Social circle of scientist types around world wars

Cybernetics, information, mind.

The other room — the fourth room — was in New York. It had no government funding, no deliverable, no urgency. It had Warren McCulloch as chairman, and over ten conferences between 1946 and 1953, it had basically everyone who would later seed the fields we now call hot.

These were people whose fields hadn’t had their revolution yet. They were doing for biology, computation, and cognition what physicists had done forty years earlier — laying the conceptual foundations of fields that didn’t exist yet. Cybernetics. Cognitive science. Information theory. Systems biology. Complex systems. Artificial intelligence.

If you sit in 2026 and list the fields that feel most alive right now — AI, neuroscience, network science, machine learning, systems thinking, biosemiotics — the seeds of every single one of them were planted in that room.

Let me introduce them one by one.

Warren McCulloch — the chairman

Without McCulloch, there is no Macy. He chaired the conferences, picked the attendees, set the tone, and held the whole thing together for seven years. He was the social and intellectual center of the room.

He was born in New Jersey in 1898 into a religious family and was supposed to become a Christian minister. As a teenager he hung out with prominent theologians and was mentored by a Quaker philosopher named Rufus Jones. The young McCulloch asked Jones a question that would haunt him for the rest of his life: “What is a number, that a man may know it, and a man, that he may know a number?”

Jones reportedly replied: “Friend, thee will be busy as long as thee lives.”

That was the question. What McCulloch wanted to know was: how does the physical meat of a brain produce abstract knowledge? How does the human body — finite, biological, made of cells — somehow grasp infinite mathematical truths? This is the same question Gödel, Russell, and Pitts were each attacking from different angles. McCulloch came at it through the brain itself.

He served in the Navy in WWI, then collected an absurd stack of credentials: BA in philosophy and psychology from Yale, MA in psychology from Columbia, MD from Columbia Physicians and Surgeons. He worked at Bellevue, then at a state hospital for the insane, then at Yale’s neurophysiology lab, then moved to Chicago in 1941 to direct the Illinois Neuropsychiatric Institute. He’d built himself into the rare person who understood philosophy, psychology, neuroanatomy, and clinical psychiatry — all in service of his single question.

Then in 1942, an 18-year-old runaway named Walter Pitts showed up at his door in Chicago. McCulloch and his wife Rook took him in. Within a year, they wrote the paper that founded artificial neural networks.

What made McCulloch the natural chairman of Macy was something rarer than intelligence. He was a connector. He had what one biographer called the temperament of “an intellectual showman” — he held court, told stories, recited his own sonnets at dinner parties, sported a spectacular beard, and made everyone in the room feel like they were part of something exciting. He invited mathematicians and anthropologists to the same conferences because he didn’t see boundaries between fields. He saw one big problem — mind, brain, knowledge — and recruited anyone he thought could attack it.

His range was almost comic. He was a serious neurophysiologist publishing on glucose tolerance and brain mapping. He was a philosopher who wrote essays with titles like “What is a number that a man may know it, and a man, that he may know a number?” and “Why the mind is in the head.” He was a published poet — his collection The Natural Fit was the first book of poetry ever published by the Chicago Literary Club. He designed buildings. He built a dam on his Connecticut farm with his own hands. He chain-smoked and drank whiskey and stayed up until dawn arguing.

He was also, by all accounts, an exceptional mentor. Pitts wasn’t his only adopted intellectual son — McCulloch took in young brilliant misfits the way some people take in stray cats. His house in Old Lyme, Connecticut became a kind of permanent salon, with grad students, postdocs, and various Macy attendees passing through.

The dark turn in his life was the 1952 break with Wiener. The two had been close collaborators — Wiener had championed McCulloch and Pitts’s work, brought them into the Macy circle, helped Pitts get to MIT. Then Wiener’s wife Margaret, reportedly out of dislike for McCulloch’s lifestyle and his closeness with their daughter, told Wiener that McCulloch’s grad students had seduced their daughter in Chicago. The story appears to have been false. But Wiener believed it, refused to ever speak to either McCulloch or Pitts again, and the closest thing the cybernetics movement had to a heart was effectively severed.

Pitts collapsed. McCulloch survived but the movement was wounded. The Macy Conferences ended the next year.

McCulloch worked on at MIT until his death in 1969. He kept publishing, kept mentoring, kept writing poetry. He just never quite got back what he had in those Macy years.

Walter Pitts — the prodigy

We already met Pitts in passing, but he was the soul of the Macy room, so he belongs here properly.

He ran away from an abusive home in Detroit as a child. At twelve, hiding from bullies in a public library, he read all three volumes of Russell and Whitehead’s Principia Mathematica — the same book the Vienna Circle was building on, the same book Gödel had partially demolished — and wrote a letter to Bertrand Russell pointing out errors in it. Russell was impressed enough to invite him to Cambridge. Pitts couldn’t afford to go.

