The Progress Nobody Ordered

When Silicon Valley’s Future Meets America’s Right to Say No

The American midterm elections are being described as a “blue tsunami,” and the expression seems appropriate given the scale of the Democratic victory and the losses suffered by Republican candidates. Beneath the familiar electoral arithmetic, however, something considerably more interesting has taken place. For the first time, Americans have voted on a massive scale against the vision of the future being promoted by Silicon Valley, and the enormous financial resources of the technology industry proved remarkably ineffective in changing their minds.

In political terms, the concept of progress championed by Elon Musk, Sam Altman, Peter Thiel, and their fellow architects of humanity’s technological destiny has suffered a crushing defeat. The American public has begun to question the assumption that every technological innovation automatically constitutes a public benefit, especially when its practical implementation requires people to surrender their land, electricity, water, and eventually their jobs.

Silicon Valley has spent decades explaining to humanity how wonderful its future will be. Humanity, it appears, has finally begun reading the terms of the offer.

The Referendum on Progress

On November 3, alongside the congressional elections, Americans in ten states voted on 41 local ballot measures concerning data centers. The proposals ranged from outright construction bans and temporary moratoriums to limits on electricity consumption, changes in zoning regulations, and requirements for direct public consent before construction could begin.

The results were extraordinary: 40 of the 41 measures went against data-center development.

In some communities, voters prohibited new facilities before any particular project had even been proposed. Apparently, the inhabitants preferred to settle the question while the surrounding fields were still fields and the local authorities had yet to discover the irresistible attractions of corporate investment presentations.

This opposition cuts across traditional political divisions. A September survey by the University of Massachusetts Amherst found that just 11 percent of Americans supported building an artificial-intelligence data center in their community, while 65 percent opposed it. Even among Republicans, traditionally more receptive to large industrial investments, opposition substantially exceeded support.

The electoral consequences were equally revealing. Candidates who had taken early, clear, and uncompromising positions against uncontrolled data-center construction generally fared better than politicians who had enthusiastically welcomed these projects. Participating in ceremonial ribbon-cuttings beside enormous server farms, once a reliable opportunity to demonstrate economic leadership, suddenly became a potentially damaging item in a candidate’s political biography.

Texas provides a particularly instructive example. For years, Governor Greg Abbott promoted the state as a major center of artificial-intelligence development, encouraging the construction of large facilities and promising substantial economic benefits. As public opposition intensified, he gradually revised his position, ordering audits and suspending certain approvals. By September, his administration had effectively imposed a temporary moratorium.

The change came too late. His Democratic opponent, Gina Hinojosa, had already made restrictions on data-center construction a central element of her program, promising a broader suspension until legislation could protect local communities and guarantee their participation in decisions about new developments. Hinojosa won, demonstrating that voters had acquired the politically inconvenient habit of remembering who discovered their concerns first.

Ohio offered another variation on the same theme. Republican gubernatorial candidate Vivek Ramaswamy attempted to make data centers more attractive by proposing that developers provide free electricity, or equivalent compensation, to neighboring residents. He also advocated stricter requirements concerning taxes, environmental protection, and the use of existing industrial sites.

From a commercial standpoint, the proposition was reasonable enough. If a massive industrial installation creates additional costs for the people living nearby, its owners should compensate them. Yet voters elected Amy Acton, who supported a conditional moratorium and stronger public protections.

Naturally, these electoral outcomes cannot be attributed exclusively to the data-center controversy. Inflation, the war with Iran, gasoline prices, dissatisfaction with the administration, and numerous other factors influenced the elections. The significance of the local results lies in the fact that a promise of technological investment, once sufficient to overcome considerable public resistance, has lost much of its former political appeal.

By autumn, defending data centers had become an increasingly delicate occupation, including for Republicans whose campaigns were financed by the very corporations developing them.

The Electricity Bill Acquires an Address

The reasons for this sudden hostility are fairly straightforward, although they extend well beyond the immediate concerns usually cited in public debate.

