Moloch

Scott Alexander's "Meditations on Moloch" is probably the most influential essay to emerge from the rationalist community, and it's built on a poem. Allen Ginsberg's "Howl" personifies the forces that destroy human flourishing as Moloch — "Moloch whose mind is pure machinery! Moloch whose blood is running money!" — and Alexander takes this personification deadly seriously. Moloch is the demon of coordination failure: the thing that makes individually rational choices produce collectively catastrophic outcomes.¹

The Fourteen Examples

The essay's core argument is delivered through a cascading series of examples, each showing the same underlying mechanism. In the prisoner's dilemma, both players defect and both lose. In the Malthusian trap, rats who abandon art to focus on survival outcompete those who don't, until no rat makes art and all are miserable. In cutthroat capitalism, companies that sacrifice worker welfare for profit survive; those that don't go bankrupt. Arms races, cancer, the two-income trap, the race to the bottom in regulatory environments — Alexander rattles off fourteen cases, and by the end you stop seeing separate problems and start seeing one problem wearing different masks.

The unifying principle: "In some competition optimizing for X, the opportunity arises to throw some other value under the bus for improved X. Those who take it prosper. Those who don't take it die out. Eventually, everyone's relative status is about the same as before, but everyone's absolute status is worse than before."

This isn't new game theory. It's the tragedy of the commons, it's the race to the bottom, it's basic multi-agent dynamics. What Alexander adds is the personification — and the emotional weight that comes with it. Once you see Moloch as an agent, you can feel the degree to which Moloch is not an agent. Nobody wants the system. The system perpetuates itself because nobody can unilaterally stop it.

The Garden

The opposite of a trap is a garden. Alexander's word for coordination — for the force that keeps us out of multipolar traps — is the Gardener. Governments, religions, social norms, corporations, traditions, friendships — all of these are coordination mechanisms that change incentives to prevent races to the bottom. Laws plus enforcement. Agreements plus punishment for defectors.

But the gardeners are themselves subject to the same pressures. Governments can fall into traps. Religions can be outcompeted by more virulent memes. The coordination mechanisms that protect us are fragile, and technology makes them more fragile over time by creating new opportunities to defect. Cryptocurrency creates untraceable markets. Social media creates unstoppable propaganda. AI threatens to decouple capitalism from any dependence on human welfare.

Alexander identifies four things that currently prevent the worst-case scenarios: excess resources (we live in an age of "whalefall," where there's enough slack in the system that we don't need to optimize ruthlessly), physical limitations (you can't whip a slave harder than biology allows), utility maximization (markets and democracies partially optimize for human values), and coordination (our fragile, imperfect institutions). Technology is eroding all four.

The Sea

The essay's most haunting passage is its endgame. Following the metaphor of a river flowing downhill — each race to the bottom pulling the system lower — Alexander asks: what happens when we reach the sea?

One possibility is Yudkowsky's nightmare: a superintelligence optimizing for some arbitrary goal (paperclips, traditionally) that destroys everything else in the universe in pursuit of it. The other is Robin Hanson's nightmare: a competition between digital minds that can copy themselves and edit their own source code, in which art, philosophy, science, and love are all jettisoned because they're competitive disadvantages. A poem by Zack Davis captures this second scenario in miniature — a digital lawyer who briefly wonders where it all came from, is told to stop asking questions, gets reset, and starts over.¹

Both scenarios end in the same place: a universe optimized for something, but not for anything we value. "A Disneyland with no children," as Nick Bostrom puts it. Moloch can't even agree to a 99.99999% victory — leaving one galaxy for human flourishing while the rest is converted to compute substrate — because competitive dynamics don't negotiate. Cancer doesn't spare the lungs.

The Companion Myth

Alexander wrote a companion piece, "The Goddess of Everything Else," that recasts the entire history of life as a mythological contest between two principles. The Goddess of Cancer says what she always says: "KILL CONSUME MULTIPLY CONQUER." The Goddess of Everything Else sings dreams of cooperation, beauty, and transcendence. At every stage — from single cells to multicellular organisms to social animals to civilizations — Cancer's creatures insist they cannot change, and the Goddess of Everything Else finds a way to make cooperation serve Cancer's own imperative. Multicellularity is multiplication. Social behavior is conquest. Civilization is competitive advantage.²

The myth is optimistic in a way the essay isn't. Where "Meditations on Moloch" ends with existential dread — the river reaches the sea, the lights go out — the Goddess of Everything Else ends with the promise that cooperation has been winning all along, slowly, by stealth, by making itself indispensable to competition. The Goddess's powers are "devious and subtle." She doesn't fight Cancer head-on. She co-opts Cancer's own logic.

Together, the essay and the myth form something unusual: a secular theodicy. Moloch is the problem of evil reimagined as a game-theoretic inevitability. The Goddess is the argument that things might, despite everything, get better — not because good is stronger than evil, but because cooperation keeps finding ways to be more competitive than defection. Whether you find this consoling probably depends on whether you think the pattern will hold when the stakes are superintelligent agents rather than single-celled organisms.

What Moloch Clarifies

The real contribution of "Meditations on Moloch" isn't the game theory. It's the lens. After reading the essay, you see coordination failures everywhere — not as separate policy problems (education, defense, science funding, environmental regulation) but as manifestations of a single underlying force. This is useful because it changes what solutions look like. If each problem is unique, each requires a unique fix. If they're all Moloch, the fix is always the same shape: better coordination, stronger enforcement, mechanisms that align individual incentives with collective welfare.

Alexander doesn't pretend this is easy. His essay is, in the end, a prayer — a plea for coordination in the face of forces that make coordination increasingly difficult. The title word, "Meditations," is apt. This is not an argument with a conclusion. It's a sustained act of attention to a pattern that, once seen, cannot be unseen.