built, and if I wasn't going to move, I had no role left at the company." That took a little bit to crawl out from. I had to tell people, "It's completely false. It didn't happen at all." He was telling people that he fired me.
Dwarkesh Patel
And when he said that, he was still at a sort of peak of hype, so to speak.
Brett Harrison
Absolutely.
Dwarkesh Patel
Did the idea of forming Architect already come to you by this point?
Brett Harrison
Knowing that I was going to leave, I started thinking about what I was going to do next. I thought, "If I think I can run a company better than Sam, I should put my money where my mouth is and start a company." I had a couple of ideas, and I had this particular idea for Architect. It was starting to really form towards the end of my time at FTX, but I hadn't started anything. I left FTX, took a little bit of time off, and then started to talk to investors about raising some money for starting this company.
There were a few investors that basically said, "Do you have Sam's blessing to do this?" I said, "Why do I need Sam's blessing? I've resigned. I don't work there anymore." They said, "We really would feel more comfortable if we could talk to Sam first, make sure things are okay, figure out what he's doing, find out if he wants to invest before we talk to you further." It was impossible to escape that Sam hype bubble, even having left the company.
Dwarkesh Patel
Why do you think they were so concerned? Were they trying to invest in FTX in the future?
Brett Harrison
They were existing FTX investors. I think it really mattered to them what Sam thought of them. If they didn't know the full story and they were being told that Sam fired me, I think they were concerned about potential conflict investing in me too.
Dwarkesh Patel
Was any part of that because Sam had a reputation that if an investor invested in somebody he disapproved of, he would get upset?
Brett Harrison
No, if that happened, I don't know about it. I think it was just that Sam had such a magical hold over the entire industry—from investors to media to politicians—that they looked to him for approval.
Dwarkesh Patel
So at this point you've left FTX US and you're starting to work on your own company. When is this exactly?
Brett Harrison
My official resignation was the end of September. I had stopped working earlier than that. I started working on fundraising for the company in October. Then a month later, the thing explodes.
Dwarkesh Patel
When did you hear the first inkling that there might be some potential trouble?
Brett Harrison
The first thing I heard was the night before the announcement that Binance might be buying FTX. I was just looking at Twitter and saw all of this fear mongering. It was like, "Okay, CZ says he's selling FTT, so FTT is going to go down." People were saying, "Well, that means Alameda is toast. Once Alameda goes under, that's going to be a problem for FTX. Pull your funds from FTX." I was just sort of laughing at this because I'm used to people saying things on Twitter that seem nonsensical. First of all, Sam and Caroline are great traders. If anything, they'll profit from all this volatility. There's no way anything's going to happen to Alameda. Also, the connection between the price of the FTT token and the ability for customers to withdraw their funds from the exchange just did not compute for me at all. I thought, "This will blow over in a couple days like everything else."
The next morning, I was actually busy talking to my own lawyers and investors for the company because we were closing up our investment round. The closing docs for my investment round went out that morning—the morning that FTX announced they were going to be bought by Binance. It was the worst timing in crypto fundraising history. I was busy all morning, then I went online and checked Twitter and saw Sam's tweet that said, "What comes around goes around, and we're going to get acquired by Binance." I felt dizzy. I had no idea what was going on in the world at FTX. I couldn't put the pieces together in my head. It just didn't make any sense to me.
Dwarkesh Patel
So before then, you did not think this was possible. You kept a bunch of money on the exchange. You were still an equity holder in FTX and FTX US. You were still very pro-FTX in spite of your experience with Sam. How did that week unfold for you? You were almost about to close your round. What happened to the round? How were you processing the information?
Brett Harrison
First, the investors and I all had to hit pause. Architect became priority number 1,000 on everyone's list. A number of those investors were trying to do damage control themselves. Either they were themselves investors in FTX or FTT, they had companies whose runway was held on FTX, or they were expecting to get investment from FTX. People were just trying to assess what was happening with their own companies. They were not writing checks into new companies anymore. I had to hit pause on the whole thing for their sake and for my sake.
What could one do in that situation except just read the news all week? Everything that came out was something brand new and unbelievable, more unfathomable than the thing before. It was a mixture of rumors on Twitter and articles coming out in major media publications and the announcement of the bankruptcy. It was information overload. It was very difficult to parse fact from fiction. It was an emotional time.
