The machines we build

From its beginnings, the modern tech industry has contained a contradiction. It has repeatedly presented technology as a means of liberating individuals from institutions while simultaneously creating some of the most powerful institutions of the modern age. Silicon Valley inherited the language of personal freedom, decentralization, and rebellion against authority. It also inherited technologies developed through government laboratories, military contracts, and large corporations. Its founders celebrated the individual while building systems that increasingly organized the lives of billions of people.

This tension is often described as a peculiarity of Silicon Valley, an awkward marriage between the hippie counterculture, libertarian capitalism, and the political right. But it is better understood as the latest expression of a much older historical pattern. New technologies frequently enlarge the sphere of individual action while also enlarging the capacity of societies, governments, and organizations to coordinate, monitor, and control. They destroy old hierarchies while making possible new ones.

The resulting problem is more philosophical than simply technological. Human beings want two things that are difficult to reconcile. We want autonomy: the ability to act, move, speak, create, and live according to our own choices. But we also want the benefits that come from collective organization: security, infrastructure, markets, law, knowledge, and cooperation on a scale no individual could achieve alone. The institutions that arise from the new technologies we create inevitably constrain the individuals who created them.

Technology exacerbates, rather than alleviates, the tension between the individual and the collective.

Silicon Valley’s contradictory inheritance

The mythology of Silicon Valley tends to begin with garages, eccentric inventors, and entrepreneurs rebelling against established corporations. There is truth in that story, but it conceals another lineage. Modern computing emerged in no small part from the American state, particularly the defense establishment. Federal research money, military procurement, Stanford University, semiconductor development, and the Advanced Research Projects Agency helped create the technological ecosystem from which Silicon Valley emerged.

The computer was therefore initially a technology of large institutions. Mainframes belonged to governments, universities and corporations. They processed payrolls, scientific calculations, census information and military data. In the popular imagination of the 1950s and early 1960s, computers could easily symbolize centralized, impersonal authority.

At almost exactly the same time, however, the San Francisco Bay Area became a center of the American counterculture. The world of communes, psychedelic experimentation, environmentalism, anti-war activism, and alternative lifestyles appeared to have little in common with military computing. But important elements of the counterculture were not simply left-wing in the conventional sense. They were anti-bureaucratic, suspicious of hierarchy, and intensely interested in personal autonomy.

Stewart Brand became one of the most important figures linking these worlds. The Whole Earth Catalog, first published in 1968, promoted the idea that individuals and small communities could become more self-reliant if they had access to the right tools. Technology, in this conception, need not belong to large bureaucracies, it could enable people to escape them.

That idea proved enormously influential.

The political significance of the personal computer was therefore larger than its technical function. The mainframe represented institutional power. The personal computer represented the possibility that computational power could be placed directly in the hands of individuals.

This reversal mattered enormously. Technology had ceased to appear simply as something done to people by large organizations. It could now be imagined as something individuals possessed and controlled themselves.

The personal computer movement of the 1970s inherited much of this ethos. The Homebrew Computer Club, early hackers and the emerging culture around companies such as Apple combined technical enthusiasm with a striking hostility toward established authority. Steve Jobs, with his interests in Zen Buddhism, India, psychedelics, and countercultural experimentation, embodied some of the contradictions particularly vividly. The same person could reject conventional corporate culture while constructing one of the most valuable corporations in history.

What happened was not so much that the counterculture disappeared as that parts of it became commercialized.

The language of political liberation was gradually translated into the language of technological empowerment. “Power to the people” became “empower the user.” Escaping institutions became owning a personal computer. Independent publishing became blogging. Community became social networking. Self-reliance became entrepreneurship.

This transformation also created the bridge between parts of the countercultural left and the libertarian right.

They disagreed profoundly on economics, property, and social organization, but they shared a suspicion of centralized authority. The hippie commune and the technology startup could therefore emerge from surprisingly similar impulses. Both said, in effect: the existing institution does not work, so we will build something else.