At fifteen, he ran away again, this time to the University of Chicago, where he attended Russell’s lectures when Russell was visiting. He never formally enrolled at any university. Never got a high school diploma, never got an undergraduate degree, never got a PhD. He just hung around brilliant people and absorbed everything.

In 1942, at 18, homeless, he met Warren McCulloch. McCulloch was 44, married, with a family. He took Pitts in. They worked together every night after dinner. Within a year they produced “A Logical Calculus of the Ideas Immanent in Nervous Activity” — the paper that founded the field of artificial neural networks, that gave us the McCulloch-Pitts neuron, that is the direct ancestor of every deep learning system in use today.

Pitts was 20 when it was published.

He then drifted into the Macy circle, became close with Wiener, was treated like a son by him. In 1952, Wiener cut off all contact with Pitts and McCulloch. Pitts, who had no family, no formal credentials, and depended entirely on these relationships, never recovered.

He destroyed his unpublished dissertation — reportedly an extension of the 1943 model to continuous-valued signals and three-dimensional connectivity, which is essentially deep learning. We had to reinvent that idea over the next sixty years.

He started drinking heavily. He died in 1969, at 46, of alcohol-related causes. The same year McCulloch died. Within months of each other. The two men who had built artificial neural networks went out together.

Norbert Wiener — the prodigy who became a prophet

Wiener gave the field its name. He coined “cybernetics” in 1948, from the Greek kybernētēs — “steersman” — to describe the science of control and communication in animals and machines. Without him, there’s no word for what these people were doing. Without him, possibly no Macy at all. He was the gravitational center of the room, even when McCulloch was technically the chair.

He was also one of the most damaged people in this entire story, and that’s saying something.

His father, Leo Wiener, was a Harvard professor of Slavic languages — a Polish-Jewish immigrant who claimed descent from the medieval philosopher Maimonides, who taught himself dozens of languages, who once translated all 24 volumes of Tolstoy into English in 24 months. Leo decided early that he would manufacture a child prodigy. He would prove that genius wasn’t a matter of birth — it was a matter of pedagogy. His son would be the demonstration.

So Leo homeschooled Norbert with brutal intensity. The child was reading Dante and Darwin at seven. He entered high school at nine, graduated at eleven, started college at Tufts at eleven, finished his BA in mathematics at fourteen. He went to Harvard for a graduate degree in zoology, discovered he was hopeless at lab work, switched to philosophy on his father’s instruction, and finished his PhD at Harvard at age eighteen. His dissertation was on mathematical logic.

The price was the child’s psyche.

Leo was an emotional tyrant. The biography by Conway and Siegelman, Dark Hero of the Information Age, draws on family documents that suggest the full picture: Leo would explode in rage at any wrong answer, would publicly humiliate Norbert in front of relatives, would take all credit for Norbert’s achievements while attributing every failure to the boy’s deficiencies. Leo did indeed produce a prodigy, but a damaged one who would spend much of the remainder of his life combating the manic depression and psychological vulnerability induced in his youth.

Wiener wrote two autobiographies about this. The first was titled Ex-Prodigy (1953). The title tells you what he wanted to escape from.

He joined MIT’s math department in 1919, at 24, and stayed there until he died 45 years later. In between, he produced an astonishing range of mathematical work — Wiener processes (the mathematics of Brownian motion, which is the foundation of modern stochastic calculus and financial mathematics), the Wiener filter (still used in signal processing today), and major contributions to harmonic analysis, ergodic theory, and the foundations of statistical mechanics.

He worked on radar-guided anti-aircraft guns during WWII. This is where the seed of cybernetics was planted. The problem with shooting down a German plane was that you couldn’t aim where it currently was — by the time the shell got there, the plane would have moved. You had to predict where the pilot would evade to. Which meant you had to model the pilot as a feedback system: the pilot sees the shells, adjusts course, you predict the adjustment, you fire ahead. The gun, the pilot, and the predictor were one coupled system with information flowing in loops.

Wiener realized this was the same structure as a thermostat, or a homing missile, or the human nervous system maintaining body temperature, or an economy adjusting prices. Feedback was the universal pattern. Information was the substance flowing through it. This was cybernetics.

After Hiroshima, Wiener did something almost unheard of for a scientist of his stature. He publicly refused to do any more military research. He wrote an open letter to the Atlantic Monthly in 1947 titled “A Scientist Rebels,” announcing he would no longer share his work with the Pentagon. He spent the rest of his career trying to redirect cybernetics toward humane uses — labor, medicine, society. His 1950 book The Human Use of Human Beings warned, decades before anyone else, that automation would displace workers and that we needed to design technology around human dignity, not corporate efficiency. Read it now and it sounds like it was written yesterday.

