# Worlds Hidden in Plain Sight **David C. Krakauer et al.** | [[Prediction]]  --- > "From one perspective, dynamical systems can be viewed as simply behaving—obeying the laws of physics. From another perspective, they can be viewed as processing information: how systems get and use information determines how they behave." This is a collection of essays from the Santa Fe Institute—the intellectual home of complexity science. The mission: to search for order in the complexity of evolving worlds. The premise: phenomena can be hidden not just in space (too small or distant) or time (too fast or slow), but through nonlinearity, randomness, collective dynamics, hierarchy, and emergence. These are worlds hidden in plain sight. Not because we can't see them, but because our intuitive and analytical tools are poorly suited to grasping them. --- ## Core Ideas ### [[Complexity as Hidden Worlds]] Traditional science finds things hidden in space (microscopes, telescopes) or time (slow-motion, geological records). But complex phenomena are hidden differently—through properties our minds struggle to grasp: nonlinearity, randomness, collective dynamics, hierarchy, and emergence. The signature concepts of complexity science—adaptation, networks, phase transitions, fitness landscapes, self-organised criticality, edge of chaos—remain more metaphorical than precise. This is intentional. The metaphor borrows language from one field and imposes it on another, and it's around the jagged edges of imperfect fit that much innovative potential lies. ### [[Explore vs Exploit]] in Adaptive Systems Every adaptive system faces the same trade-off: exploration versus exploitation. To explore a new niche requires untried action sequences. This can only happen at the cost of departing from sequences with established payoff rates. The ratio of exploration to exploitation relative to available opportunities shapes the life history of any system—biological, economic, or cognitive. This applies at every level: organisms adapting to environments, companies testing markets, individuals balancing career options. ### [[Emergence Through Thresholds]] > "Because of the stimulus-threshold relationship of behavior, division of labor—the hallmark of social organization—is an inescapable property of group living." Social organisation doesn't require special genes or new neural systems. It emerges from simple stimulus-response threshold relationships. When individuals with different response thresholds interact, division of labour emerges automatically—the individual with the lowest threshold for a task ends up doing more of it, reducing the stimulus for others. The evolutionary transition from solitary to social life doesn't require new features. Solitary insects already have all the behavioural components for organised social living. The organisation emerges from interaction. ### [[Ergodicity and Time Averages]] Boltzmann coined "ergodic" for situations where time averages and ensemble averages are identical. But not every situation is ergodic—and non-ergodic situations are profoundly counterintuitive. The ensemble average includes those few lucky copies of yourself whose enormous gains make up for your likely losses. But you're not playing across parallel universes—you're playing through time. In investment contexts, the difference becomes critical when risks increase, when leverage amplifies fluctuations, when reward structures incentivise excessive risk. Small catastrophes are probably essential for maintaining system health. Suppressing all forest fires today sets up larger catastrophes tomorrow. Financial systems are likely no different. ### [[Cities as Social Reactors]] Cities have unique transformational power. What's happening inside them is massive information exchange across social and economic networks, all on a complex infrastructural landscape. Two effects mirror each other: economies of scale in infrastructure and increasing returns to scale in social interaction. The ultimate function of cities: they're social reactors, places where interactions among many different strangers can be realised and sustained. That accelerating dynamic creates the buzz of a great city. ### [[Hierarchy and Information]] At all levels of complexity in life, we see hierarchical structure where higher regulatory structures sharpen or direct lower-level constructive processes. This information hierarchy appears to be universal in biological systems—possibly the key to life's greatest mystery, the origins of biological complexity. Each new level of organisation typically brings new functionality—a feature with positive payoff for the system as a whole or its components. Hierarchies are constructed on a building block principle: subsystems at each level combine small numbers of subsystems from the next lower level. --- ## Key Insights **Metaphors drive innovation.** The road to novel theoretical work winds through a forest of metaphors. Though often dismissed as rhetorical, metaphors reveal potentially fruitful connections. The fit is always inexact, and it's around those jagged edges that innovative potential lies. **Incentives and morals are not separable.** Economists assume incentives and morals are "additively separable"—the effects of one don't depend on the level of the other. But studies show that rewarding kids for helping behaviour actually decreases helping. External incentives can crowd out intrinsic motivation. **No vertebrate has evolved wings without giving up something.** There are no hexapod vertebrates. Bats and birds had to sacrifice forelimbs to produce wings. Evolution often can't get "there from here"—constraints on what's achievable shape what actually evolves. **Life's robustness depends on variation.** Systems that suppress or lose diversity are prone to collapse. This applies to ecosystems, economies, and organisations. Evolution is about preparing for the unknown, because the scope of possible environmental changes is too immense to predict. **Alliances and patron-client relationships are different.** In true alliances, partners treat each other as equals around a shared goal. Patron-client relationships start with a gift—and there's no free gift. The receiver becomes obligated. Understanding which type of relationship you're in matters enormously. > "To win a war, it helps to have three things: more troops than the enemy, intelligence about the enemy's plans, and superior technology." **A single coordinate system can't cover complex topologies.** You can't cover a sphere with one smooth coordinate system—you need two and rules for connecting them where they overlap. A single universal representation can't capture spaces with non-trivial topology. This is a metaphor for understanding: sometimes you need multiple frameworks and explicit rules for how they connect. --- ## Connects To - [[Making Sense of Chaos]] - complexity economics shares the SFI intellectual lineage - [[Antifragile]] - systems that gain from disorder, small catastrophes as necessary - [[Linked]] - network science as one lens on complex systems - [[The Fifth Discipline]] - systems thinking as a related but different tradition - [[Systemantics]] - how systems resist control through emergent behaviour --- ## Final Thought Complexity science is premised on a bold assumption: seemingly disparate phenomena—natural and social, evolved and constructed—can be understood through a common conceptual framework. The same patterns appear in ant colonies, cities, immune systems, and markets. The signature concepts remain metaphorical rather than precise. But that's a feature, not a bug. Language is thick with the corpses of dead metaphors—"corporation" from corpus (body), "strategy" from strategos (general). Today's loose metaphors become tomorrow's rigorous definitions. What makes SFI's approach distinctive isn't withdrawal to a vantage point with a bigger picture—seeing the big picture necessarily loses information at higher resolution. It's developing new ways of directing attention, seeing and manipulating the tension between foreground and background. The worlds are hidden not because they're invisible but because we haven't learned how to look.