The Social Meaning of AI

Capital, Government and the Chinese Contrast

The contrast is difficult to miss. In China, artificial intelligence appears to have been adopted rapidly and with considerable public confidence. People use it at work, incorporate it into ordinary digital routines and encounter it as another stage in the country's continuing technological development. In the United States, AI has generated a remarkably dense collection of anxieties. Will it take my job? Will relying on it diminish my ability to think? Will companies use it to appropriate my personal information? Will its data centers consume intolerable amounts of electricity and water? Will synthetic images, voices and documents make it impossible to know what is real?

These attitudes are often treated as cultural differences. Chinese people are supposedly enthusiastic about technology, while Americans are skeptical and individualistic. The contrast becomes more intelligible when placed within the political and institutional systems through which people encounter technological change. Americans have good reason to expect that corporations will capture the benefits of AI while workers, consumers and communities absorb its costs. Chinese people have more reason to expect technological development to be regulated, coordinated and incorporated into a public project. American anxiety about AI is therefore a form of political knowledge: people are anticipating what powerful institutions are likely to do with the technology.

The empirical difference is substantial. Stanford's 2026 AI Index found China among the countries with the highest levels of excitement, workplace use and trust in AI. More than 80 percent of Chinese respondents reported using AI regularly at work, while 78 percent expected it to reduce the time required to accomplish tasks. Chinese respondents expressed some nervousness, but at levels below those found across much of North America and Europe.

The American pattern runs in the opposite direction. A Pew Research Center survey conducted in June 2026 found that 52 percent of Americans were more concerned than excited about AI, while only 9 percent were more excited than concerned. Seventy-one percent expected AI to produce fewer jobs during the next twenty years. The Stanford report also found that only 31 percent of Americans trusted their own government to regulate AI responsibly, the lowest figure among the countries surveyed.

This last result provides a key to the others. Americans are evaluating AI through their accumulated experience of government, corporations and technological change. They have watched productivity rise while wages stagnated, employment become more precarious, unions decline, public services deteriorate and wealth accumulate at the top. They have watched the internet develop from a publicly funded communications system into an infrastructure dominated by advertising platforms, private surveillance and a handful of immensely powerful companies. When another transformative technology arrives under the control of many of those same institutions, apprehension is a reasonable response.

Five experiences of institutional abandonment

The fear that AI will take people's jobs begins with the structure of American employment. An employer that uses AI to increase output can distribute the resulting gains to executives and shareholders while dismissing the workers whose labor has become unnecessary. The displaced worker must then manage the loss of income, health insurance, retirement contributions and professional identity. Retraining also becomes a personal responsibility, undertaken without any assurance that the newly acquired skills will remain valuable. Under these conditions, greater productivity appears as a threat because the institutions controlling production have no obligation to share its benefits.

The fear that AI will make people stupid expresses a related distrust. Schools and employers operating under financial pressure have strong incentives to use AI to reduce costs, accelerate production and automate evaluation. Their institutional interest lies in obtaining more measurable output with fewer teachers, researchers, writers and other skilled workers. Whether AI helps a person develop judgment, memory, interpretive ability or intellectual independence remains largely outside the calculation. The individual is expected to resist dependency while living inside systems designed to reward speed and convenience.

The fear of data theft follows several decades in which American companies have treated human activity as a source of extractable information. Searches, purchases, movements, conversations, relationships and preferences have been collected, combined and sold through systems that few people can meaningfully understand or refuse. AI increases the value of these accumulated records and creates new uses for information originally surrendered for entirely different purposes. Americans are consequently asked to protect their privacy through passwords, settings, consent screens and individual vigilance within an economy built around surveillance.

Environmental anxiety follows the same distributional pattern. AI companies and investors receive the profits produced by new data centers, while surrounding communities may experience increased electricity demand, water consumption, noise, land use and infrastructure costs. Individuals are encouraged to consider the environmental consequences of their own AI queries even though they exercise almost no control over model design, energy sourcing, cooling systems or the location of computing facilities. An industrial planning problem is translated into a consumer's moral dilemma.

Misinformation may be the most revealing anxiety because it concerns the conditions of shared reality. Generative AI makes it possible to produce synthetic photographs, voices, witnesses, documents and entire histories at enormous scale. At the same time, commercial journalism has been hollowed out, social platforms profit from circulation rather than accuracy, and public institutions have lost much of their authority. The citizen is expected to inspect sources, identify deepfakes, recognize propaganda and authenticate every apparent piece of evidence personally.

