What this covers
Sam Bankman-Fried discusses the constraints on effective altruism and the principles underlying his approach to both cryptocurrency and philanthropic strategy. The conversation centers on a core diagnosis: that capital and ideas are abundant, but the scarcest resource is capable founders and leaders willing to execute the unglamorous work required to build organizations. This framing shapes how he thinks about where Effective Altruism should allocate resources, how companies like FTX and Alameda found room to operate, and what traits separate successful ventures from failed ones. The discussion moves across philosophy, organizational theory, and practical finance, with Bankman-Fried drawing examples from baseball statistics to pandemic preparedness to exchange design.
The conversation covers both normative claims about how altruistically-minded people should structure their careers and technical positions on how organizations should be built and funded. On the altruism side, he argues that the world's Diminishing Marginal Utility differs from an individual's, making Risk Neutrality for Altruists a sound principle, and that Effective Altruism should weight theories of impact that remain robust under uncertainty rather than betting everything on a single model of how the world works. On organizational design, he examines why large teams suffer from Diffusion of Responsibility, why Context as Productivity Multiplier makes concentrated work superior to distributed effort, and how re-granting programs can reach small high-impact grants while reducing institutional bias. In crypto and finance, he discusses the balance between efficiency and risk in exchange architecture, the likelihood that blockchain elements will migrate into traditional finance rather than replace it, and the systemic risks posed by uncentralized clearing structures.
Bankman-Fried argues that the binding constraint on doing good in the world is capable founders willing to do brutal grunt work, and that effective altruism should fund robustly impactful bets while staying flexible across moral and physical uncertainty.
- The world is full of inefficiencies and the missing ingredient for most projects is the right person to lead, not capital or ideas
- Altruistically-minded people should take higher-variance careers because the world's marginal utility of money does not diminish like an individual's
- EA should weight moderately more robust theories of impact because chains of reasoning contingent on current models can fall apart under uncertainty
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
When choosing among life-saving interventions, the correlation between the outcomes of separate beneficiaries is irrelevant to expected value: a person in Mexico and a person in Bangladesh each simply want to live, so whether their chances of being saved are correlated or anti-correlated should not change how much you try to help each.
“the person in Mexico isn't saying, “I only want to be saved in the cases where the person in Bangladesh is or isn't saved.” That’s not relevant. They want to live.”
A core trait of great founders and early teams is willingness to do enormous amounts of dumb, annoying, broken grunt work—like spending hundreds of hours opening bank accounts or stationing someone senior in a bank branch to manually instruct wire transfers—because if the CEO refuses such work it simply won't get done and the company can't operate; the founder must reframe it not as low-prestige but as the important, valuable thing to do.
“One of the things at early Alameda that was really crucial to our ability to make money was having someone very senior spend hours per day in a physical bank branch, manually instructing wire transfers. If we didn't do that, we wouldn't have been able to do the trade.”
Most companies that grow headcount very fast become an absolute mess: diffusion of responsibility means no one takes charge and nothing gets done, so firms hit the negative marginal utility of employees—where adding more people reduces total output—which explains why large incumbent exchanges failed to act on SBF's risk-improvement proposals.
“in practice, most of the time that we see a company grow really fast, really quickly, and get really big in terms of people, it becomes an absolute mess.”
Cross-collateralizing positions in one venue gives capital efficiency, but the downside is that the risk of one position can affect another; traders uncomfortable with a single venue's risk algorithm can spread collateral across venues, so it is a balance rather than a binary, and no single exchange will win because different exchanges make different decisions better suited to different situations (e.g., bespoke risk models for physical commodities like corn the standard model can't understand).
“you get capital efficiency for cross-margining them—for putting them in the same place. But, the downside of that is the risk of one can affect the other. There's a balance there, and I don't think it's a binary thing.”
The earlier view that EA causes are multiplicative rather than additive has become more complicated, especially around AI: factors like economic growth can advance both AI safety and AI capabilities research, so whether they help net safety is ambiguous; where causes are multiplicative you should fund every piece, but where they are not you should identify and advance only the most impactful pieces.
