Norbert Wiener, 1950. The popular companion to Cybernetics (1948). Two volumes, one author, deliberately split: the math version and the applied-to-society version. This page walks chapters 1β10, with interactive exhibits for the load-bearing concepts, plus a parallel track on Teilhard de Chardin β the Jesuit paleontologist writing in the same window with a strikingly similar shape: directional cosmic evolution, a planet-scale layer of mind, convergence as the goal of complex systems. Different vocabulary, same questions.
Wiener opens by positioning cybernetics as the science of communication and control in the animal and the machine β the same theory covers nervous systems, factories, conversations, and feedback amplifiers, because they all involve information traveling through channels and acting on receivers.
The historical setup is the move from Newton to Gibbs. Newtonian physics is reversible and deterministic β clockwork. Run the equations backward and you get yesterday; run them forward and you get tomorrow, exactly. Gibbs's statistical mechanics (1870s onward) broke that picture: the universe is fundamentally probabilistic, irreversible, and trending toward higher entropy. Time has a direction.
Once the universe is statistical, information becomes a real physical quantity, not a metaphor. Knowing something means having local order β a configuration improbable enough to carry meaning against the background noise. Entropy is the load-bearing concept for everything that follows in the book.
The universe trends toward heat death (max entropy). But locally, ordered structures can persist by importing energy and information and exporting entropy. Life and civilization are local pockets running against the gradient. Below: particles drift toward disorder. The orange "eddy" maintains a low-entropy region by pulling particles in and pumping heat out. Pause the eddy to watch what happens.
This is the chapter that gets cited everywhere later. Wiener takes the second law of thermodynamics seriously and follows the implication. The universe heads toward heat death. Progress in the 19th-century sense β civilization rising endlessly toward perfection β is not what physics actually predicts.
What life and civilization are, from a thermodynamic perspective, is local pockets running against the gradient. An organism imports food (low-entropy chemical energy), uses it to maintain its own order, and exports waste (high-entropy heat). A civilization does the same at a larger scale. Both are temporary eddies in a stream that flows toward disorder.
This reframes the question from "will progress continue?" to "what does it take to sustain a viable system?" The rest of the book is variations on that question: applied to learning organisms (Ch 3), to language (Ch 4), to identity (Ch 5), to law (Ch 6), to societies under information control (Ch 7), and to automation (Ch 9).
We are but whirlpools in a river of ever-flowing water. We are not stuff that abides, but patterns that perpetuate themselves. β Wiener, paraphrased
Wiener distinguishes two patterns of communicative behavior:
The political subtext starts here. Brittle organisms die when their environment shifts; brittle institutions do too. Wiener is writing five years after the war, watching the early stirrings of automation, and is already worried about systems whose rigidity will outlast the conditions that justified them.
This is also the chapter that defines feedback as a technical concept: output sampled, compared to a goal, and fed back as input to drive correction. A thermostat is the simplest case. Learning is feedback over a longer timescale, where the goal itself can be revised.
Two systems, same target, same disturbance, same drive strength. The rigid one only watches the error and pushes proportionally β it can't see itself moving. The adaptive one feeds back its own velocity and brakes when it's moving fast. Drag the slider or hit "Shock both" and watch the red line ring while the blue line settles.
Language, for Wiener, is a coded transmission. He breaks it into three layers:
Translation problems aren't surface β they're code-mismatch. Two people using "the same words" can have entirely different semantic and behavioral layers. The same theory, Wiener insists, covers humanβhuman, humanβmachine, and machineβmachine communication. The receiver doesn't have to be conscious for the framework to apply.
This generality is what later lets cybernetics absorb information theory (Shannon, 1948), control theory, and what would eventually become AI. The decoder doesn't need to be alive, but the same questions apply: how much information got through? How much was lost to noise? How is meaning preserved across the channel?
This is the philosophical core of the book β the chapter most often quoted. Wiener argues that identity is pattern, not substance. The atoms in your body turn over completely on roughly seven-year cycles. What persists is not the matter; it's the organization of the matter.
The thought experiment: in principle, a person could be telegraphed. If you could fully encode the pattern β atomic positions, chemical states, memory traces, neural connectivity β the message is the person. Reception and reconstruction would be (massive but tractable) coding problems. Wiener doesn't claim this is feasible in 1950. He claims the question of feasibility is engineering, not metaphysics.