The Macy years were probably the happiest of his life. He had McCulloch as a co-conspirator. He had Pitts as a kind of son. He had a circle of brilliant friends — von Neumann, Shannon, Mead, Bateson — who treated him as a peer rather than as a freak ex-prodigy. He was finally inside a community.

And then in 1952, he ended it himself.

The story is murky and sad. Wiener was deeply attached to his elder daughter, Barbara. His wife Margaret — a German immigrant whom Wiener had married in 1926, reportedly an admirer of Nazi ideology even after the war — disliked McCulloch and the bohemian atmosphere of his Chicago house. Margaret told Norbert that during Barbara’s visit to Chicago, McCulloch’s male graduate students had seduced her. This appears to have been false; the story doesn’t hold up to scrutiny, and Margaret had motive to fabricate it.

But Wiener believed it. And the man who had spent his whole life trying to escape his father’s emotional violence proceeded to inflict the same kind of violence on his closest collaborators. He sent McCulloch a furious letter, ended all communication, refused to attend any Macy meeting McCulloch would be at. He cut off Pitts too — the kid he’d been a father figure to for a decade.

The Macy Conferences ended the next year. The cybernetics movement lost its momentum. Pitts began the slow disintegration that would kill him. McCulloch never quite recovered. Wiener himself spent the rest of his life trying to claim sole credit for cybernetics and writing increasingly defensive autobiographies.

He died in 1964 in Stockholm, of a heart attack, while accepting an honor.

The shape of his life is the most unbearable kind. He was treated like a manufactured object as a child. He grew into a brilliant adult who saw, more clearly than anyone, that feedback and information were the structure of everything — including human suffering. He built a community that gave him, for the first time, the kind of love and equality he’d never had. And then, when his wife told him a story that triggered the old wounds, he destroyed the community with the same emotional violence his father had used on him.

His final book, God & Golem, Inc., published the year he died, was a meditation on whether scientists were creating machines they could not control. He won the National Book Award for it. He was thinking, at the end, about whether the things we build out of our deepest desires turn into monsters that consume us.

He was probably also thinking about himself.

John von Neumann — the universal mind

Most accounts of von Neumann eventually give up trying to describe him and just start quoting his peers. Hans Bethe said: “I have sometimes wondered whether a brain like von Neumann’s does not indicate a species superior to that of man.” Eugene Wigner, a Nobel laureate himself, said that von Neumann was the only genuine genius he had ever met. Edward Teller said: “If a mentally superhuman race ever develops, its members will resemble Johnny von Neumann.”

These were not casual compliments from outsiders. These were people who routinely worked alongside the most brilliant minds of the twentieth century, comparing notes and saying yes, the brightest one is him.

He was born János Neumann in Budapest in 1903, into a wealthy Jewish family — his father was a banker who later received a noble title from the Habsburg emperor, which is where the “von” comes from. By age six he could divide eight-digit numbers in his head. By eight he was doing differential calculus. He spoke Hungarian, German, French, English, and Italian fluently, and read ancient Greek and Latin. His party trick as a child was memorizing entire pages of the phone book and reciting them back to dinner guests on request. As an adult, he could recite from memory long passages of Goethe’s Faust or the entirety of Dickens’s A Tale of Two Cities.

Budapest in his generation produced an absurd cluster of geniuses — the so-called “Martians of Budapest”: von Neumann, Edward Teller, Leo Szilard, Eugene Wigner, Theodore von Kármán, Paul Erdős. All Hungarian Jews, all born within a few years of each other, all ending up at the heart of American science. There’s a joke from Los Alamos that the Martians were a superior alien race who chose to call themselves Hungarians because nobody else could pronounce their names. Von Neumann was the brightest Martian.

His father, sensibly worried that pure mathematics wouldn’t feed a family, made him study chemistry simultaneously. Von Neumann obliged. He earned a chemical engineering degree from ETH Zürich and a PhD in mathematics from Budapest in the same year, 1926, at age 22. His PhD thesis solved a deep problem in set theory that had been open for decades.

By his mid-twenties he had reshaped multiple fields. He gave quantum mechanics its rigorous mathematical foundation (his 1932 book Mathematical Foundations of Quantum Mechanics is still the standard text). He founded modern game theory. He proved the minimax theorem. He invented operator algebras (now called von Neumann algebras). He produced foundational work in ergodic theory.

He arrived at Princeton in 1930. When the Institute for Advanced Study was founded in 1933, he was one of its first six professors, alongside Einstein and Weyl. He was 30. At the IAS, he was reportedly the only person Einstein deferred to on mathematics.

Then came the war, and von Neumann revealed a side of himself that nobody had quite seen before: he turned out to be a brilliant practical problem-solver. The pure mathematician became a weapons designer. At Los Alamos, working with Oppenheimer’s team, he solved the implosion problem for the plutonium bomb. The plutonium “Fat Man” device needed to be compressed to supercritical density by a precisely shaped explosive lens — the geometry was fiendishly difficult, and von Neumann figured it out. Without his math, the Nagasaki bomb does not work.