“if what you think matters is something like the relative progress of AI safety research versus AI capabilities research, a lot of things are going to have the same impact on both of those, and thus confusing impact on safety as a whole.”
Even for profit-minded projects, the missing ingredient is most frequently the right person or team to lead, not ideas or capital, because being a founder is brutal and requires a specific, extensive set of skills that are therefore in high demand.
“the missing ingredient quite frequently for them is the right person or team to take the lead on it. In general, starting something is brutal.”
Altruistically-minded people should pursue higher-variance, higher-risk careers because, unlike individuals whose marginal utility of money declines quickly, the world's marginal utility of money does not diminish, making expected value rather than personal security the right criterion.
“To the extent you're just thinking of personal criteria, that's going to argue heavily in favor of a safer career path because you have much more quickly declining marginal utility of money than the world does. So, this kind of path is specifically for altruistically-minded people.”
Because many chains of altruistic reasoning are contingent on our current and possibly mistaken understanding of the universe, EA's tendency to fixate on one specific model is risky; one should weight moderately more toward theories of impact that behave robustly when underlying assumptions are jostled, while still avoiding the hand-wavy 'diffuse good' reasoning that usually masks unrigorous thinking.
“people end up fixating on one particular understanding of the universe, of ethics, of how things are going to progress. But, all of these things have some amount of uncertainty in them. And when you jostle them, some theories of impact behave somewhat robustly and some of them completely fall apart.”
To handle adverse selection from an easy application process, Future Fund vets all large grants and does deep dives into randomly sampled subsets of small grants, which provides a statistical sense of whether they are facing significant adverse selection while still scaling up grant numbers.
“We do deep dives into randomly sampled subsets of them—which allows us to get a good statistical sense of whether we are facing significant adverse selection in them.”
A re-granting program—distributing pots to domain experts who allocate semi-autonomously—is valuable because it gives more stakeholders a voice, lets a small team reach the many small grants ($25,000 to launch an org) that are often the most impactful per dollar but too numerous to evaluate centrally, and reduces the bias toward funding only people the central team already knows.
“sometimes the smaller grants are the most impactful per dollar and there are a lot of cases where someone really impressive has an exciting idea for a new foundation or a new organization that could do a lot of good for the world and needs $25,000 to get started.”
Inverting the traditional structure—putting an automated clearinghouse at the center with FCMs that may use discretion at the edges—contains systemic risk because initial margin must always reach the clearinghouse under clear programmatic rules, so one intermediary's bespoke credit decisions cannot pose risk to other intermediaries, unlike the current system where each FCM is exposed to every other FCM's manual decisions (as seen in the LME nickel blowup).
“The goal of this is to prevent contagion between different intermediaries. Whatever credit decisions one intermediary makes, with respect to their customers, doesn't pose risk to other intermediaries.”
Pandemic prevention requires tens of billions of dollars and international cooperation on pathogen surveillance and equitable vaccine distribution, which makes it too large to fund individually and necessitates government involvement, because in a global pandemic incomplete coverage anywhere lets disease spread everywhere.
“we're talking tens of billions of dollars of scale necessary to start putting this infrastructure in place. So it's a pretty big scale thing—which is hard to fund to that level individually.”
Career advice often references visible outward factors of others' success, but the actual drivers of people's career decisions are heavily personal—family, friends, and people close to them—rather than the public success factors observers cite.
“When we talk with people about their careers, personal considerations and the advice of people close to them weigh very heavily on the decisions they end up making.”
Having multiple effective-altruist donor organizations is valuable because every organization has blind spots that a diverse set of donors can cover, even when their focus areas overlap rather than being identical.
“Every organization has its blind spots, and you can help cover those up if you have a few.”
Unlike pitching, knowledge work doesn't physically injure people, but context is a huge component of effectiveness: holding all the context of the company in your mind lets you operate far more efficiently, so one full-time employee beats two half-time employees and concentrated, uninterrupted work produces better output than constantly broken-up work.
“if you could have either two half-time employees or one full-time employee, you're way better off with one full-time employee because they're going to have more context than either of the part-time employees would have—thus be able to work way more efficiently.”
Losing the Carrick Flynn race produced only limited updates because there is a mathematical limit to how much one can update from a single occurrence: if you believed odds were 50-50, the outcome can revise your estimate by at most a factor of two.