Inflation is an abstract economic phenomenon, accompanied by percentages, charts, central-bank explanations, and the familiar assurances that conditions are improving at a pace apparently invisible to the people paying the bills. A data center, by comparison, is a very tangible object: an enormous industrial building occupying land that may recently have been farmland, consuming substantial quantities of electricity, sometimes requiring considerable water resources, and placing additional demands on infrastructure maintained by local taxpayers.

Suddenly, rising electricity costs acquire a physical address.

The questions people begin asking are entirely reasonable. Why should a family pay more for electricity because a corporation worth hundreds of billions of dollars has decided to construct a computing facility nearby? Why should local water resources be placed under additional pressure? Why should agricultural land disappear, and who, precisely, authorized such a transformation of the community?

That last question has proved especially troublesome for the technology industry.

Many data-center projects were negotiated through confidential arrangements between developers and local authorities. Nondisclosure agreements restricted public access to information while decisions concerning land, water, electricity, tax incentives, and infrastructure were already taking shape. In some communities, residents learned the practical details only after their elected representatives had committed themselves to the projects.

The referendums consequently became a means of reversing decisions that local officials had made on behalf of the people who elected them. The voters appear to have developed a somewhat old-fashioned understanding of representative government, under which elected officials are expected to represent the interests of their constituents.

Once the public discovered how much had already been decided, opposition became considerably more difficult to manage through conventional public-relations campaigns.

The Factory That Employs Almost Nobody

There is another fundamental problem with Silicon Valley’s proposition, and it reaches directly into the economic meaning of technological progress.

Traditional industrial enterprises had obvious disadvantages. A steel mill, chemical plant, or automobile factory could pollute the air, consume water, create noise, and impose considerable costs on the surrounding environment. Yet such enterprises also employed people, often hundreds or thousands of them, whose wages supported local businesses, housing markets, schools, and public services. Over time, the factory became part of the economic organism of the community.

A modern data center operates according to a very different model. Its construction may generate substantial temporary employment, particularly for skilled trades, while the permanent workforce required to operate the completed facility is usually modest in proportion to its enormous physical size and financial value.

Moreover, the principal product generated by these installations is intended, at least in part, to automate activities currently performed by human beings.

From the standpoint of a local resident, the proposed bargain becomes increasingly peculiar. A corporation arrives, occupies a considerable area, consumes valuable resources, creates relatively few permanent jobs, and develops technologies that may eventually eliminate employment elsewhere in the same community. The inhabitants are then invited to celebrate this arrangement as another triumph of progress.

For the investor, automation promises productivity, efficiency, and expanding returns. For the worker whose occupation may disappear, the same technological achievement can sound remarkably like a dismissal notice delivered several years in advance.

The contradiction has begun to influence the labor movement. Some unions have supported candidates demanding restrictions and greater accountability from data-center developers, while construction unions, whose members benefit directly from building these facilities, have often taken the opposite position. Organized labor therefore finds itself confronting a choice between immediate employment generated by construction and the longer-term consequences of technologies whose commercial value depends partly on reducing the need for human labor.

The industry understands that rising energy costs have become politically poisonous and has begun revising its commercial proposals accordingly.

Some corporations now promise to provide their own generating capacity or pay for the infrastructure required to serve their facilities. Others are considering dedicated power plants, including natural-gas installations operating independently of the public electricity grid. The political message is that local residents will be protected from higher utility bills.

It is a sensible adjustment, particularly when the original proposal effectively required ordinary households to subsidize the operating costs of some of the wealthiest corporations in human history.

The adjustment, however, leaves several important problems unresolved. Independent generating facilities still require land, fuel, equipment, environmental permits, and supporting infrastructure. Water consumption may remain substantial, while air pollution, noise, and the broader consequences of industrial development continue to affect neighboring communities.

Most importantly, providing one’s own electricity leaves unanswered the question of who has the authority to decide how the community develops.

Silicon Valley has responded to a crisis of public legitimacy with a revised financial offer. The election results suggest that the disagreement has already moved well beyond the price.

When Money Becomes a Political Liability

The midterms have also exposed a rather serious weakness in the technology industry’s traditional approach to American politics: its confidence that sufficient campaign spending can eventually overcome almost any obstacle.