This is the move that sets up everything later science fiction has been chewing on for seventy years. It's also the move that makes cybernetics applicable to organizations: a company is its pattern, not its current employees. A nervous system is its connectivity, not its specific neurons. Pattern is what persists across substrate change.
Below: a simple shape held together by an organizational rule. Each "atom" lives a short while, then is replaced. The atom count completely turns over within seconds. The shape doesn't change.
Wiener treats law as a feedback system regulating society. Disputes are signals; legal decisions are responses; the system as a whole is supposed to maintain some kind of social homeostasis (justice).
His sharpest distinction: common law's case-by-case precedent IS adaptive feedback (the Ch 3 family). It updates as new cases arrive and reveal where prior decisions don't fit. Rigid, fully codified law is the brittle pattern β comprehensive on paper, increasingly out of register with reality as conditions change.
The chapter is a quiet argument against legal systems that try to anticipate every case in advance. Anticipation buys uniformity at the cost of adaptation. Without a mechanism for the law to learn, it ages badly.
The Cold War context is at the surface here. By 1950 the United States had a vast classification apparatus and was actively deepening it. Wiener pushes back hard.
His argument is mostly thermodynamic. Information wants to flow; preventing flow is expensive and incomplete. Classification damages the field that produces the knowledge β the people who can't share, the references that can't be cited, the redundancy that gets prevented. Secrecy is leaky in practice and corrosive to the institution that maintains it.
Information theory gives this a quantitative edge. If you treat secrecy as a channel β the classified domain transmitting to itself, blocked from the open scientific community β you can ask how much capacity is being wasted. The answer is "a lot," and the cost compounds with time.
This is the most personal chapter. Wiener was deeply involved in WWII weapons research β fire-control systems for anti-aircraft guns, which is where much of cybernetics actually came from β and was openly walking away from military work by 1950. The chapter is partly a critique of his colleagues, partly a self-defense, partly a call.
The argument: scientists are responsible for the use of their work. The military-industrial pressure on science (already obvious in 1950, and accelerating) creates a pipeline where curiosity-driven discovery feeds directly into weapons systems with little reflection from the people generating it. Wiener finds this morally untenable.
He's not Luddite. He's not arguing science should stop. He's arguing scientists should refuse to be passive components in someone else's control loop β which is, of course, the same wedge he's been running the whole book. Be an adaptive system, not a rigid one. Have a S5 (policy) layer of your own.
Here Wiener cashes the synthesis. He frames automation as the second industrial revolution:
Mechanization. Steam, looms, factories. Machines replaced muscle. Human role: increasingly tend the machine, supply judgment.
Automation + control systems. Machines replace judgment and control. Human role: unclear, and Wiener says so directly.
In 1950 β with computers as room-sized academic curiosities and no commercial industry to speak of β Wiener writes about job displacement, dangers of automated decision systems, and what happens when control is delegated to machines that have no human stake in the outcome. This chapter is genuinely better than most 1990s writing on the same topics.
The connection back to Ch 2 is direct: automated systems are also entropy-fighting eddies. They sustain themselves by importing energy and information and producing local order. The question is whether their order is aligned with the bigger eddy (society) or eats it.
The final chapter is more concrete. Wiener walks through specific machines and prosthetics he was thinking about: hearing aids, reading machines for the blind (specifically the experimental ones he'd been working on), control systems for prosthetic limbs, automated translation, and the early conceptual ground for what would later be called robotics.
The framing throughout: these aren't replacements for human function β they're extensions of the nervous system into a new medium. A hearing aid isn't a separate device; it's part of the human's own perception loop. A control prosthetic isn't a tool; it's a feedback system spliced into the user's sensorimotor system.
The chapter ends quieter than the polemic chapters that precede it. The book has been about how sustained order is possible in a universe trending toward disorder; the answer Wiener has been building is better feedback, more flexibility, more adaptive coupling between humans and the systems they live inside. The communication machines are an example of doing it right β provided we don't lose track of who they're for.