The commune attempted to exit mass society. The startup attempted to disrupt an incumbent.

This common hostility toward bureaucracy helped produce what Richard Barbrook and Andy Cameron later called the “Californian Ideology,” a mixture of countercultural individualism, technological optimism, entrepreneurial capitalism and libertarian economics.

The early internet seemed to vindicate this worldview. Anyone could publish and geography mattered less. Information could circulate without passing through conventional newspapers, broadcasters or governments. Traditional gatekeepers appeared destined to lose power.

The great promise of the internet was therefore not merely connectivity, rather it was decentralization.

But the outcome was more complicated.

The internet weakened mainstream newspapers and broadcasters, but produced Google, Meta, and other digital intermediaries with immense influence over information. Online commerce weakened traditional retailers but created Amazon. Personal computing undermined centralized mainframes only for much of computing eventually to migrate back into enormous cloud platforms. Social media gave billions of people the ability to publish, while a small number of companies acquired extraordinary influence over what those billions of people would see.

The history of Silicon Valley therefore contains an extraordinary inversion. A movement partly animated by suspicion of centralized institutions created some of the most centralized information infrastructures ever constructed.

This is not necessarily the result of conscious hypocrisy. Network effects, economies of scale, capital accumulation and infrastructure costs all reward concentration. But it does reveal a recurrent mistake in technological thinking: disruption is often confused with decentralization.

A new technology may destroy an old hierarchy without eliminating hierarchy itself.

The pattern has become even more visible as parts of the technology industry have moved politically toward the right. Some technology entrepreneurs increasingly regard universities, bureaucracies, mainstream media and regulatory institutions as a new establishment. Their rhetoric of rebellion can sound strikingly similar to the counterculture of the 1960s: question authority, reject gatekeepers, build alternative systems, decentralize power.

The political content has changed, but the structure of the argument remains familiar.

Cryptocurrency promises to escape banks and central banks. Private spaceflight challenges government monopolies. Distributed energy promises independence from utilities. Artificial intelligence promises to place capabilities once monopolized by large professional organizations into the hands of individuals.

The dream remains the same: technology will permit exit from institutions.

History suggests that the dream is never entirely false, just incomplete.

Technology does not merely communicate, it reorganizes power

The freedom-control paradox is not confined to communications. It appears throughout technological history because technologies expand human capacity. They make actions possible that were previously too expensive, difficult or slow.

The key political question is therefore not whether technology produces freedom or control. It is whose capacity it enlarges.

Agriculture itself offers an early example. The development of settled farming created food surpluses, permanent communities, specialization, and eventually cities. Human beings gained greater productive capacity and some protection from the uncertainties of hunting and gathering.

But surplus could also be taxed. Stored grain could support rulers, bureaucracies, priesthoods, and armies. Permanent populations could be counted, organized, and governed more effectively than mobile communities.

One of humanity’s greatest increases in productive capacity therefore helped make possible one of humanity’s most consequential institutions: the state.

Roads illustrate the same duality. Roman roads allowed goods, people, and information to move efficiently across enormous distances. They integrated markets and territories. But they also allowed armies and tax collectors to move rapidly. The road that liberated movement also extended imperial authority.

The history of transportation repeatedly displays this pattern.

Ocean-going ships opened the world to commerce. They also enabled conquest and colonization. Advances in navigation simultaneously expanded the horizon trade and empire.

The railway was similarly ambiguous. Railways allowed ordinary people to travel farther and faster than almost any previous generation. They lowered transport costs, integrated markets, and helped distribute food and manufactured goods.

But railways also strengthened states. Governments could move troops rapidly, administer distant regions and integrate peripheral territories into national economies. Railway timetables encouraged standardized time. Stations, tickets and bureaucratic systems made movement more organized and more legible.

The railway increased mobility while also increasing the capacity of institutions to organize mobility.