He also, by some accounts, picked the bombing targets. He sat on the committee that selected Hiroshima and Nagasaki. He argued for choosing cities with intact infrastructure so the damage could be measured precisely. He never seemed to be tortured by this the way some others were — Oppenheimer agonized; Szilard turned against the bomb; von Neumann pushed straight on, advocating for hydrogen bombs and even, briefly, for a preventive nuclear war against the Soviet Union before they could develop their own bombs. (He later softened on this, but never repudiated it cleanly.)

The same brain was also, simultaneously, inventing modern computing. His 1945 “First Draft of a Report on the EDVAC” defined the architecture every computer uses to this day: stored programs, separate memory and processor, sequential execution. We literally still call computers “von Neumann machines.” He oversaw the construction of the IAS machine, one of the first stored-program computers ever built, and used it to run the calculations for the hydrogen bomb.

He attended every Macy Conference. He was deep in conversation with McCulloch and Pitts about the relationship between brains and computers. His 1958 posthumous book The Computer and the Brain asked, in clean technical language, what it would mean for the brain to be a computational device. Almost every modern conversation in AI alignment, computational neuroscience, and theoretical computer science traces back to questions he was raising in the early 1950s.

He also invented cellular automata — the idea of grids of simple cells following simple rules that produce complex behaviors — which directly led to John Conway’s Game of Life, Stephen Wolfram’s work, and modern artificial life research. He proved that machines could in principle self-replicate, opening the door to nanotechnology and synthetic biology.

He did all this while throwing constant loud parties. He was famous for them. His house in Princeton hosted weekly gatherings where mathematicians, physicists, politicians, and military brass drank and argued until 2 AM. He drove cars badly — he was in multiple accidents because he liked to do arithmetic while driving. He preferred sloppy mathematics that worked over elegant mathematics that didn’t. He was the rare pure mathematician who openly enjoyed money, power, and the company of generals.

His personality was almost the opposite of every other figure in this essay. He was warm. Sociable. Funny. Confident. Generous to younger mathematicians. Unburdened by visible neurosis. Where Gödel saw poisoners in every shadow, von Neumann saw cocktail parties to attend.

And then in 1955, the universe corrected.

He developed shoulder pain. It turned out to be bone cancer — possibly caused by radiation exposure at the Bikini Atoll bomb tests he had observed. By 1956 it had metastasized to his spine and then his brain. The brain that had computed the implosion lens, the von Neumann architecture, game theory, quantum foundations — that brain was being eaten by cancer.

He didn’t take it well.

This is one of the most disturbing endings in the history of science. The man who had been confident and rational and unflappable his entire life fell apart completely as his mind failed. He had been a non-practicing Jew his whole life, sometimes openly contemptuous of religion. As death approached, he summoned a Catholic priest and was received into the Catholic Church. (Asked why, he reportedly cited Pascal’s Wager — if God exists, there’s infinite payoff to belief; if not, you’ve lost nothing. It was a game-theoretic conversion.)

He screamed in terror at night. His hospital room at Walter Reed Army Medical Center was guarded by military personnel because his colleagues were genuinely afraid he would mutter classified information in his pain or delirium. Generals sat by his bed taking notes in case he said something operationally important.

His brother read to him from Goethe’s Faust in the original German as he was dying. He could still complete sentences his brother started. The brain was failing but the deep memory remained, dredging up lines memorized fifty years earlier in Budapest.

He died on February 8, 1957, age 53. Just at the moment when his ideas about computers, automata, and self-replication were about to become the foundation of the digital age he had largely invented.

Gödel, who almost never wrote to anyone except Einstein, sent him a letter a few months before his death. After perfunctory inquiries about his health, Gödel got to what he really wanted to say: he had been thinking about a problem. Could every mathematical question that can be checked quickly also be solved quickly?

This is now called the P vs NP problem. It is the most important open question in theoretical computer science. Gödel sent it, in the form of a private letter, to the dying von Neumann. He thought, correctly, that von Neumann was the one person who might understand it.

We don’t know if von Neumann was lucid enough by then to read it. The letter went unanswered.

The two greatest logicians of the century, in their final exchange, were puzzling over the limits of computation — the field von Neumann had just founded. Sixty-five years later, we still don’t know the answer.

Claude Shannon — the playful one

If everyone else in this essay seems to have lived life at extremes — Pitts in alcohol, Gödel in paranoia, Wiener in his father’s shadow, von Neumann at the bomb tests — Shannon is the one figure who seems to have just enjoyed himself.