Elon Musk reportedly committed approximately $474 million to supporting Republican candidates and political organizations during the election cycle. The results were hardly an advertisement for the efficiency of the investment, particularly in constituencies where public dissatisfaction with the technology industry had already become entrenched.

The cryptocurrency industry received a similar education.

In 2024, crypto-backed political action committees spent roughly $40 million helping defeat Ohio Senator Sherrod Brown, a persistent critic of the industry. They succeeded, reinforcing the belief that politicians who interfered with powerful technological interests could be removed through sufficiently generous campaign financing.

Two years later, the same industry committed even greater resources to preventing Brown’s return to the Senate. The combined spending reached approximately $64 million, yet Brown won.

Apparently, the political marketplace had undergone an adjustment that the industry’s financial models failed to anticipate.

Artificial-intelligence interests experienced similar difficulties. The Leading the Future network, supported by prominent technology investors, reduced spending in competitive general-election races during October and increasingly concentrated its resources on candidates whose prospects were already secure.

Even political action committees, it seems, eventually develop an appreciation for risk management, especially when the money being managed is their own.

A further complication concerned the increasingly sensitive question of how political contributions were being distributed. Campaign-finance records showed substantial payments to recently established corporate entities with limited public visibility. The Campaign Legal Center filed a complaint alleging that affiliated political committees had used intermediary companies to conceal the ultimate recipients of certain expenditures. The committees disputed wrongdoing, and the allegations remain unresolved.

Whatever the eventual legal outcome, the broader political development is revealing. Campaigns that once welcomed Silicon Valley money as evidence of privileged access to the technological future began acquiring reasons to keep such relationships discreet.

The industry had become accustomed to treating financial resources as a reliable measure of political influence. Now, in some constituencies, association with technology billionaires has become a liability in its own right.

For a politician or public-relations consultant, accepting Silicon Valley money increasingly requires calculating the potential damage to reputation alongside the benefits of the contribution. The industry has entered the unfamiliar situation in which additional spending may attract precisely the kind of public attention its beneficiaries would prefer to avoid.

This is a significant change from the political environment of just two years ago, when the same financial resources could often determine the outcome of a closely contested race.

Silicon Valley Meets the Owner

Why has this happened?

The explanation, I believe, reaches deeper than electricity bills, environmental concerns, or dissatisfaction with the companies building data centers. It lies in the difference between the relationship Silicon Valley has spent decades cultivating with its customers and the relationship that citizens expect to have with the communities in which they live.

The technology industry has grown accustomed to dealing with users.

A user can be persuaded through convenience, discounts, free services, improved functionality, carefully designed interfaces, and an endless variety of incentives. If the initial offer fails, the company adjusts its terms. If resistance persists, it develops a more sophisticated proposition, eventually finding the combination of benefits and psychological inducements that produces the desired decision.

Underlying this approach is the assumption that consent has a price and that the essential problem consists of discovering the amount and the method of payment.

This philosophy has produced extraordinary commercial success. It has also encouraged the belief that virtually any human relationship can be understood and managed through the same mechanism, provided the corporation possesses sufficient resources, data, and expertise.

The technology industry has now encountered a category of people its commercial philosophy has rarely had to accommodate: citizens who regard themselves as owners of the land, resources, and institutions that technology companies have grown accustomed to treating as assets available for acquisition.

For these people, the issue concerns the authority to decide what happens in the place where they live. Their expectations include consultation before an agreement is concluded, meaningful participation in decisions affecting communal resources, and the ability to refuse proposals whose consequences they consider unacceptable.

An owner expects his consent to precede the transaction, and he may decide that the transaction should never take place, regardless of the compensation offered.

This is precisely the kind of decision Silicon Valley’s established commercial model struggles to understand. Its success has been built on identifying what people want and developing products capable of persuading them to accept the offer. The current confrontation concerns people who have begun questioning whether they want the product, the development, or the transformation of their communities at all.

The industry can offer cheaper electricity, generous tax arrangements, public infrastructure, and promises of future prosperity. Citizens can still conclude that their land and community are worth preserving in their present form.