Dwarkesh Patel
I understand, man. So we've done the whole story of you joining to the company collapsing. I want to do an overview of what exactly was happening that caused the company to collapse and the lessons there. In the aftermath, SBF has been saying that FTX US is fully solvent. They wanted to start processing withdrawals. He had a Substack recently in January where he said it had $400 million more in assets than liabilities. What is your understanding of the relationship between the solvency of FTX US?
Brett Harrison
The short answer is I really don't know. If you had asked me about the solvency of FTX US at the time that I left, I would have said, "Why are you asking about this? Of course everything's fine." Now it's very difficult to understand what is going on. The level of deception that was created by this inner circle of Sam's—reported through the various complaints and indictments from the DOJ—they were doing things to intentionally manipulate internal records in order to fool employees, auditors, and investors. So everything's out the window at that point. Secondly, it sounded like in the week prior to the bankruptcy, there was a flurry of intercompany transfers. Given all that's happened, it's impossible to say what state we are in now compared to where we were several months prior.
Dwarkesh Patel
Who took over management of FTX US when you left?
Brett Harrison
I'm not sure. I've totally cut off from everything FTX at the time that I left.
Dwarkesh Patel
Before you left, were the assets of FTX US custodied separately from FTX International's assets?
Brett Harrison
Yes, they were. We had a separate set of bank accounts, a separate set of crypto wallets. The exchange itself was a separate database of customers. It ran on a different AWS cloud than the international exchange ran on.
Dwarkesh Patel
And you had full access to this and it checked out—more assets than liabilities?
Brett Harrison
Right. But remember, the thing that makes them not separate is the fact—and this was completely public—that Sam was the CEO of FTX and FTX US, Gary was the CTO of FTX and FTX US, and Nishad was the Director of Engineering for FTX and FTX US. As long as there wasn't this completely separate walled-off governance that we were trying to establish while I was there, there was never going to be perfect separation between the companies. This was a known problem. That's what makes it so difficult to understand the nature of what was potentially happening behind the scenes.
Dwarkesh Patel
We've been talking about the management and organizational issues of FTX. Were these themselves not a red flag that something really weird could be happening here, even if it wasn't fraud? These people are responsible for tens of billions of dollars worth of assets and they don't seem that competent. They don't seem to know what they're doing. Was that itself not something that concerned you?
Brett Harrison
It concerned me, and I tried to raise concerns multiple times. If you raise concerns multiple times and they don't listen, what can you do other than leave? But you have to understand that every company I've ever worked at—and I would think any company anyone's ever worked at—has management problems and growing problems. Especially for a super high growth startup like FTX, it's a very natural progression to have the visionary CEO who brings that product to product-market fit, who enjoys that explosive success, and then the reins of the company are eventually handed over to professional managers who take it into its maturation phase.
I thought, "I'm not that person because Sam and I have interpersonal issues. But there are 100 plus major investors in FTX. Someone will figure out how to install the correct management of this company over time and bring it to a good place. One way or another, this is going to succeed. There are too many people with vested interests in doing so." No, I wasn't concerned that FTX wouldn't somehow figure this out. I still thought FTX had an extremely bright future.
Dwarkesh Patel
There might be these sorts of visionaries. A lot of them might have problems, to put it in that kind of language. But I don't know how many of them would make you suspect that there are mental health issues or addiction issues. For somebody who's in charge of a multi-billion dollar empire, it seems like something that's especially concerning.
Brett Harrison
I can't quite speak to whether people would think there are mental health issues of other figureheads of large companies. But remember, at this point, Sam is not leading the day-to-day operations of the company. Many other people are. As the public figurehead of the company, Sam was obviously doing a very good job. He was extremely successful at raising money. He was extremely successful at building a positive image for the company. In that sense, that was all going fine. The rest of the company was being run by other people. I didn't witness the addiction stuff firsthand. I definitely thought he was not as happy a person as I met a long time ago. But could you blame a person for inheriting a $20 or $30 billion company and not taking it super well when you're 29 years old? I think so.
Dwarkesh Patel
You mentioned that the fact that hundreds of accredited investors presumably had done good due diligence gave you some comfort about the ultimate soundness of the company. But potentially, those hundreds of investors were relying on the experienced high-level executives that SBF had brought on. They're thinking, "If somebody from Citadel and Jane Street is working at FTX, that's a good indication they're doing a good job." In some implicit way, you're lending your credibility to FTX. Was there this circle of trust where the investors are assuming if this person who has tons of leadership experience in traditional finance is coming to FTX, they must have done the due diligence themselves? And then you are assuming the investors have done this. So it's like nobody's role to be the guy who's like, "This was my job."