The automobile seemed to break precisely this constraint. Unlike trains, cars did not operate according to centralized timetables or fixed routes. The motorist could leave when he wished and travel where roads permitted. Few technologies became more closely associated with personal freedom.

But mass auto-mobilization required an enormous institutional architecture. Roads had to be financed and constructed. Drivers had to be licensed. Vehicles had to be registered. Traffic laws had to be created and enforced. Insurance systems developed. And fuel supply became strategically important.

The individual gained freedom of movement by entering a vast system of collective coordination.

Today the contradiction is stronger still. GPS navigation gives drivers unprecedented independence while simultaneously making precise location tracking possible. Automatic toll systems, license-plate recognition, mobile phones and connected vehicles make motorists increasingly visible.

The technology associated with personal mobility has therefore also become an infrastructure of observation.

Energy technologies reveal another version of the same dynamic.

The steam engine liberated production from the limits of human muscle, animal power, wind and flowing water. It multiplied productive capacity and helped create modern industrial society.

But industrial machinery also transformed the organization of labor.

The factory required workers to assemble at particular places and times. The rhythms of agricultural and household production gave way to the clock. Tasks could be standardized and measured. Managers could supervise large numbers of people performing repetitive functions.

Mechanization freed human beings from some forms of physical labor while simultaneously making labor more disciplined and observable.

Frederick Taylor’s scientific management took this logic further. By breaking work into measurable components, productivity could be increased. But measurement is also a form of surveillance. To optimize work, organizations had to observe workers more closely.

This relationship remains visible in contemporary platform labor. Ride-sharing and delivery applications can provide workers with flexibility unavailable in the traditional factory. But the same software can monitor location, response times, productivity, customer ratings and working patterns with extraordinary precision.

Technology can decentralize the workplace while centralizing management.

Electricity produced a similar bargain. Electric lighting reduced dependence on daylight, electric motors transformed homes and factories, household appliances reduced enormous amounts of physical labor.

But electrification also created large infrastructures: power stations, transmission networks, utilities and regulatory systems. The individual household became more capable while becoming increasingly dependent on a system too complex for any household to reproduce independently.

This pattern might be expressed as a general principle: technology often frees people from one dependency by inserting them into another.

The railway freed people from the limits of walking and horse travel but made them dependent on railway networks.

The automobile freed them from railway timetables but made them dependent on roads and fuel.

Electricity freed households from daylight and localized mechanical power but made them dependent on the grid.

The smartphone frees the user from fixed telephones, physical maps, libraries and offices while making the user dependent on cellular networks, operating systems and digital platforms.

Cloud computing frees companies from maintaining their own servers while increasing their dependence on a small number of enormous infrastructure providers.

Each technological emancipation therefore contains a potential new dependency.

Medicine, identity and the problem of legibility

Medicine provides another important dimension of the freedom-control relationship.

Modern medicine has dramatically increased individual autonomy in some of the most fundamental areas of human existence. Vaccination reduced vulnerability to infectious disease. Antibiotics transformed survival. Anesthesia made complex surgery possible. Contraception gave individuals greater control over reproduction.

Medical technology expanded the choices available to individuals.

But modern medicine also required new forms of organization and classification. Governments and health institutions collected information about births, deaths, diseases, disabilities and vaccination. Hospitals created records. Insurance systems classified individuals according to risk.

The same knowledge that allows institutions to help individuals also allows institutions to categorize and regulate them.

This touches on a broader characteristic of modern society: legibility.

Large societies cannot rely entirely upon personal familiarity and informal trust. Millions of strangers must cooperate. States, markets and institutions therefore need ways of identifying people and making their actions predictable.

Names, addresses, identity documents, property registers, passports, contracts, and databases are technologies of coordination.

They are also technologies of control.

A property register allows someone to prove ownership of a house. Without it, property may be vulnerable to arbitrary seizure. But the same register allows the state to identify taxable property.