“If you thought the odds were 50-50, and it turns out to be close in one direction or another, there's a maximum of a factor-of-two update that you have on that.”
FTX found that an employee's prior experience has little correlation with how much mentorship they need; far more predictive is how they think, how well they understand new and different situations, and how hard they try to integrate FTX's workings into their understanding.
“we found that the experience of the people that we hire has not had much correlation with how much mentorship they need. Much more important is how they think, how good they are at understanding new and different situations”
Early-stage startups should not be maximizing the chance of at least a little success, which requires founders to accept the personal fallout of failing; therefore communities like EA must normalize and be okay with failure, which most communities currently are not.
“Most startups and especially very early-stage startups should not be trying to maximize the chances of having at least a little bit of success. But that means you have to be okay with the personal fallout of failing and that we have to build a community that is okay with that.”
Baseball evidence shows pitchers worsen over a game and that performance depends on very recent usage, so breaking up outings—three innings every three days rather than six innings every six days—yields better average pitching for the same total innings, a shift the sport has broadly adopted as starter pitch counts have fallen over the last 5-10 years.
“If you could choose between someone throwing six innings every six days, or throwing three innings every three days, you should use the latter. That's going to get the better pitching on average, and just as many innings out of them”
Rather than crypto subsuming traditional finance or a decentralized utopia, much of what's happening in crypto will migrate into traditional finance: blockchain settlement, stablecoins as settlement mechanisms, tokenization of more assets, and composable on-chain programs are all likely, while the intersection of decentralized ecosystems and regulation remains genuinely uncertain.
“a lot of properties of what is happening crypto today will make their way into Trad Fi to some extent. I think blockchain settlement has a lot of value and can clean up a lot of areas of traditional market structure.”
Combining the exchange and broker into one entity, despite apparent conflicts of interest, mostly serves the same goal of a great customer experience, because when two separate entities must hand off to each other the customer gets the least-common-denominator of support, whereas synchronizing them lets the combined entity support the full experience well.
“When you have two different entities with two completely different businesses but have to go from one to the other, you're going to end up getting the least common denominator of the two as a customer.”
People focused on helping the world will be disproportionately represented among both the most and least successful, partly because risk-neutral utilitarian philanthropists lack diminishing returns to consumption that would otherwise push them toward safer paths.
“People who are looking at how they can help the world are going to end up being disproportionately represented amongst the most and maybe least successful.”
FTX and Alameda's consecutive success reflects that the world has many low-hanging inefficiencies generally, not merely a one-time property of crypto markets at a particular moment.
“I think it's more of the former, there are just a lot of inefficiencies.”
Beyond some point, a single donor giving heavily to a candidate produces diminishing or even negative marginal returns due to negative PR.
“At some point, I think that's probably true.”
Having a 'RAM-skewed' mind has helped FTX adapt quickly and identify what's most important at any moment—focusing on product, regulation, licensing, and branding—but its downside is needing people around him to remember long-term important things that get lost day-to-day, requiring periodic stepping back to recover the big picture.
“I need to have a lot of people around me who will try and remember long-term important things that might get lost day-to-day.”
Living in the Bahamas and building specific infrastructure with strong government support has updated SBF positively on the feasibility of charter cities, making them seem more doable than he previously thought, though building a full city from an empty plot of land remains far harder than adding pieces to an existing welcoming country.
“It's made me feel like it is more doable than I previously would have thought. But it's a lot of work.”
The most likely way Future Fund fails to meet expectations is becoming 'lame'—funding decent things that everyone else already funds while its innovative bets on new causes and new organizations don't work, leaving it ineffective at starting new things.
“We end up giving the same that everyone else is giving. We don’t turn out to be effective at starting new things, we don't turn out to be effective at thinking of new causes or executing them.”
If he had to restart earning to give without crypto, the odds of becoming a billionaire are pretty decent, depending on how aggressive he chose to be, because dedicating himself to starting businesses had a decent chance of success even though safe paths would not have ended there.
“if I dedicated myself to starting up some businesses, there would have been a pretty decent chance of it.”