Brett Harrison
I was the person in charge. Remember, regardless of how experienced or inexperienced people in the company are, regardless of how many or few investors there are, how many senior lateral hires there are—if a very small group of individuals who are very smart and very capable intentionally put forth schemes that deceive people within the company and outside the company about the veracity of records, what can you do? What is one supposed to do in that situation?
If the public reporting matches private reporting, if investors have done their own diligence, if we've joined the company and we see nothing wrong within the company from a financial perspective, if we can see the public volume on the exchange and it all matches up with our internal reporting and we know how much fees we're able to collect and all that, and it seems like a lot of income compared to our expenses for a two or three hundred person company—at what point do you go against all of that and say, "In spite of the overwhelming evidence to the contrary, I think something is wrong"?
Dwarkesh Patel
Someone might look at this and say, "Brett, you weren't a junior trader right out of MIT who just joined FTX. You have more than 10 years of experience in finance. You saw Lehman happen. You've managed really large teams in TradFi. You have the skills and the experience. And not only that, your position in FTX as President of FTX US—if somebody with your skills and experience can't see it coming, and maybe you couldn't, whose job was it to see it coming?" It doesn't seem that anybody other than the inner circle could have been in a better position. Is there somebody outside of the inner circle you think should have been able to see it coming?
Brett Harrison
It's a good question. When a major fraud happens in such a way where it was expertly crafted to be hidden from the people who could have done something about it, what should one do? One answer is never trust anyone. Every company I ever work for in the future, every time we say we've done some transaction, I will ask them to show me the bank records and give me the number of the bank teller I can call to independently verify every single banking transaction. This is impractical and ridiculous. It doesn't happen.
The counterfactual here is one where, first, I have to believe that there is some kind of fraud, which I don't. Then I have to say, "I would like to start auditing all bank transactions for the company that I don't work for. Also, I want to disbelieve audited financials from respective third party auditors. I also want to look into the possibility that Sam is lying under oath in congressional hearings about segregation of customer funds. Also, I should disbelieve all of the trusted media outlets and also a hundred financial institutions that have invested in FTX." The chain you have to go through in order to get to a point where you start to be able to figure out something is wrong is really impossible.
The bottom line is there should be mandated, at certain stages of company growth, independent boards. A lot of that has to do with where the nexus of control of the company really is and making sure it's in a place where there is appropriate regulatory oversight and appropriate financial oversight. That maybe could have helped. But besides that, I think this is ultimately a job for enforcement. People will commit crimes, and there's nothing one can do to stop all people from committing all possible future crimes. What one can do is come up with the right structures and incentives so that we can build a trust-based system where people can innovate and build great companies, and people who are bad actors can get flushed out, which is ultimately what I think is happening.
Dwarkesh Patel
But they're not letting you hire people. They're overseeing writing the actual code for FTX US from the Bahamas. That doesn't make you think, "Why are they doing this?" It's a little odd.
Brett Harrison
I just thought it was not the right way to run the company. There's a very large chasm between "I don't think they're doing a good job running the company" and "I think that customer funds are at risk."
Dwarkesh Patel
Fair enough. What should someone who sees bad organizational practices—no board, a ton of really weird investments and acquisitions, and most importantly, they are responsible for managing tens of billions of dollars worth of other people's assets—what should somebody do when they're seeing all this happening?
I mean, obviously it's very admirable that you put this in writing to him, you gave it to him and then you resigned when he refused to abide by it. But maybe the answer is just that. Is there something else that somebody should do?
Brett Harrison
I would say within any company, and I would expect that the overwhelming majority of companies, if you see bad management, it does not imply fraud. But there are lots of places with bad workplace culture and people making bad management decisions. It should be that if you find yourself in that position, there is someone you can go to talk to. It might be your manager, your manager's manager, or someone in your HR department. There should be a designated person within the company for you as an employee that you know is a safe space to bring complaints about the workplace and about the company strategy.
Then you should see how they handle it. Do they take it seriously? Do they make changes? Do they look into the stuff you're talking about? Do they encourage cooperative, positive discussion? Do they threaten you? Do they retaliate against you? Do they start excluding you from conversations? Do they threaten to withhold pay? If you're in that latter camp, what do you do at that point? It's easy for me to say, and I've been in fortunate positions within companies and have personal flexibility. It might not be so easy for the average person to get up and leave a job. But I do think that at some point you have to start making plans. Because what can you do in the face of a giant organization that you disagree with other than leave?