Click any node to see how it descends from (or runs parallel to) Wiener's 1950 frame. The most direct citation chain runs through Ashby and Beer; the most independent is Prigogine, who arrived at compatible territory from a different department.
Pierre Teilhard de Chardin (1881β1955), Jesuit priest and working paleontologist, wrote the cosmic-evolution version of what Wiener wrote as cybernetics. Same window, same questions, different vocabulary β and a much rougher institutional ride. The Vatican prohibited his publishing during his lifetime. The books came out posthumously and exploded in the 1960s. He's having a moment again now because of and the AI-as-new-industrial-revolution framing his agenda is reportedly built around.
Teilhard was simultaneously a Jesuit priest and a serious paleontologist β on the team that excavated Peking Man (Homo erectus pekinensis) at Zhoukoudian in the 1920sβ30s. This isn't incidental. He was building a unified cosmology that absorbed Darwin, geology, and Christian theology into one frame, and he had hands-on data from deep evolutionary time.
He's writing in roughly the same window as Wiener and Shannon. Le PhΓ©nomΓ¨ne Humain (The Phenomenon of Man) was finished by 1940, circulated in samizdat through the 1940s, and published posthumously in 1955 β five years after Human Use of Human Beings. The conversations with cybernetics never happened directly because Teilhard couldn't publish. But the convergence is uncanny.
Teilhard's central move is to treat the planet as a sequence of layers, each emerging from and wrapping the one below. Geosphere is rock and chemistry. Biosphere is the layer of life β Vernadsky coined the term, Teilhard ran with it. The new claim is the third layer: the noosphere, a planet-scale layer of mind/thought, made of connected human cognition.
He's not being metaphorical. The noosphere is, in his frame, a real emergent layer with its own dynamics β the way the biosphere has its own dynamics distinct from chemistry. As humans network with each other (in his time: writing, telegraph, radio, books), they form a thinking envelope around the planet that has properties no individual mind has. He's describing the internet in 1938. Without an internet.
The cybernetics parallel is direct. Wiener's frame is information-as-physical-quantity flowing through coupled systems; Teilhard's noosphere is what you get when you stack enough such systems on a planet's surface and let them interconnect. Different language, structurally close enough that 21st-century writers (Marshall McLuhan, then everyone) reached for "noosphere" to describe digital networks.
Teilhard's geosphere/biosphere/noosphere stack with a network of minds forming on the surface. Add nodes to grow the population. Toggle "connect" to see the noosphere wire itself together β isolated minds become a coupled system. The "convergence" slider pulls the network toward an Omega-Point attractor: from broadcast-shape (many edges to a few hubs) through dense-mesh to total integration. Watch what happens to variety at each stage.
Teilhard's most contested move: cosmic evolution has a direction, and the direction is convergence. From scattered matter to chemistry to life to thought to β eventually β total integration into what he calls the Omega Point. He framed it theologically: Christ as the attractor pulling history toward unity. Strip the theology and you have "complex systems evolve toward higher coordination," which is structurally close to Prigogine's directional time-arrow and to Beer's recursive-viability nesting.
The non-trivial claim isn't that complexity increases β physics already gave us entropy going up. It's that complexity also organizes, against the gradient, and that this organization has a shape (convergent, hierarchical, self-aware). This is Wiener's Ch 2 position β life as eddy against entropy β extended into a cosmic teleology. Wiener is agnostic about destiny; Teilhard isn't.
The folie-Γ -deux check: Friston's free energy principle (2006+), Prigogine's dissipative structures (1967), Beer's recursive VSM (1972), and Teilhard's Omega Point (1940s) are all structurally similar claims about directional self-organization. That four independent formalisms converged on the same shape is evidence β but it's also the kind of evidence that should make us suspicious of our own pattern-matching. Convergence on a shape isn't proof the shape is real; it can also mean the shape is a strong attractor for human cognition under certain abstractions.
Teilhard wrote his major theological-cosmological works in the 1920sβ40s and was prohibited from publishing them by his Jesuit superiors. The Vatican of that period was dominated by neo-Thomism β a tightly worked-out scholastic philosophy that treats God as the eternal source before creation. Teilhard's frame, where God-as-Omega-Point pulls evolution forward toward a future unification, sounded to the Thomists like making God the result of cosmic process. That smells like pantheism, and pantheism is a hard line for the Vatican.
