Simulation and Emergence

The question that ties this section together: how do simple rules produce complex behavior? A grid of cells following one update rule builds a universal computer. Ants following local encounter rates build a colony with 30-year memory. A statistical physics model with three parameters reproduces the entire wealth distribution of the United States. The phenomenon is everywhere, and it's either the deepest insight in complex systems theory or a fancy name for "we don't understand what's happening." Probably both.

The Mechanics of Emergence

Cellular Automata provides the purest examples — Conway's Game of Life, Gosper's HashLife algorithm, and the Neural CA work from Distill that inverts the traditional relationship by training emergent behavior through gradient descent. Artificial Life extends this with physarum simulations, Particle Life (where breaking Newton's third law makes particles behave like living things), and the genuinely unsettling xenobots — real biological cells doing something that looks like artificial life. Morphogenesis shows the same activator-inhibitor mechanism producing patterns from nanometer-wide bismuth crystals to leopard spots to bacterial biofilms — Turing patterns as a universal property of systems with competing feedback at different spatial scales.

Emergence is the hub article, collecting examples from whale vocal clans (cultural memory through biased social learning), ant colony memory (distributed state through encounter rates), biofilm architecture (physics-driven collective form), and the prestige economy (self-interested agents producing coordinated groups). The running thread: when the constraints are similar, distributed systems converge on similar solutions whether the components are ants, neurons, or firms.

Systems Thinking

Leverage Points provides the practical framework — Meadows' hierarchy of where to intervene in a system, from parameters (least powerful, most fought over) through feedback loops and information flows to paradigms (most powerful, hardest to change). Systems That Eat Themselves adds Gall's laws: systems develop their own goals, operate in failure mode, and resist reform through Le Chatelier-like inertia. Ecological Modeling applies both to the question of whether we can build predictive models of living systems — and finds that the gap between simulation and understanding is widening.

Model Hierarchies proposes the solution: not a single model that captures everything, but a ladder of models where each rung is simple enough to understand and complex enough to be informative. The analogy to biology (bacteria → fruit fly → mouse → human) and to ML (Mechanistic Interpretability studying small models before large ones) is deliberate.

Scaling and Growth

Scaling Laws covers power laws in economics and biology. Superexponential Growth goes deeper on Roodman's extraordinary finding that the human economy has been growing superexponentially for 12,000 years — with mathematical projections that imply infinite output around 2047. Chaos And Universality shows how even the simplest nonlinear equation (the logistic map) produces behavior so complex it contains universal constants (Feigenbaum's 4.669...) and connects to the Mandelbrot set. Patterns that hold for an extraordinarily long time before suddenly failing — Borwein integrals, plateau erosion — serve as a cautionary tale about mistaking persistence for stability.

Competition and Cooperation

Intransitive Competition explains why biodiversity is self-reinforcing: rock-paper-scissors dynamics where space is everything. Food Webs And Trophic Cascades maps the network structure of ecosystems — including the hidden half that parasites contribute. Epistemic Logic shows how the blue-eyed islanders puzzle reveals the gap between mutual knowledge and common knowledge — the same gap that Common Knowledge in the Rationality section explores for coordination failures. And Game Theory And Cooperation — now housed in the Rationality section, where its focus on strategic interaction fits more naturally — connects Cantor's diagonal argument to the prisoner's dilemma through the shared structure of self-referential limitation.

The Computation Thread

Information And Computation is the section's philosophical anchor — Chaitin's Omega, Landauer's principle, the arrow of time as the arrow of memory, compression as intelligence. The article argues that information is physical: it has thermodynamic cost, logical limits, and mathematical weight. This connects to Predictive Processing (Bayesian updating has minimum physical cost), to Mechanistic Interpretability (the model you're interpreting knows more about text than you do), and to Spacetime And Information in the Physics section (the fabric of space may be a quantum error-correcting code). SAT Solvers and Hash Function Design are the practical side — what happens when you apply computation to search, and what you discover when you let search discover your algorithms for you.

What's Not Here

The section would benefit from an article on network science proper — the mathematics of small-world networks, preferential attachment, and community structure that underlies food webs, social networks, and neural architectures. Computational Astronomy is a beautiful standalone but somewhat orphaned from the rest of the section.