Dwarkesh Patel
Let's talk about regulators and your relationship with them while you were at FTX. As head of FTX US, I imagine you were heavily involved with talking to them. What was their attitude towards FTX like before it fell?
Brett Harrison
All the regulators were, I think, in the common belief that crypto was a large and viable asset class, and in order for it to grow in a responsible way, it needs to come within the regulatory envelopes that already exist, in whatever way that's appropriate for crypto. And crypto can mean a lot of different things. We have to distinguish between centralized and decentralized finance here. I would say regulators saw FTX as at least one of the companies that was very willing to work directly and collaboratively with regulators, as opposed to trying to skirt around the regulatory system.
Dwarkesh Patel
When I was preparing to interview SBF, I got a chance to learn about your proposal to the CFTC. We were just talking about it earlier—the auto-liquidation and cross-margining system, bringing that not only to crypto in the US, but to derivatives for stocks and other assets. I thought, and I still think, it's a good idea. But do you think there's any potential for that now, given that the company most associated with it has been blown up? What is the future of that innovation to the financial system?
Brett Harrison
Yeah, I definitely think it's been set back. It's interesting. Walt Lucan from the Futures Industry Association, in a conference shortly after the collapse of FTX, talked about FTX in a speech, but specifically made a call to the fact that in spite of what happened, the idea of building a future system that can evolve with a 24/7 world is still a worthwhile endeavor, something we should consider and pursue and be ready for. We are 3D printing organs and coming up with specially designed mRNA vaccines, but you still can't get a margin call on a Saturday for an S&P 500 future. There's a real lack of evolution in market structure in a number of areas of traditional finance. I think it's still a worthwhile endeavor to pursue it. I think the LedgerX proposal makes a lot of sense. I understand where some of the concerns were around how that could dramatically alter the landscape for derivatives regulatory structure and market structure. There were still unaddressed questions there. But I still think it was the right idea.
Dwarkesh Patel
During those hearings, the establishment—CME and others—brought up criticisms like, "We have these sort of bespoke relationships with our clients. If you just have this algorithm take its place, you can have these liquidation cascades where illiquid assets start getting sold, that drives their price even lower, which causes more liquidations from this algorithm, and you have this sort of cascade where the bottom falls out." Even though this might not be an accurate way to describe what happened with FTT and FTX, because there was obviously more going on, do you think they maybe had a point, given how FTX has played out?
Brett Harrison
A lot of FCMs have auto-liquidation. There's one particular one where they actually automatically close you every day at 4 PM, and they do it in a really bad way. So the idea of auto-liquidation is not new. The idea of direct-to-customer clearing is not new. The idea of cross-collateralization is not new. The thing I think was novel about FTX was putting it all together: direct-to-customer, margining, cross-collateralization, auto-liquidation. For the regulators to get comfortable with the application, they had to understand that FTX as one entity was performing the roles that typically multiple different entities perform. You always need to ask yourself the question: was there something worthwhile about having those different entities be separate? Or is it just legacy regulatory structure? I think that remains to be seen. We don't have enough experience, especially in the US, with that kind of model to be able to say whether it actually works better or worse. Either way, it was worth a try.
I think the biggest misconception about the application was that if we got approved, it meant suddenly FTX would list everything from corn to soybeans to oil to S&P 500 overnight and completely destroy the existing derivatives landscape. What would have actually happened was FTX would have gotten permission to list one contract on small size, there would have been experience with the platform, and it would have been assessed compared to the alternatives on traditional CCPs. If it was worse, changes would have been made. If it was better, it would evolve, and the market would basically decide what people wanted to trade on. I do think the auto-liquidation part was the main piece people were hung up on: how do you provably show the worst-case scenario in an auto-liquidation cascade? Then again, on large CCPs now in the US and Europe, there are margin breaches all the time. The current system is far from perfect.
Dwarkesh Patel
Backing up to regulation more generally. Many people saw crypto as a sort of regulatory arbitrage. Because regulations are so broken in finance—evidence would be that you're not allowed to do this manually, right? You had to go through this lengthy approval process if you're a giant company to begin with. The entire point of crypto was to get around the regulators and not go through them to get approval for things and hand over that kind of full process to them. Do you think that working with the regulators, and then also being part of crypto, kind of defeated the point of crypto?