He gave us information theory. Not “contributed to” — gave us. There were people before him thinking vaguely about communication, but Shannon’s 1948 paper “A Mathematical Theory of Communication” essentially landed from outer space, fully formed, and created an entire field that didn’t exist the week before. He defined the bit. He proved you could send arbitrarily large amounts of information through a noisy channel with arbitrarily small error if you encoded it cleverly enough — a result that seemed impossible and turned out to be true. Every cell phone call, every Wi-Fi connection, every Mars rover signal, every DNA sequencing run, every compression algorithm is built on Shannon’s math.

His master’s thesis at MIT, written when he was 22, applied Boolean algebra to the design of electrical circuits. It has been called “possibly the most important master’s thesis of the twentieth century.” He showed that any logical operation could be implemented by switches arranged in series and parallel — turning logic into engineering and engineering into logic. Every digital circuit ever designed traces back to this thesis. He did it in his spare time.

He was born in 1916 in Michigan. As a child he ran a telegraph line between his house and a friend’s house out of barbed wire on top of a fence. He grew up reading Edgar Allan Poe — his favorite story was “The Gold Bug,” about a man decoding a cryptogram to find buried treasure. He stayed interested in cryptography for the rest of his life; during WWII he worked at Bell Labs on encrypted voice communication (the famous SIGSALY system that let Roosevelt and Churchill talk securely) and met Alan Turing for several months in 1943 — they had lunch together every day, talking around the edges of their separately classified projects.

After the war, he wrote the information theory paper, attended the Macy Conferences, and then — this is the part nobody quite expected — he basically stopped doing serious work and devoted himself to having fun.

He took a professorship at MIT in 1956. He didn’t really like teaching, and after a few semesters told MIT he’d prefer to stop. They let him. He spent the rest of his career in a kind of self-designed retirement, in his basement workshop in Winchester, Massachusetts, building things for no particular reason.

The list of things Shannon built for fun is one of my favorite documents in the history of science:

• A mechanical mouse named Theseus that could navigate a maze and remember the solution. It is plausibly the first machine to demonstrate machine learning. He built it in 1950.
• THROBAC (THrifty ROman-numeral BAckward-looking Computer) — a calculator that did arithmetic entirely in Roman numerals, because he thought it was funny.
• A juggling W.C. Fields automaton that was supposed to juggle but mostly didn’t work.
• A “Useless Machine” — a box with a switch on it; when you flipped the switch, a mechanical hand emerged from the box, flipped the switch back off, and retracted. Marvin Minsky designed an early version; Shannon built and refined it. There is no purpose.
• A two-seater unicycle — nobody wanted to ride it with him.
• A unicycle with an off-center hub, which made him bob up and down like a duck as he rode it. Colleagues came out into the corridors to watch.
• Flame-throwing trumpets.
• A rocket-powered Frisbee.
• Plastic foam shoes he used to walk across a nearby lake. From a distance it looked like he was walking on water.
• A motorized pogo stick — he said this would let him retire the unicycle his colleagues feared, but he never did.
• A wearable computer that he and Ed Thorp used to beat the roulette tables in Las Vegas. (Yes — Shannon was a successful card and roulette cheater. He and Thorp’s wearable computer worked by predicting where the ball would land based on its initial velocity and rotation. They split the winnings.)
• A chess-playing machine — his 1950 paper on computer chess essentially founded the field. Every chess engine since traces to Shannon.

He famously rode his unicycle up and down the long corridors at Bell Labs, often juggling at the same time, nodding politely at colleagues as he passed. Nobody knew whether the juggling-while-unicycling was preparation for some new theorem or whether he just liked it. The answer was probably: both.

He fell in love with juggling specifically. He developed a mathematical theory of juggling — a relationship between number of balls, number of hands, time in hand, time in flight, and time empty. He published it in 1980. He could juggle four balls reliably. He wanted five but his hands were too small; he often complained about this.

In his eighties his memory began to fail. He was diagnosed with Alzheimer’s. His wife Betty cared for him at home as long as she could. He died in 2001, at 84. By the end, he didn’t really remember that he had invented information theory or that the entire digital civilization being built around him owed him its existence.

What I love about Shannon is that he is the one figure here who doesn’t seem to have been driven by suffering, ambition, or fear of failure. He just kept doing what amused him. The same brain that produced the most important master’s thesis of the twentieth century also produced flame-throwing trumpets. He didn’t apparently see a difference. Both were objects of curiosity. Both were fun to figure out.

In the social geography we’re mapping in this essay — the bomb-builders driven by urgency, the cyberneticians driven by ambition, the loners driven by inner demons — Shannon stands slightly apart. He attended Macy. He was respected by everyone. But he was never in any single project the way Wiener or McCulloch were. He just happened to live nearby and drop in.

He’s the reminder that you don’t have to be tortured to do great work. You can also just be a slightly strange person who likes building things and following your curiosity, and the great work happens as a side effect.

