Boeree, Liv
▶ Watch on YouTubeCore ideas
Boeree opens by explaining that luck's influence in poker shrinks as sample size grows: over ten hands a stronger player might win only slightly more than half the time, but over ten thousand hands they'll win nearly every time, because skill compounds while luck averages out. She grounds this in game theory proper — Nash equilibria as unexploitable strategies where no unilateral deviation helps, illustrated via rock-paper-scissors — and describes how modern top pros now train primarily by running Monte Carlo simulations of billions of self-play hands to derive game-theory-optimal (GTO) betting ratios, memorizing them the way a chess player might memorize opening theory. She's careful to distinguish GTO play (loss-minimizing and unexploitable) from maximally profitable play, noting the latter still requires reading opponents' deviations from optimal strategy and adjusting — the domain where intuition remains essential, since even the best AI hasn't reached perfectly optimal play. She discusses Phil Hellmuth as a fascinating outlier: a player whose plays often look mathematically indefensible yet who has won far more World Series of Poker bracelets than anyone else, and offers (only half-seriously) the idea that his extreme self-belief may function as a kind of reality-bending confidence rather than deep GTO mastery — contrasted with Daniel Negreanu's edge, which she attributes to reading live physical/verbal tells built from immense experience.
On bluffing, she recounts her early career highlight of successfully bluffing Phil Ivey in a high-stakes tournament, describing the physiological "white noise" of intense stress that shuts down deliberate logical thought once fear crosses a threshold, leaving a contest of nerve — and she draws a parallel (with Lex) to other high-stakes forms of deception, from spousal infidelity to, historically, Hitler's deceit of Stalin before Operation Barbarossa. She reflects that despite poker's built-in deception, professional players are, in her experience, unusually honest people outside the game, since the table gives everyone a sanctioned outlet to "scratch the itch" of being deceptive without it bleeding into real relationships. Discussing loss, she says she trains herself to calibrate the emotional intensity of her reactions to actually match the objective stakes involved, and describes her most difficult period in poker (roughly 2016–2018) as one where she'd lost the love of the game but kept chasing self-imposed metrics (like being the "winningest female player") that no longer reflected genuine passion — a "search for meaning" she compares to what other elite competitive figures (she specifically raises Magnus Carlsen) may face after decades of identity built around winning.
A large portion of the conversation is devoted to "Moloch," a concept she has built a video series around (crediting Scott Alexander's essay Meditations on Moloch and researcher Daniel Schmachtenberger): a game-theoretic "generator function" describing multi-agent competitive systems where each participant is individually incentivized toward a strategy that leaves everyone collectively worse off (a multipolar trap), such as Instagram beauty filters, social media's optimization for outrage-driven engagement, or nuclear arms races. She argues anger is disproportionately viral online because it's an "active" emotion that drives sharing (unlike fear or disgust, which tend to make people disengage), and that platforms structurally reward this regardless of founders' intentions, since ad-based business models require maximizing engagement rather than subscription revenue. As a counter-concept, she proposes "Win-Win" (formally "Omnia"): a playful, wise character representing bounded, healthy competition alongside cooperation, as the meme she hopes can compete against Moloch — while acknowledging she doesn't have a concrete mechanism for how such positive-sum memes could be engineered to go viral as reliably as outrage does.
On existential risk, Boeree ranks engineered pandemics as the threat that worries her most (over AI or nuclear war), criticizing a proposed biosecurity project ("Deep Vixen"/DEEP VZN-style efforts) to hunt for the most dangerous natural pathogens and publish their genomes openly, arguing the assumed benefit (faster vaccine development) is marginal given mRNA technology while the downside — effectively handing bioweapon blueprints to bad actors — is catastrophic; she attributes this to naivety within parts of the scientific community rather than malice. On COVID specifically, she argues the most damaging legacy wasn't the virus but the erosion of institutional trust caused by public-health officials' "white lies" (e.g., early mask guidance) and a failure to admit mistakes candidly, which she believes has left society more vulnerable to a future, more lethal pandemic by fostering reflexive distrust. On AI, she describes feeling "bipolar" about the arms-race dynamic between labs (noting some trust in the safety-conscious culture she's observed at DeepMind specifically), and discusses Nick Bostrom's "orthogonality thesis" — that superintelligence and super-wisdom are not guaranteed to correlate — as the crux of AI risk, hoping capability growth proceeds slowly enough to allow course-correction.
Boeree also shares two personal anecdotes she frames with explicit epistemic humility rather than certainty: a premonition-like inner voice telling her she'd win a major tournament in 2010 (which she then did, six days later), and a 2018 episode where a stranger's informal "energy healing" session at Burning Man coincided with the resolution of a painful, medically undiagnosed ear condition (possibly Ménière's disease) that had plagued her for months — she explicitly declines to offer a causal explanation for either, framing them as experiences that pushed her away from her earlier strict atheism toward greater epistemic humility, while still endorsing scientific inquiry as the right tool for investigating such claims further. Discussing extraterrestrial life, she cites researcher Toby Ord's revised, log-scale Bayesian analysis of the Drake equation (correcting for a statistical bias in naive point estimates) suggesting a roughly 50-50 chance humanity is alone in the observable universe. She closes with reflections on the "paperclip maximizer" thought experiment (debating with Lex whether such a system could still generate emergent complexity), advice for young people to seek "win-win" framings and to be wary of self-servingly rationalizing zero-sum choices as positive-sum, and her own tentative answer to the meaning of life: increasing the richness, complexity, and diversity of experience in the universe — likening it to maximizing the "length of code required to describe the universe" — while enjoying the process along the way.
Excerpts
On the psychological experience of a high-stakes bluff against Phil Ivey:
“I think in reality, because I was so scared... there's a certain threshold of nervousness or stress beyond which the logical brain shuts off, and now it just gets into this — it feels like a game of wits, basically, of nerve: can you hold your resolve... I call it the white noise.”
On why she believes anger dominates social-media virality:
“If someone's disgusted or fearful, they actually tend to... be less likely to actually engage and share it, whereas if they're enraged by a thing, well now that triggers all the old tribalism emotions... that's how things get spread much more easily, they out-compete all the other memes in the ecosystem.”
On her tentative answer to the meaning of life:
“I think [it's] to explore, have fun, and understand, and make more of... whatever this is — more of experience, ideally positive experience, and more complex... make it so that the length of code required to describe the universe is as long as possible... highly complex and therefore interesting.”