Brett Harrison
I disagree with the premise. I don't think the point of crypto is regulatory arbitrage. I think that while crypto remains unregulated, it is easier to get something done in crypto than if it were regulated. That's tautological. I also think that most people, especially on the institutional side, who trade crypto, believe that we are in a temporary state that cannot last forever, which is that crypto is largely unregulated or has a weird patchwork of regulatory authority. Maybe it's the 50 state regulators in the US, or some combination of money transfer and CFD or broker-dealer activity in Europe. So I think it's absolutely a worthwhile endeavor, knowing that there's going to be some regulation for at least part of the crypto ecosystem, to work with regulators to make sure it's done well.
Dwarkesh Patel
One very important question about the whole FTX thing is: what did the conventional narrative about how it went down and why it went down get wrong? Given that you were on the inside, what do you know was different than what has been reported?
Brett Harrison
I actually think not much. The reason for that is, typically when something like this goes wrong and it becomes this media frenzy, there's plenty of opportunity for misinformation to spread. But to the credit of the investigators working on this case, they moved so quickly. They had unsealed indictments within what, two months of this going down? Having a lot of the truth spelled out in facts in a public written document quelled a lot of the opportunity for misinformation to proliferate. Whether that's from a Twitter troll or from Sam Bankman-Fried himself trying to spread misinformation about what happened, a lot of it wasn't really given the room to breathe.
Dwarkesh Patel
What did you make of his press tour in the aftermath? Why did he do it? What was your impression of it?
Brett Harrison
I'm not going to speculate what's inside Sam's head. I think Sam had built up his empire partly through his control over media. He did that by being available all the time and being ostensibly open and honest with them all the time. He probably thought, "Why can't the same strategy work now? Maybe I can sway public opinion. If I can sway public opinion, maybe I can sway regulators and law enforcement." It turns out that is definitely not true. I don't really know. It could just be an addiction to the media. He couldn't stand people talking about him and him not being part of the conversation.
Dwarkesh Patel
Yeah. The earlier account you gave of his day-long media cycles kind of lends credence to that mentality. I have a question about the future of the people who worked at FTX. There are many organizations whose alumni have gone on to have incredible careers in technology. Famously, PayPal had a so-called "mafia"—they went on to found YouTube, Elon Musk went to SpaceX and Tesla. So many other companies came out of the people who came out of PayPal. Bernd Hobart has this interesting theory: the reason that happens is when a company exits too fast, you still have these people who are young and ambitious who go on to do powerful things with the network and skills they've accumulated. Do you think there will be an FTX mafia?
Brett Harrison
The number of the most talented people within FTX are leaving FTX in a slightly different position than the people exiting PayPal in an acquisition. I would say they're in positions more like actual mafia people. So I'm not sure it's going to be some giant FTX mafia. But I do think there are a ton of talented people at FTX who are going to look to do something with their careers. A lot of those people came from very impressive backgrounds prior to FTX. I expect them to want to continue to get back on track and build something great. I do think you're going to see at least a couple of people who emerge from this and do something really great.
Dwarkesh Patel
That's a good note to close the FTX chapter of the interview on.
Brett Harrison
Sure.
Dwarkesh Patel
Let's talk about Architect, your new company. Do you want to introduce people to what Architect is doing and what problem it's solving?
Brett Harrison
Sure. The goal of Architect is to provide a single unified infrastructure that makes it really easy for individuals and institutions to access all corners of the digital asset ecosystem. That's everywhere from individual crypto spot exchanges that are centralized, to DeFi protocols, to qualified custodians and self-custody, and everything in between.
Dwarkesh Patel
I'm not sure I have enough context to understand all that. Could you bring it down a few grade levels?
Brett Harrison
Sure. Let's back up to a very high level. Let's say you are someone who wants to trade crypto in some way. What do you actually have to do? Imagine you want to do something slightly more than just sign up for Coinbase and click the buttons there. Let's say you would like to find the best price across multiple exchanges. Let's say you not only want to find the best price across multiple exchanges, you also want to occasionally do borrow and lending from DeFi. Maybe not only that, you also want to store your assets in off-exchange custody as much as possible. Aside from doing all that manually by opening up all the different tabs in your browser and clicking all the buttons to move assets around and connect all these different exchanges, you actually want to build a system that unifies all these things.
You have this build-versus-buy choice. The build choice looks like: hire five to ten software developers and get them to write code that understands all the different protocols and different exchanges, all the idiosyncrasies of them, that downloads market data, stores the market data, connects to these different custodians, bridges all these things together, and provides some kind of user interface that pulls it all together. It's a significant amount of work that, up till now, basically all of these different companies are just reproducing. They're all solving the same problem. As a trader, you want to focus your time on strategy development and alpha and monetization, not on how to connect to a random exchange. The goal of Architect is to build this sort of commodity software that people can deploy out of the box that gives them access to all of these different venues at the same time.