Gregory Bateson and Margaret Mead — the anthropologists who came home

Most of the figures in this essay are people I’d describe as central to one thing — Wiener to cybernetics, Gödel to logic, Shannon to information theory. Bateson and Mead are different. They were anthropologists who came at the Macy questions sideways, from human cultures rather than from mathematics or biology. And their lives are tangled with each other in a way that makes it almost impossible to tell either story alone.

Mead was the more famous of the two. She was probably the most famous anthropologist of the twentieth century, period. Born in Philadelphia in 1901, daughter of academics, she did her PhD at Columbia under Franz Boas — the founder of American cultural anthropology, the man who established that “race” was a social construction rather than a biological category. Mead was Boas’s intellectual heir.

At 23 she went alone to American Samoa to study adolescent girls. The resulting book, Coming of Age in Samoa (1928), was a bombshell. She argued that the angst, sexual repression, and identity crises that American teenagers were going through were not biological inevitabilities but products of American culture. In Samoa, she reported, adolescence was easy, sexuality was relaxed, and young women moved into adulthood without trauma. The implication — explosive in 1928 — was that American sexual mores were arbitrary, and we could choose differently.

The book sold huge numbers. It made Mead a public intellectual at 27. For the rest of her life she would be the anthropologist most Americans had heard of, the one who showed up on TV, testified before Congress, wrote columns for Redbook magazine. She made anthropology mean something to the general public in a way it never had before and never has since.

Her personal life kept pace with her ideas. She married three times. Her first husband, Luther Cressman, was a theology student she married at 22 and divorced at 27. Her second was Reo Fortune, a brilliant and difficult New Zealand anthropologist she met on a boat back from Samoa.

It was with Fortune, in 1932, deep in the Sepik River region of New Guinea, that she met Gregory Bateson.

Bateson was English aristocracy, in the intellectual sense. His father, William Bateson, was a famous biologist who coined the word “genetics.” William named his son Gregory in honor of Gregor Mendel, the monk who had founded genetics. The pressure on the boy must have been intense, and it got worse: Bateson had two older brothers, both of whom were expected to be the scientific stars of the family. The eldest died in WWI. The second committed suicide. Gregory, the unloved third son who was supposed to be the family disappointment, was suddenly the only son left.

He went to Cambridge, studied biology to please his father, switched to anthropology when his father couldn’t stop him, and ended up — through a strange chain of circumstances — in New Guinea among the Iatmul people, alone, learning their language and studying a ritual called naven in which men dressed as women and women dressed as men to celebrate the achievements of their nephews and nieces.

Mead and Fortune turned up at his fieldsite. They had been miserable together for months. Bateson was lanky, brilliant, English in the way that Mead, who was small, intense, and American, found exotic. The three of them sat together in mosquito nets in the jungle for weeks, theorizing about culture and personality. It was an actual love triangle in a real jungle. Mead and Bateson fell in love. By the end of the trip, Mead’s marriage to Fortune was finished.

She divorced Fortune in 1935. Married Bateson in Singapore in 1936. They went immediately to Bali for two years of fieldwork. They produced together one of the most extraordinary ethnographic archives ever assembled — 25,000 photographs, 33,000 feet of motion picture film — documenting Balinese culture, child-rearing, trance, dance, and the entire texture of village life. They invented modern visual anthropology in those two years.

They had one daughter, Mary Catherine Bateson (who later became an important anthropologist herself).

The marriage didn’t survive World War II. The exact reasons are murky — there are letters indicating Mead’s bisexuality and her ongoing emotional entanglement with the anthropologist Ruth Benedict; there are accounts of Bateson’s complicated personality and his difficulty with the constant Mead-as-celebrity dynamic. They separated in 1947 and divorced in 1950. They remained professionally close and friendly for the rest of their lives.

What’s relevant for our story is what happened after the divorce. Both of them, separately, found their way into the Macy circle.

Mead and Bateson had been at the very first Macy meeting in 1942 — they were among the original organizers, helping Frank Fremont-Smith plan it. They brought to the cybernetic conversation something nobody else in the room had: the anthropological mind. Where Wiener and von Neumann thought about feedback in terms of guided missiles and computers, Mead and Bateson thought about it in terms of human cultures, family dynamics, ritual cycles, mother-infant interactions.

This turned out to matter enormously.

Mead’s specific contribution to Macy was insistence that human social systems could be analyzed with the same cybernetic tools as machines. She pushed the field toward what would become family systems theory, organizational behavior, and modern sociology of communication. Without Mead, cybernetics would have stayed an engineering discipline.