A passport allows someone to demonstrate citizenship abroad and travel internationally. But it also allows governments to determine who may cross borders.

A bank account allows an individual to transact securely with people thousands of miles away. But it makes financial activity more visible.

Biometrics can protect against identity theft, terrorism, and fraud, while making anonymity increasingly difficult.

The paradox is therefore unavoidable: the more effectively society can recognize individuals, the more effectively it can serve individuals, but also the more effectively it can regulate them.

Digital technology intensifies this bargain because identification and observation become cheaper.

A state no longer needs an official physically present to know where a person is. A company no longer needs an employee in a shop to observe purchasing behavior. Data generated as a by-product of ordinary activity can make populations extraordinarily legible.

The distinction between convenience and surveillance often becomes a matter of perspective.

Weapons and the redistribution of coercive power

Military technology offers perhaps the starkest illustration of the same process.

Weapons redistribute coercive capacity. The spread of gunpowder weapons weakened certain forms of aristocratic military dominance. A relatively ordinary infantryman equipped with a firearm could become militarily effective against warriors whose power depended on years of specialized training. In that sense, firearms could democratize violence.

But gunpowder also strengthened centralized states. Cannons allowed monarchs to destroy castles, standing armies became larger and more expensive. States thus developed more sophisticated taxation and administrative systems to sustain them.

A technology that weakened one hierarchy helped create another.

Technology could therefore simultaneously restrain violence and magnify its consequences.

Again, the relevant question is not whether the technology is liberating or controlling. It is how it redistributes power.

The recurring cycle of technological liberation

Across these cases a recognizable historical pattern emerges.

A new technology initially disrupts established arrangements. Old gatekeepers do not yet understand it. Entrepreneurs, dissidents, and outsiders exploit opportunities.

Then institutions adapt.

The technology generates new forms of economic organization. Governments regulate it. Large firms consolidate around it. Standards develop. Infrastructure becomes necessary. Systems of observation and enforcement follow.

A rough sequence repeatedly appears:

Innovation — liberation — disruption — consolidation — regulation — new dependency.

Then another technology emerges promising escape from the system that has just consolidated.

This pattern should not be treated as a deterministic law. Technologies differ enormously, and some genuinely produce lasting decentralization. Printing permanently lowered the cost of publication. Cars genuinely increased individual mobility. Personal computers genuinely distributed computational capacity.

The important point is that decentralization rarely occurs in every dimension simultaneously.

A technology can decentralize production while centralizing distribution.

It can decentralize expression while centralizing attention.

It can decentralize mobility while centralizing infrastructure.

It can decentralize energy generation while centralizing grid management.

It can decentralize employment while centralizing workplace surveillance.

It can decentralize access to knowledge while centralizing the computational systems that produce it.

Artificial intelligence may prove an especially powerful example.

For the individual, AI can be radically democratizing. Capabilities that once required programmers, researchers, translators, designers, lawyers, or analysts can increasingly be accessed by a single person.

The user experiences an extraordinary increase in autonomy.

But advanced AI also requires enormous amounts of computing power, electricity, data, specialized chips and capital. Those requirements favor concentration among large companies and powerful states.

The experience at the edge is decentralization.

The infrastructure underneath may be highly centralized.

This is the same contradiction that has followed technology throughout history.

The individual and the collective

These recurring technological paradoxes point toward a deeper political and philosophical problem.

Human beings are individuals, but we are also social animals whose greatest achievements depend upon cooperation. An individual cannot build a national electricity grid, maintain a legal system, create a modern hospital network or defend a large territory alone. Collective organization gives human beings capacities vastly greater than the sum of isolated individual abilities.

But collective organization requires rules. Traffic works because drivers surrender the freedom to drive wherever they please. Property markets work because societies agree on rules governing ownership and transfer. Contracts work because courts enforce obligations. Money works because people accept standards and institutions that establish value. Cities function because people accept building codes, zoning, sanitation systems and taxation.