Silicon Valley has discovered that some decisions remain beyond the reach of an improved subscription package.

The Future Nobody Asked For

There is also a deeper political transformation underway in the public’s attitude toward technological progress itself.

For decades, the technology industry offered products that made everyday life easier, faster, more entertaining, and sometimes substantially more prosperous. Personal computers, the internet, smartphones, and digital services established a powerful association between technological innovation and the expansion of individual opportunities.

Artificial intelligence introduces a more complicated proposition.

Its most enthusiastic advocates promise unprecedented abundance, dramatic productivity gains, and a world transformed by automation. Machines will perform increasing numbers of tasks previously entrusted to people, while society will supposedly benefit from the resulting efficiency and prosperity.

The problem is that people are beginning to examine their own place in this remarkable future.

A young worker facing uncertain employment prospects, rising housing costs, and an increasingly expensive education is unlikely to find much reassurance in being told that artificial intelligence will soon perform much of the work for which he has spent years preparing. A community asked to surrender land, water, and electricity for the construction of an AI data center may reasonably wonder how much of the promised prosperity will remain within its boundaries.

The promise of abundance becomes considerably less persuasive when accompanied by the prospect of personal economic obsolescence.

For the technology investor, eliminating the need for human labor represents a substantial improvement in productivity. For the people whose labor becomes unnecessary, the same development raises questions about employment, income, social standing, and the possibility of building a secure future.

The contradiction is particularly striking because Silicon Valley’s vision of technological progress is being promoted at a moment when many younger Americans already doubt that they will enjoy the standard of living achieved by their parents.

They are being offered a future of extraordinary productivity while experiencing growing difficulty securing the ordinary material foundations of an independent life. The distance between these two promises has become politically significant.

In this environment, another announcement about the imminent arrival of artificial general intelligence may generate less enthusiasm than its authors expect, particularly among people wondering whether their profession will still exist when the technology arrives.

The American midterms have demonstrated that the public is increasingly willing to distinguish between technological capability and social benefit. A technology may be remarkable, commercially valuable, and strategically important while still imposing costs that citizens regard as unacceptable.

The existence of a technological opportunity does not automatically establish an obligation for society to reorganize itself around it.

The Right to Refuse

The 2026 elections will undoubtedly be remembered for their immediate political consequences in Washington. Their longer-term significance may lie in the confrontation they revealed between technological ambition and democratic consent.

Americans have demonstrated a willingness to challenge the assumption that technological development, merely by presenting itself as progress, acquires an automatic right to transform their communities. The referendums were local, yet the questions they raised reach far beyond individual construction projects.

Who controls the resources required by the new economy? Who receives the benefits of automation, and who bears its costs? What obligations do technology companies have toward the people whose occupations, communities, and ways of life they intend to transform? And perhaps most importantly, who possesses the authority to decide which changes society should accept?

Silicon Valley has considerable resources with which to address these questions. It possesses money, technology, political connections, and a seemingly inexhaustible supply of presentations explaining how wonderful the future will be. What it is beginning to lose is the assumption that the public will accept the answers simply because they have been supplied by people who understand the technology better than everyone else.

This resistance may eventually spread far beyond the United States. Other developed societies face similar pressures on employment, energy, housing, water, and local infrastructure. They will encounter the same fundamental dispute over whether technological progress should develop through decisions in which citizens participate or through arrangements concluded by corporations possessing superior financial and computational resources.

The transformation is especially important because the technology industry has become accustomed to treating public acceptance as a problem of communication, persuasion, and price. The election results suggest that a growing number of people regard it as a question of political authority and ownership, which requires a very different kind of conversation.

For decades, Silicon Valley has trained the world to click a button marked “I Accept the Terms,” usually after presenting an agreement that almost nobody has read and that few people possess the knowledge or inclination to negotiate.

The American electorate has now demonstrated that communities retain a rather more consequential form of consent, exercised through the ballot box and supported by the simple conviction that the land, water, and institutions surrounding them belong to the people who live there.

Silicon Valley may have developed extraordinary expertise in persuading users to accept its products. It now faces the considerably more difficult task of persuading owners to surrender control over their own future.