Bateson went deeper. After the divorce, he moved to California, ended up working at a VA hospital in Palo Alto studying schizophrenia, and in 1956 published the paper that made his name: “Toward a Theory of Schizophrenia.” In it, he proposed the double bind theory — the idea that schizophrenia could arise from chronic exposure to contradictory communications you can’t escape or comment on. (E.g., a mother who says “I love you” while physically withdrawing from her child, and who punishes any attempt to point out the contradiction.) The theory turned out to be wrong about schizophrenia specifically, but it founded family therapy as a discipline and reshaped how we think about communication and pathology.

He kept going. He studied dolphin communication. He studied octopus learning. He thought about ecology long before it was fashionable. His book Steps to an Ecology of Mind (1972) is a collection of essays that includes some of the most important early writing on how minds are not just brains — they are processes distributed across organisms and environments. He’s the great-grandfather of what’s now called 4E cognition (embodied, embedded, enactive, extended). He’s also the intellectual hero of much of the environmental movement, the systems thinking tradition, and the school of psychotherapy that runs through Milan, Palo Alto, and modern family therapy.

In the 1970s, Bateson became something nobody predicted: a guru figure for the California counterculture. He lived at Esalen Institute for a while. He gave lectures to crowds of hippies who were trying to figure out how minds, ecosystems, and societies fit together. He talked about epistemology — the question of how we know things — with the seriousness of someone who realized the planet’s survival might depend on getting it right. He died in 1980 at the Zen Center in San Francisco, in the kind of place where his late-life thinking had finally found a home.

Mead kept her own arc. She became a kind of public moral authority — a small, sharp, white-haired woman with a forked walking stick that became her trademark. She advocated for women’s rights, civil rights, nuclear disarmament. She popularized the idea that culture shapes character. She was so visible that the FBI maintained a file on her. She died in 1978, of pancreatic cancer, at 76.

In their final years they had circled back to each other. They had been friends, lovers, collaborators, ex-spouses, co-parents, intellectual peers. They had been on opposite coasts but in constant correspondence. When Mead died, Bateson outlived her by less than two years.

What’s worth understanding about them, for the Macy story, is that they were the anchor in human reality for a movement that could easily have floated off into pure mathematics. McCulloch and Pitts could have built artificial neurons forever without anyone asking what they meant for human lives. Wiener could have stayed inside his equations. Von Neumann could have stayed inside his bombs. Mead and Bateson kept dragging the conversation back to: what does this mean for how mothers talk to children? What does this mean for cultures? What does this mean for ecosystems?

They are the reason cybernetics seeded family therapy, environmental thought, organizational psychology, and the whole sprawling complex-systems tradition that goes through Maturana, Varela, and onward. The pure mathematicians made the math. The anthropologists made it matter.

Heinz von Foerster — the Viennese magician

Some of the people in this essay you’d want to have had as a teacher. Heinz von Foerster is the one you’d want to have had as a grandfather.

He lived to be 90. He was small, charming, played the piano, did stage magic tricks, loved Wittgenstein (he had memorized large parts of the Tractatus), and spoke with a Viennese accent until the day he died. He was a kind of human bridge between the world of pre-war Vienna — Wittgenstein, the Vienna Circle, Schoenberg, Klimt — and the world of late-20th-century California cybernetics. The world that started with Gödel quietly attending Carnap’s lectures ended with von Foerster lecturing at Esalen to crowds of curious hippies. He carried that thread.

He was born in Vienna in 1911 into an aristocratic family of architects, artists, and intellectuals. His great-grandfather had designed half of Vienna’s Ringstrasse. His mother was a feminist activist. Ludwig Wittgenstein was a distant cousin and, by some accounts, a childhood playmate. As a teenager, Heinz occasionally sat in on Vienna Circle meetings — the same meetings the young Gödel was attending silently — and absorbed the logical positivist project from inside, even as he came to distrust its certainties.

He studied physics at the Technical University of Vienna. Then the Nazis came.

Von Foerster was partly Jewish on his grandfather’s side. In Vienna this would have been a death sentence. He did something audacious: he moved to Berlin, the capital of the Reich, on the theory that the safest place to hide was in the middle of the storm. With the help of an employer who declined to check his papers too carefully, he worked in radar laboratories throughout the war, hiding his ancestry in plain sight. He completed his physics PhD at the University of Breslau in 1944, in the final year of the war. His wife and three sons survived with him.

When the war ended he went back to a ruined Vienna and tried to start a career. The city had nothing to offer him. In 1949, he emigrated to the United States and arrived at the University of Illinois with almost no English.

His arrival at Macy is one of the great accidents of cybernetics history. He had written a short book in German called Das Gedächtnis (Memory) — a quantum-physical theory of how human memory might work. McCulloch read it and was charmed. McCulloch invited him to the next Macy Conference, immediately, despite the fact that von Foerster’s English was so poor he could barely follow the conversation. McCulloch then assigned him to be the editor of the Macy conference proceedings, with the explanation that this would force him to learn English fast. It worked. Von Foerster edited the proceedings of the last five Macy conferences. He learned English by listening to the cleverest people in America argue about cybernetics.