This amounts to the privatization of epistemic security. Even an unusually careful and educated person cannot independently verify an unlimited stream of synthetic material. Once the informational environment becomes radically untrustworthy, private skepticism cannot reconstruct the common world required for public judgment. Yet American institutions increasingly assign that impossible task to each individual user.

These five fears describe different aspects of the same experience. Americans expect corporations to develop and deploy AI according to the demands of capital accumulation. They expect public authorities to assist that process, moderate some of its most conspicuous abuses and leave individuals responsible for surviving the resulting disruptions. Each anxiety asks, in a different form, who will protect ordinary people when private technological power becomes socially destructive.

AI within the Chinese developmental project

China places AI inside a different relationship among government, technology and society. Official policy presents it as part of a coordinated developmental program encompassing manufacturing, transportation, healthcare, education, agriculture, public administration, poverty reduction and environmental management. China's Global AI Governance Action Plan explicitly calls upon the public sector to guide AI deployment in healthcare, education and transportation while developing standards for safety, privacy, energy efficiency and water use.

This resembles the future presented in Chinese planning halls. Technology appears within a publicly narrated account of collective development. Digital infrastructure, transportation, housing, environmental restoration, healthcare and education are represented as interdependent components of a common project. AI can therefore be understood as another social capacity through which that future will be constructed. Americans encounter the same technology primarily through corporate product announcements, investor valuations, employer directives and warnings that everyone must adapt or become obsolete.

Chinese regulation also establishes a clearer expectation that technology companies are subject to public authority. The country's rules for generative AI encourage innovation while imposing responsibilities concerning personal information, discrimination, harmful content, accuracy and transparency. Providers must protect user information, respond to complaints and take corrective action when their systems produce unlawful material. China's Personal Information Protection Law provides rights to access, correct and erase personal data, prohibits unreasonable algorithmic price discrimination and allows penalties reaching 5 percent of a company's annual turnover.

Environmental consequences are similarly treated as matters of infrastructure planning. China has established efficiency requirements for large data centers, coordinated their geographical distribution and connected computing development with renewable-energy policy. Its government recognizes the enormous energy requirements of AI, while retaining the administrative capacity to influence where facilities are built, how efficiently they operate and which energy systems support them.

Chinese people still confront employment pressures, algorithmic management, intrusive technologies and misinformation. Their relative optimism indicates greater confidence that these consequences belong within the sphere of governance. When a technology creates a social problem, the government is expected to intervene, establish standards, discipline companies, redirect investment or provide a collective response. The corresponding American problems generally return to the individual as instructions to reskill, protect personal data, verify information, supervise children, preserve intellectual independence and monitor personal environmental consumption.

Some of the contrast also reflects different historical experiences of development. India, Indonesia, Nigeria and several Southeast Asian countries report relatively high levels of AI optimism as well. Societies that have recently experienced rapid improvements in infrastructure, communications and public services have reason to associate technology with expanded collective capacity. In a deindustrialized society where technological disruption has repeatedly brought layoffs, weakened communities and greater inequality, productivity itself acquires a threatening meaning.

The politics contained within the technology

The current American government's own language reinforces these expectations. Its AI Action Plan describes the national objective as "winning the AI race," maintaining American dominance, accelerating construction of data centers, removing "onerous" regulations and asking private industry which additional rules should be eliminated. The operative priorities are corporate expansion, infrastructure construction, exports and geopolitical competition. There is little comparable institutional machinery ensuring that displaced workers will retain their income, that productivity gains will be broadly distributed, that communities will control the environmental consequences or that the informational commons will remain usable.

The question "Are you optimistic or pessimistic about AI?" consequently conceals most of what matters. AI has no single social meaning independent of the institutions that own, deploy and regulate it. A technology used to reduce working hours while maintaining wages would be experienced differently from one used to eliminate workers. A public educational resource would create different capacities from an automated system for reducing teaching costs. A publicly accountable information infrastructure would have different consequences from a proprietary engine inserted into an advertising economy.

American anxiety about AI reveals an expectation that privately controlled innovation will concentrate benefits and distribute harms. Chinese confidence reflects a stronger expectation that technological change will be coordinated and incorporated into a broader project of social development. Neither attitude arises from the technology alone. Each expresses a historically grounded judgment about government, business and the probable distribution of technological power.

Americans worry about AI because they understand that no reliable institution stands between them and its possible consequences. The deepest question raised by artificial intelligence is therefore also the oldest political question: who controls society's productive powers, for whose benefit, and who will be responsible for the people harmed along the way?