His most prominent critic was RΓ©ginald Garrigou-Lagrange β the leading Thomist of his era and a heavy presence in Vatican theological gatekeeping. The Society of Jesus effectively exiled Teilhard to China for paleontology fieldwork, partly to keep him out of European theological circles where he was making waves. Books circulated in typescript; formal publication had to wait until after his death in 1955.
After Teilhard's death, his books exploded. The Phenomenon of Man (1955), The Divine Milieu (1957), and the rest sold extensively in the 1960s β riding the Vatican II wave and then 60s counterculture interest in cosmic-evolutionary frameworks. The Vatican responded in 1962 with a monitum (a warning, not a formal condemnation) flagging errors but stopping short of putting him on the Index of Forbidden Books.
The arc since then has been steadily warmer. Cardinal Ratzinger (later Benedict XVI) cited Teilhard favorably in The Spirit of the Liturgy (2000) β using his cosmic-Eucharist framing to talk about the church's relationship to creation. Pope Francis cited Teilhard in Laudato Si' (2015), the encyclical on ecology and climate. And Leo XIV (elected 2025) chose his name explicitly evoking Leo XIII's Rerum Novarum (1891) β the great encyclical responding to the industrial revolution. AI is the new analogous transformation, and Teilhard is the one Catholic thinker with a framework where planetary information networks are part of the spiritual story.
Among 20th-century religious thinkers, Teilhard is the one whose vocabulary lines up cleanly with cybernetics, systems theory, and now AI. Directional cosmic evolution, planet-scale information layer, system-level convergence β every concept has a structural neighbor in Wiener (1948), Beer (1972), Prigogine (1967), or Friston (2006+). That's not because Teilhard was reading them; he was writing earlier and from a different direction. It's because complex systems theory and process theology were both reaching for shapes the older mechanistic frames couldn't provide.
The reason he keeps surfacing in 2025β26 conversations specifically is that the AI industrial revolution framing has made his frame useful in a way it wasn't during the post-Vatican-II period. If you're trying to do cybernetics-flavored thinking inside a Catholic intellectual tradition, Teilhard is the bridge that lets the science and the theology talk without breaking either. That's why he keeps showing up in , in phenomenology departments, and now in arguments about what the next pope is going to do.
Confident on: Bergson influence on Teilhard, the China exile, Garrigou-Lagrange dynamic, Benedict citing Teilhard in Spirit of the Liturgy, the 1962 monitum, the noosphere/ biosphere stack, the Wiener-era timing parallel.
Less confident on: exact statements by Leo XIV about AI specifically (I have the name-choice symbolism and reported general direction, but if you want to cite something specific in a reply, the cite needs verification before sending). Also: the "structurally close to Prigogine/Friston" claim is my framing, not something Teilhard or those authors signed off on.
Wiener writes The Human Use of Human Beings in 1950, between two world wars worth of evidence about what brittle systems do. The book's argument is simple in shape and very hard to apply: living systems sustain themselves by adapting, communicating, and importing energy against a universal entropy gradient. The same framework covers organisms, organizations, languages, laws, and machines. The same failure modes recur at every scale.
The synthesis Wiener opens β entropy plus feedback plus organization β is still being groomed. Beer's VSM is one branch. Prigogine's dissipative structures is a parallel branch arriving at the same questions from chemistry. Friston's free energy principle is the most ambitious modern attempt to merge them into a single formalism. Whether that succeeds is genuinely open. But the company is real, and the book that started it is unusually clear-eyed for 1950.
And running parallel to all of this, on a different continent of intellectual life, is Teilhard's cosmic-evolution frame β same questions, different vocabulary, no citation traffic. The convergence is the interesting thing. When complex-systems theory and process theology, working from completely different starting points, both keep arriving at "directional self-organization toward higher integration," something is going on β though convergence on a shape isn't proof the shape is real, only that it's a strong attractor for human reasoning under certain abstractions. The folie-Γ -deux check stays open. Which is, in the end, the same wedge Wiener was running the whole book: stay adaptive, keep your S5 layer alive, don't lock in a frame just because it feels right.