The mechanisms that allow collective cooperation are therefore also mechanisms of constraint.

This creates a fundamental bargain:

Individuals surrender some autonomy in order to gain the greater capacities produced by collective organization.

The tension can be described as autonomy versus coordination.

But that formulation remains incomplete because institutions do not remain passive tools.

People create institutions, but institutions subsequently shape the people who inhabit them.

No single individual designed the modern market, bureaucracy, state, internet or financial system in the form that it currently exists. These systems emerged through countless interactions, decisions, compromises, and accidents.

Once established, however, they structure individual choices.

Individuals create society.

Society shapes individuals.

Individuals modify society.

Society reshapes individuals.

The relationship is recursive.

Technology accelerates this process because it changes what both individuals and institutions are capable of doing.

The technology begins as a choice and eventually becomes part of the environment within which choices must be made.

A technology that initially offers freedom can eventually become infrastructure. Once it becomes infrastructure, opting out becomes increasingly difficult.

That is why the idea of “exit” is so central to technological politics.

The hippie commune attempted to exit bureaucratic industrial society.

The startup seeks to exit or bypass incumbent firms.

Cryptocurrency seeks to exit traditional finance.

Distributed energy seeks to reduce dependence on centralized utilities.

Artificial intelligence may allow individuals to bypass professional intermediaries.

The attraction is obvious. If institutions restrict autonomy, technology can provide an alternative.

But successful alternatives frequently become institutions themselves.

Google began by helping users navigate information more effectively than existing directories. It became an infrastructure through which information itself was organized.

Facebook allowed individuals to bypass traditional media gatekeepers. Then it became a gatekeeper.

Uber challenged taxi systems. It became the system through which many drivers and passengers encountered one another.

The alternative becomes the institution, and the tool becomes the infrastructure, then the rebel becomes the governor.

The oldest political question in technological form

Seen in this way, arguments about technology are often versions of very old arguments in political philosophy.

Thomas Hobbes asked how much liberty individuals must surrender in exchange for security.

Jean-Jacques Rousseau asked how people could obey collective rules while remaining free.

John Stuart Mill asked where society’s authority over the individual should end.

Karl Marx examined how systems created through human activity could acquire an apparently independent power over the individuals who created them.

Friedrich Hayek emphasized the limits of centralized knowledge and the advantages of dispersed decision-making.

The counterculture asked whether individuals and communities could escape bureaucratic mass society.

Silicon Valley asked whether networks could replace hierarchies.

Libertarians asked whether technology and markets could perform functions previously monopolized by governments.

These questions belong to very different ideological traditions, but they orbit the same problem.

Human beings want the freedom of individuality and the power of collective organization simultaneously.

Without collective institutions, individuals remain vulnerable and limited. They cannot enjoy the infrastructure, security, specialization and cooperation that make complex civilization possible.

But as collective institutions become more powerful, individuals risk becoming subordinate to systems they ostensibly created.

The history of technology is therefore not a linear movement from control toward freedom, nor a descent from freedom into technological domination.

It is an ongoing renegotiation of the relationship between individual autonomy and collective power.

The plow, road, ship, factory, railway, automobile, electrical grid, vaccine, passport, computer, smartphone, and artificial intelligence system all alter that relationship in different ways.

Each expands the range of what human beings can accomplish.

Each forces societies to decide how those expanded capabilities will be organized.

And every time individuals use technology to escape an institution, there is a possibility that the system they build in its place will eventually become another institution from which a future generation will seek to escape.

This is why the strange relationship between hippies, Silicon Valley, and the technological right is historically revealing rather than merely ironic. Beneath their considerable political differences lies a recurring feature of human engagement with new technology.

Technology changes the tools, but the underlying problem remains ancient: how can human beings obtain the extraordinary benefits of living together without becoming prisoners of the systems that make living together possible? No technology has solved that problem. The most consequential technologies merely force us to confront it again.