Dwarkesh Patel
So it sounds like this is a solution for people with significant levels of assets and investment in crypto. I'm assuming it's not for...
Brett Harrison
I think that's the place we want to start. But one phenomenon in crypto that I think is somewhat new and exciting is this: if you want to get into sophisticated equities trading, what do you have to do? You usually have to either establish a broker dealer or get hooked up to an existing broker. You need to get market data, which can be very expensive. The full depth of book feed from NASDAQ costs tens of thousands of dollars per month. If you want to compete on latency, you have to get a co-located server in the NASDAQ colo, which is also going to cost you tens of thousands of dollars per month. There's a significant time and money overhead to doing all this, which is why it's so hard to compete in that space against all the incumbent players.
In crypto, many of the markets are just in the cloud—in Amazon's cloud or Alibaba's cloud. You can very cheaply and easily spin up a server in there for a couple of dollars a month and have the same access as a big HFT. All the market data is free. The order entry is free. The protocols are usually fairly simple. You can use JSON over a web socket, as opposed to speaking FIX over some private line. As a result, there is this large and growing class of semi-professional individual traders. These are smart individuals who have some wealth amassed, and they want to be able to do professional trading, whether that's manual quick trading or simple algos using Python or whatever. They can do that and experiment easily because of the open access of crypto markets. There's a much wider customer base for something like this, which includes these high-powered individuals in addition to small, medium, large hedge funds, prop shops, and different institutions.
Dwarkesh Patel
Is the goal that crypto is the ultimate market you're targeting, or are there other asset classes you also want to provide the service to?
Brett Harrison
We're building very general infrastructure. We think crypto is a viable asset class, but it's one of many. Our goal is to provide institutional-grade connectivity and software to anyone who wants to participate in trading in a semi-sophisticated way. Over time, we'll want to grow our offering as much as possible, given the fact that NASDAQ or whatever already have these barriers.
Dwarkesh Patel
Is it possible for someone to remove those barriers with a solution like yours? An analogy that comes to mind is nobody before Mark Cuban's pharmaceutical company tried to go outside the insurance system and directly sell drugs to people. Is it possible to do something like that for NASDAQ?
Brett Harrison
You can't connect to NASDAQ without connecting to NASDAQ. You can't not go through a broker dealer. But I think we could eventually try to get the appropriate licensing required to be an intermediary that is focused on being a technology-forward interface towards people being able to do more program trading. If the mission of our company is to provide better access, we can do so within the existing system.
Dwarkesh Patel
This raises a broader question. If you're initially trying to solve for the problems that these exchanges should natively have solutions to, or at least some of the problems are the ones these exchanges just natively have solutions to—why haven't these exchanges built this stuff already? You were part of one such exchange, and maybe it functioned better than the others, but they're highly profitable. They have a bunch of money. Why haven't they invested in making their infrastructure really great?
Brett Harrison
In many cases, their infrastructure is very good. It's more a question of what's the incentive of the exchange. No single exchange is going to have all the market share. There's always going to be this market fragmentation problem. The question is, whose responsibility is it to make software that helps solve that problem? If I'm some centralized exchange, my incentive is not to build software to make it easier for my customers to go to all the other exchanges. It's to make my exchange better. I'm going to put all of my R&D dollars into providing new products, offering new different kinds of services, or maybe investment advisory, or lowering the barrier to entry to connect into my own exchange. But not creating this sort of pan-asset class, pan-exchange interconnectivity software.
Dwarkesh Patel
Got it. Given the fact that you're trying to connect these different exchanges, currently most of the volume in crypto is in centralized exchanges. What is your estimate of the relative trading volume of CEX versus DeFi? Do you think it'll change over time?
Brett Harrison
I do think it'll change over time. My view is I can't predict which way it's going to change. People after FTX asked me, "Why don't you try to start your own exchange? Take all your knowledge from FTX US and maybe even raise money to buy the IP out of bankruptcy and start an exchange." My feeling is I don't want to bet personally on the exact direction of crypto trading. I could see it continuing status quo where your Coinbases and Binances of the world maintain market share. I could see it moving significantly to DeFi where people feel like this is the true spirit of crypto—a non-custodial, fully decentralized trading environment. I could also see it going the complete opposite direction and having the existing highly regulated exchange players like NYSE, NASDAQ, and CBOE start to enter the game on spot trading. Where is the ultimate flow going to end up between these three possibilities? I have no idea. I'm much more excited about providing the connectivity layer to all of them and saying, regardless of where the liquidity ends up, we'll be able to facilitate it.
