Systems Thinking

Adam Curtis's three-part BBC documentary is the single most important video companion for this course -- and this first episode is where it all begins. Curtis traces how Ayn Rand's philosophy of rational self-interest merged with cybernetic theory in Silicon Valley to produce a vision of self-organizing networks that could replace political authority. The documentary shows how systems metaphors -- feedback loops, self-stabilizing networks, emergent order -- were adopted by California technologists who believed computer networks would create stability without central control. A 1991 experiment by Lauren Carpenter, where hundreds of people coordinated a game of Pong without any leader, seemed to prove the dream was real. But Curtis's argument is devastating: when Alan Greenspan, a member of Rand's inner circle, applied this systems thinking to financial markets by deregulating them, the result was not self-correcting equilibrium but the Asian financial crisis. This is meta-systems-thinking: a documentary about how the very frameworks this course teaches were used -- and misused -- to reshape the world. Students will never look at the word 'feedback' the same way again.

Nicholas Christakis opens this Big Think lecture with a move that perfectly mirrors Unit 1's central insight: he describes shifting from treating patients one at a time to asking why whole populations become sick. This is Simon's ant parable in medical form -- the complexity is not in the individual but in the network. Christakis then introduces the concept of emergence through suicide notes, showing how even the most seemingly private act -- taking one's own life -- is shaped by social structure. His key claim, that 'collective phenomena are not mere aggregations of individual phenomena,' is the foundational principle of systems thinking stated in sociological language. He demonstrates how interventions that target network structure rather than individual behavior produce fundamentally different outcomes, connecting directly to the course's treatment of leverage points in Unit 15. For students who think systems thinking is abstract, this lecture grounds it in the most concrete human realities: illness, death, and the invisible connections that shape both.

This PBS Nova documentary stages a direct confrontation between the rational-agent model of economics and the bounded rationality framework that Herbert Simon pioneered -- the exact intellectual battle that drives Units 2 through 4. The film opens with an experiment that students will never forget: a $20 bill auctioned for $28, because the second-highest bidder also has to pay. This is satisficing failure made visceral. The documentary then traces the history from Adam Smith's rational self-interest through the University of Chicago's mathematical formalization to the 2008 crash, showing how models built on the assumption of perfect rationality failed catastrophically when confronted with actual human behavior. Gary Becker defends the rational model; Richard Thaler demolishes it with billiard-ball experiments. The crash footage is devastating and makes the stakes of this intellectual debate concrete: the question of whether humans optimize or satisfice is not academic -- it determines whether financial systems are designed to be stable or fragile. Watch this after Unit 2 to see bounded rationality's real-world consequences unfold on trading floors.

This FRONTLINE documentary is the best available visual case study of nearly every concept in Units 8 through 13. It traces how artificial intelligence emerged from the same problem-space-and-search framework that Simon and Newell pioneered with Logic Theorist in 1956, then shows what happens when those ideas are scaled to billions of parameters and deployed across entire economies. The documentary profiles DeepMind's AlphaGo, which defeated the world's best Go player not through brute-force search but through pattern recognition -- the exact distinction Unit 12 draws between novice calculation and expert perception. It covers neural networks as hierarchically organized systems of near-decomposable modules (Unit 9), the attention economy and algorithmic design of information environments (Unit 13), and the systemic risks that emerge when complex adaptive systems interact in ways their designers did not anticipate. The two-hour runtime is justified: this is the documentary that shows students how Simon's mid-century insights about bounded rationality and problem spaces became the most powerful technology of the twenty-first century.

Roger McNamee, a veteran Silicon Valley venture capitalist, delivers a thought experiment at the Ostrom Workshop that reads like a systems-thinking diagnostic of the technology industry's transformation. He traces how the removal of engineering constraints around 2003 -- when Moore's Law finally delivered unlimited processing power, memory, and bandwidth -- produced a phase transition in Silicon Valley's culture, shifting from building tools that empowered users to building platforms that extracted from them. His key insight is structural: every major technology sub-sector exists because of an antitrust intervention (IBM's consent decree created Silicon Valley, the AT&T case created the internet, the Microsoft case created web platforms), which means the system we treat as a free market is actually a product of deliberate design. This connects Unit 10's natural-versus-design-science distinction to Unit 15's leverage points: McNamee is arguing that antitrust regulation operates at Meadows' highest leverage level -- changing the rules of the system. The Ostrom Workshop setting is fitting: Elinor Ostrom's work on governing the commons is systems thinking applied to institutional design.

Pinker opens with an argument that directly engages the course's intellectual framework: progress is not a force in the universe but the result of rationality deployed against entropy, parasites, and human nature. This is design science (Unit 10) stated as civilizational principle -- the difference between understanding the world as it is and creating the world as it might be. Pinker's catalogue of progress metrics (longevity, poverty reduction, literacy, declining violence) provides empirical data for evaluating system-level interventions, while his acknowledgment that progress 'hasn't happened in a straight line' maps onto the oscillation and delay patterns of Unit 14. His treatment of cognitive biases and the 'availability heuristic' connects to Units 3 and 12, and his framing of rationality as a collective achievement enabled by language and shared knowledge systems is deeply compatible with Simon's vision of human cognition as embedded in institutional and informational environments. Students who have just finished the first few units on bounded rationality will find Pinker's macro-scale argument both a complement and a provocation.

This FRONTLINE documentary is the best available case study of what happens when a single actor attempts to intervene in a complex adaptive system without understanding its near-decomposable structure. Musk's $44 billion purchase of Twitter provides a real-time illustration of nearly every failure mode the course teaches: ignoring the architecture of complexity (Unit 8) by dismantling teams whose functions were invisible from outside, disrupting feedback loops (Unit 14) by eliminating content moderation systems whose stabilizing role only became apparent after removal, and choosing weak leverage points (Unit 15) like rebranding over structural interventions. The documentary traces how a platform that functioned as a global information system was transformed by someone who treated it as a simple mechanism rather than a complex adaptive system. For the synthesis unit (Unit 16), this is invaluable: students can apply every framework they have learned -- bounded rationality, organizational decision-making, attention economics, feedback dynamics -- to a single, richly documented case.

Hank Green does something in this video that systems thinkers will recognize immediately: he maps a problem space. He walks through more than a dozen concerns about artificial intelligence, explicitly ranking each by likelihood and severity, acknowledging uncertainty about his own rankings, and discovering that the concerns interact in ways that make them impossible to evaluate independently. This is procedural rationality (Unit 4) performed in real time -- Green is not optimizing for the 'right' answer about AI risk but following a transparent process of intelligence-gathering, design, and choice. His discussion of algorithmic black boxes as systems whose decisions cannot be audited connects to Unit 9's treatment of modularity and interface: what happens when a system's modules are opaque? His treatment of intellectual property, platform concentration, and AI-induced psychosis demonstrates stocks and flows of power, information, and trust across interconnected systems. At 93 minutes, this is the longest video in the showcase, but it models the kind of systematic, uncertainty-aware analysis the course demands.