Podcast notes – Sam Bankman Fried (FTX founder) with Tyler Cowen

Guest: SBF – FTX and Alameda Research founder, crypto billionaire
Host: Tyler Cowen

What is his production function?
World is messy, huge power in getting quantitative / numerical when it’s hard to do so

Worked at Jane Street before crypto
Lot of interview questions are game questions
You’re always making probabilistic decisions, focus on the factors that matter and ignore the rest

Trading is grinding to do a good job
Goal is making money
Constantly think of solutions and try them out

Parents are famous Stanford law professors

Utilitarian perspective – turn things into numbers, and decide which number is the biggest, even in a messy system

Crypto today is enormous fraction of discourse in politics and regulation
Understand the nuances, possible directions
200 countries – each one is trying to figure out how to regulate crypto

Thought he was a good manager until he started Alameda
Seems easy when things are going well
What to do when two employees vehemently disagree?
First, think of the object level truth / answer
Find the right answer
Company culture should focus on that right answer, being ruthless about it
Need agreed upon process to make hard decisions

Some sports teams outperform every year
eg, Belichick + Patriots
They’re known for doing the right thing – what play optimizes our winning – it’s not about conventions or what’s popular
In sports, a lot of fruit is low hanging, focus on that

Supporter of effective altruism
Lot more funding for it now than before
Still need more innovation in how to properly spend money
Also need more good people – to start and lead projects and take responsibility for their success
A leader who can get everyone to take leap together, putting flagpole down

Next $million, where do you give it?
Two things
1. Pandemic preparedness (covid response screwed up, haven’t improved for next one, we got lucky all things considered). We just shut everything down. Vaccine was fast by vaccine standards but not by “people are dying” standards. Lobbying Washington for more spending on pandemic preparedness. Funding vaccine capacity / vaccine stockpiles / early detection systems
2. ?

He’s a pure Bentham-ite utilitarian
Tyler: Why spend money on life extension? Easier to grow utils than save them
SBF: Not compelled by life extension research, not necessary for world flourishing

SBF is vegan
No problem with eating meat if you’re net happy
When he switched to vegetarian / vegan, took 6 months to make real progress
Easier to go cold turkey than steadily wean off it
Doesn’t like cognitive load of deciding whether to eat meat

Everything gets weird at infinity when you think about expected values
Infinity matters, but no one has good account of it

Crypto
Can you coherently regulate stablecoins?
All stablecoins backed by dollars, in regulated accounts – solves most of current problems
But lots of gray area, like what’s difference between PayPal and USDC?
Banks can re-hypothecate, but what about stablecoins?
How do you regulate a coin that doesn’t call itself stable, but in reality has a very stable price?
A stablecoin is stable in both directions (asymmetric for consumer)

DeFi
Tyler: where is implicit (hidden) leverage?
In 2008, lots of bespoke OTC swaps that weren’t properly accounted – these exist in crypto too
Everything’s on-chain, but it’s inconsistently tracked
Dai – algo stablecoin, backed by crypto assets, not perfectly stable – but can also be used itself as collateral for other loans / leverage

Do we have rational theory for crypto prices?
Dogecoin – does away with pretense – can’t do DCF – bizarre and wacky but powerful – it’s just supply and demand (memes and anti memes)
Elon’s greatest product is TSLA (the stock, not the car)
Are there other asset classes aside from crypto that have price models that are similarly artificial / unreliable?
Meme stocks show this

Bitcoin’s price comp with gold is useful
Currently trading ~10% of gold
Swiss franc or bitcoin – which has more impact on world? Now the answer is obvious, but it wasn’t 2-3 years ago

What does it take for global scale blockchain?
take NYSE, Nasdaq, Google, Facebook
if billion users, 10 clicks per day, 1M TPS (transactions per second) – need something of that scale
most blockchains can’t support that today
ETH can support 10 – off by 6 orders of magnitude
ETH2 will support thousands of TPS per shard, but the shards don’t talk to each other

Growing up, McDonalds fries were his favorite
hard to find good fries in Asia
Tyler’s favorite fries are in Patagonia (also Belgium)

likes Beyond Burger – huge fraction of all veggie burgers sold, taste a lot better than most veggie burgers

currently in Bahamas – somewhat bullish on it, great place for some types of businesses
investing in schools, health infrastructure
nimble small state, robustly democratic
people are very nice
close to Miami
how would he give away $10M in Bahamas?
Health and hospitals
cash transfers are always a reliable answer