Dwarkesh Patel
Speaking of FTX, how has your experience with FTX informed the development of Architect?
Brett Harrison
First of all, working at FTX gave me an appreciation for just how behind a lot of the infrastructure is on other exchanges. People really liked trading on FTX, institutions especially, because the API made sense. It really did follow the standard state machine of any kind of financial central limit order book that you would see on a place like NASDAQ or CME. Whereas there are a lot of exchanges that have crazy behavior. You send a cancel for an order. You get acknowledged that your order has been canceled. And then you get a fill. You actually get traded on something you supposedly thought you canceled. Things that you think shouldn't be possible are possible. My time at FTX gave me an appreciation for how to design a good API for institutions to want to be able to trade all the time. The protocols on some of these other exchanges aren't quite as good. It's informed how much the focus of Architect should be on wrapping up the complexity of these different exchanges and providing a really good API for institutions and individuals on our side.
Thing two is obviously what happened with FTX. People are much less likely to trust a centralized institution with their personal information, especially things like the keys that allow you to trade on their exchange account or the keys that give you access to their wallet. We're thinking a lot about how to design Architect such that the user can connect to all these places and hook up their wallets without needing to ever give us any of their private credentials. That's another particular inspiration from everything that's happened at FTX.
Dwarkesh Patel
What is your opinion of crypto in general at this point? How has your perception of its promise changed at all, given the things that have happened?
Brett Harrison
I feel the same way now as I did then. It's a one to three trillion dollar asset class that is traded by every major institution, that is being invested in by every major institution. It's totally viable and it needs good, mature infrastructure to support its growth.
Dwarkesh Patel
But is the motivation also informed by a view that crypto is going to be even bigger, or in some way going to solve really big use cases? Or is it simply that this market exists, you don't know what it's going to be good for, but it needs this solution?
Brett Harrison
I certainly do believe that is a likely future for crypto. But to me, the interest in it starts with knowing that this is a huge asset class that needs better infrastructure for trading it.
Dwarkesh Patel
In the aftermath of FTX and other things, all crypto companies have special scrutiny on them. Fairly or unfairly, if there are FTX alumni, it'll be even more so. How are you convincing potential clients and investors that crypto is safe, that FTX alumni are safe?
Brett Harrison
On the FTX alumni side, I just personally haven't had those issues. In recent months as we've been building out Architect, I hired three—almost five now—employees from formerly at FTX to come work with me.
Dwarkesh Patel
Is that some sort of arb? Basically, the overall hiring market has overcorrected on them?
Brett Harrison
100%. And not just FTX. Right now, it's March 2023 as we're recording this. There is a huge arb in the hiring market. All the layoffs in tech and crypto, all the fear around various financial services means we basically didn't need to work on recruiting. I had the best people who worked for me at FTX US. I had ex-colleagues of mine from former jobs come work with me here. We actually didn't have to do any formal recruiting efforts because of just how much supply there now is on the job market, especially in tech and finance.
Luckily, I've had a long career history prior to FTX. Even at FTX, we built really great stuff. We had very good connections and relationships with our customers and our investors. There would be times where on Twitter I would answer a customer's support question at two o'clock in the morning. I maintained a lot of those relationships even through the collapse. These are the kinds of people who are reaching out offering support, offering to test out stuff, who want to be customers, who are also having problems with existing crypto tools and looking for something better. All that stuff has remained intact. I don't really have a concern there.
Dwarkesh Patel
What is institutional interest in crypto like at this point, given what's happened?
Brett Harrison
I think it is just as great as it was before. Every major investment bank in the US has announced some plan to do something with blockchain technology, still, even post-FTX. The top trading institutions of the world are all continuing to trade it. As of what we're speaking about right now, volumes are down because people are generally fearful. But I expect that to turn around pretty quickly. The institutional interest still remains really high. People are definitely expecting and waiting for proper regulatory oversight, especially in the US. That's still happening. People are waiting for higher grade professional tools that make it safe for them to trade and invest in crypto. That's obviously in the works for various things, Architect and otherwise. But otherwise, the interest is all completely there.
Dwarkesh Patel
A broader question somebody could have about crypto at this point, or maybe not crypto generally but crypto trading, is: why is it good to have more crypto trading? With stocks and bonds and other kinds of traditional assets, as we were talking about earlier, you can tell a story that liquidity helps capital allocation—projects that need funding will get funding, and so on. Why is it good if the price of Bitcoin is efficient? Why is building a project that enables that something that's socially valuable?