It was von Foerster who, after Wiener published his 1948 book, suggested that the Macy conferences adopt “cybernetics” as their official name. The brand-naming of the field is partly his.

After Macy ended, von Foerster founded the Biological Computer Laboratory at the University of Illinois (1958-1975), which became the institutional home of post-Macy cybernetics. He published densely, taught generations of students, and developed his most distinctive idea: second-order cybernetics.

Here’s the move. First-order cybernetics studies feedback systems — thermostats, missiles, brains — as if they were objects you could observe from the outside. The observer is separate from the observed. Second-order cybernetics asks: what about the observer? The observer is also a feedback system, also a cognitive process, also subject to the same dynamics. You can’t cleanly separate “the system being studied” from “the system doing the studying.” Knowledge is not a mirror of reality; it’s a construction by an organism trying to maintain itself.

This sounds esoteric but it has huge downstream consequences. Second-order cybernetics is the philosophical foundation of radical constructivism in epistemology, of autopoiesis in biology (Maturana and Varela explicitly built on von Foerster), of family therapy in psychology (where the therapist is part of the system they’re treating), and of large parts of complexity theory. The idea that observers are part of what they observe — which is a recurring intuition you find in quantum mechanics, in anthropology, in psychology, in modern AI alignment thinking — has von Foerster as one of its main twentieth-century formalizers.

He retired in 1975, moved with his wife to a remote ranch in Pescadero, California, and spent the next 27 years as one of the most genial elder statesmen of complexity theory. He gave interviews. He hosted visiting scholars in his living room. He did magic tricks. He told stories about Vienna in the 1920s. He coined aphorisms — his favorite was “act always so as to increase the number of choices.”

He died in 2002 at 90, the last of the original Macy core to go. He outlived all of them by decades — Pitts by 33 years, McCulloch by 33, Wiener by 38, von Neumann by 45, Mead by 24, Bateson by 22. He was the last living witness to that room.

W. Ross Ashby — the British psychiatrist who built brains

Ashby is the least famous of the names in this section and the most professional in his life — no scandals, no tragedies, no breakdowns. Just a quiet British psychiatrist who built one of the most important conceptual machines of the twentieth century.

He was born in London in 1903. Trained as a medical doctor. Specialized in psychiatry. Worked for decades at the Barnwood House Hospital and the Burden Neurological Institute in England, treating patients while thinking about the deeper question: how does a brain adapt to its environment?

In 1948, he built a machine he called the Homeostat — a small electromechanical device that could re-stabilize itself when disturbed. You’d push it out of equilibrium and it would, through internal feedback, find a new equilibrium. It wasn’t pre-programmed for any specific environment; it would adapt to whatever conditions you put it in. Ashby called it “the closest thing to a synthetic brain so far designed.”

The Homeostat was, in a real sense, the first artificial self-organizing system. It was an existence proof that you could build a machine which exhibited learning-like behavior without being told what to learn. Modern reinforcement learning, modern adaptive control, modern artificial life — all of them have the Homeostat in their ancestry.

He attended the Macy Conferences as one of the British representatives. He brought a peculiarly British combination of empirical psychiatry and pure mathematics. He wasn’t flashy. He didn’t tell stories. He just kept showing up with rigorous formal models of how complex systems could self-regulate.

Two books made his reputation: Design for a Brain (1952) and An Introduction to Cybernetics (1956). The latter is still the cleanest textbook introduction to cybernetic thinking ever written — generations of complex-systems researchers have started there. He formulated the Law of Requisite Variety: only variety can absorb variety, meaning that for a control system to handle a complex environment, the control system itself must be at least as complex as the environment. This sounds obvious until you realize it’s a hard mathematical constraint that explains why simple bureaucracies fail at complex problems, why ecosystems resist single-cause interventions, and why one-size-fits-all policies break.

Ashby moved to the University of Illinois in 1961, joining von Foerster’s Biological Computer Laboratory. He spent the rest of his career there, with quiet collaborations and steady publishing. He died in 1972 at 69, of natural causes, in his sleep, at home in England where he had returned to retire.

His life has no biographical drama at all. No childhood trauma, no breakdowns, no rivalries, no scandals. He married once, stayed married, had two daughters, did his work, retired, died. The Wikipedia article on him is short.

But the ideas he produced are still pulling weight sixty years later. Every time someone designs an adaptive system — a thermostat, an autopilot, an AI training loop, an ecological policy intervention — they are working in territory Ashby mapped. He’s the reminder that the Macy Conferences weren’t just full of larger-than-life characters. Some of the people in the room were just quiet professionals doing extraordinary work because they took the questions seriously and stayed at them.

You don’t need to be a tragic figure to change the world. Sometimes you just need to build the Homeostat.