Brett Harrison
I think it boils down to: first of all, do you think it's important for people to be able to trade commodities? How important is it for the world that people can trade gold efficiently or trade oil efficiently? I think the answer is, if people have a use for the underlying commodity, then it's important. So what's the use of crypto? Each crypto token might have its own use. I don't think everyone has a good use. I think there are a bunch that do. But if you believe in Bitcoin as a store of value and a great medium of exchange, then it's important that there's a good, fair price for Bitcoin to enable that. If you think the Ether token is important for the gas fees required to run a decentralized computer, and you think the programs running on a decentralized computer are important, then it's important for there to be a good price for Ether that's fair. It really depends on if you believe in the underlying use case at all, in the same way you would for any commodity.
Sometimes there are tokens that have more security-like properties where they are like trading Apple stock. There was an initial offering of that token. If people bought it, it actually directly funded the product behind the token. The efficient trading of that token is a barometer for the health of that particular company. They can sell more tokens to raise more capital in secondary offerings, in which case it looks very much exactly like the stock market.
Dwarkesh Patel
That's a great lead into the next question. Will there ever be a time when things that are equivalent to stocks or other kinds of equities will be traded on chain in some sort of decentralized way?
Brett Harrison
I think it's likely. I think the primary reason is that existing settlement systems in traditional markets seem to be very slow and built on very outdated technology. That's highly non-transparent and very error prone. So equities are a prime example of this. It still requires two business days to settle a stock.
A frequent occurrence when I was at trading firms was that you would get your settlement file that told you what trades were settled and also told you things like if any corporate actions had occurred overnight, like paying a dividend or a share split. And they would frequently be wrong. The dividend would be wrong, or it would be missing, or the share split amount was for the wrong amount, or they missed the day that it happened, or they messed up some trades and didn't get reported properly. There are frequently mistakes.
It feels like there should be some easy, transparent, open, decentralized settlement layer for all things that you can trade. Rather than try to retrofit the existing settlement technology to be faster and better, starting from scratch with something like blockchain could make a lot of sense. That's why you hear about a lot of these investment banks working on settling fixed income products on chain, because fixed income products have even worse settlement cycles than equities.
Dwarkesh Patel
Should the marginal crypto trader stop trading? Maybe this is a good question to ask about the marginal trader on Robin Hood or something, but I have a couple of thoughts about this.
The first is that I don't think crypto markets are as efficient as equity markets. So there's probably more opportunities for short and long term edge as a trader in crypto than there would be in equities. That being said, I think there's still an enormous amount of room in both traditional and crypto markets for even individuals to have and derive information that gives them profitable trading ideas.
I actually think it's the wrong conventional wisdom to think that if you are not Jane Street or Citadel or Hutton River or Tower or Jump Trading, then you have no chance of being able to profit in markets except for luck. I do think there are a lot of people who trade and it's pure speculation. It's not really on me to tell them that they shouldn't speculate. They probably derive some personal enjoyment from speculation besides the opportunity for profit.
But I do think the access to more sophisticated instruments and information has helped what have traditionally been players that have been unable to compete in the market actually be able to do so in a way that's sort of systematically profitable.
Okay, so that is, I think, a good point to end the conversation. We got a chance to talk about a lot of things. Let's let the audience know where they can find out more about Architect and also where they can find your Twitter and other sorts of links.
Brett Harrison
Sure. Yeah. So Architect's website is architect.xyz. We also have architect_underscore_xyz on Twitter, and I'm brettharrison88 on Twitter.
Dwarkesh Patel
Okay, perfect. Awesome. Brett, thank you so much for being on Lunar Society. This was a lot of fun.
Brett Harrison
Thank you so much.
Outro begins
Dwarkesh Patel
Hey, everybody. I hope you enjoyed that episode. Just wanted to let you know that in order to help pay for the bills associated with this podcast, I'm turning on paid subscriptions on my Substack at dwarkeshpatel.com. No important content on this podcast will ever be paywalled. So please don't donate if you have to think twice before buying a cup of coffee. But if you have the means and you've enjoyed this podcast or gotten some kind of value out of it, I would really appreciate your support.
As always, the most helpful thing you can do is to share the podcast. Send it to people you think might enjoy it, put it on Twitter, in your group chats, et cetera. Just spreads the world. Appreciate you listening. I'll see you next time. Cheers.