As voices questioning artificial intelligence grow louder in the United States, the deeper concerns behind them touch political nerves. AI could well become a new dividing line in the American political spectrum, much like gun control and abortion.
Since the beginning of the year, voices questioning artificial intelligence have grown louder in the United States. Major polls show that Americans, particularly young people ages 18 to 29, have a more negative view of AI than they did a year ago. Overall, public acceptance and trust in AI also appear to be lower in the United States than in China. Some have compared the trend with the backlash against industrialization, warning that a new generation of Luddites could emerge to resist AI and its disruptions. The issue has already become part of political life in many parts of the United States, including New York and Texas, where AI a subject that candidates and voters cannot afford to ignore ahead of the midterm elections.
The U.S. is at the forefront of the artificial intelligence revolution. One might therefore expect Americans to be more receptive to AI than people in other countries. Why, then, are voices questioning AI also so prominent in the country? What lies beneath this apparent paradox?
When it comes to concerns about artificial intelligence, environmental impact and employment are often the first things that come to mind. They are also the issues discussed most widely and the ones most likely to resonate with the public. But the debate in the United States has clearly moved beyond those two sectors. It is now closely intertwined in the political matrix that matters most to Americans.
First, transparency. For many Americans, the way artificial intelligence models data and powers applications is essentially opaque and difficult for ordinary people to understand. As embodied AI becomes more widespread, people may likewise struggle to understand why robots take certain actions or make particular decisions. The companies that develop the models, applications and robots, however, know how their systems work—and many of those companies are controlled by large corporate interests and wealthy individuals. With the Trump administration now maintaining close ties with the AI industry, some Americans are concerned that these interest groups could gain greater influence in Washington and ultimately come to exert control over U.S. politics.
Second, wealth distribution. Artificial intelligence has created a new breed of extraordinarily wealthy individuals: Where billionaires were once the norm, there are now trillionaires. The personal fortunes of figures such as Elon Musk and Jensen Huang exceed the annual GDP of many countries, yet the distribution of wealth in the United States has lagged far behind this transformation, leaving many young Americans disillusioned. Many also argue that the foundational models and AI-powered applications developed by technology companies are voraciously mining and extracting the vast store of knowledge created throughout human history, while paying virtually nothing for it and showing little regard for copyright.
Third, the right of choice. In most cases, artificial intelligence can improve efficiency and reduce costs, making it the latest avenue for businesses to boost profit margins. Companies and organizations across the United States are accelerating the deployment of AI applications and requiring employees to learn how to use AI tools. But not everyone is willing to embrace them. Some complain that when people used Google Search, they could choose from a wide range of links. Now, AI can process search results before users even see them and provide a direct answer, effectively taking the choice away from users.
Fourth, economic system. In an interview in March, Jensen Huang, CEO of Nvidia, said he would be frustrated if an engineer earning $500,000 a year consumed only $5,000 worth of tokens over the course of a year. In his view, a reasonable level of consumption would be half the engineer’s salary, or $250,000, which would be necessary to generate a substantial increase in productivity.
Some Americans see this as a possible indication of how economic relations may work in the future: Tokens could become a new kind of consumable input. To create value, one would have to consume tokens. Conversely, if you could not consume tokens on a large scale, you would have little value to the economic system, much as someone who cannot afford to buy a car or a home would. But not everyone will have the ability, opportunity or willingness to consume tokens on a large scale. As a result, some people could be left behind. This may be more worrying than the temporary unemployment caused by AI, because it goes to the underlying logic of the economic system itself.
In fact, it remains unsettled whether the concerns Americans have raised—including those over the environment and employment—will ultimately prove to be real problems. The Industrial Revolution displaced workshops and handicraft workers, leaving large numbers of textile workers unemployed, but it went on to create an industrial working class on a scale unprecedented in human history. Factories once polluted the air and rivers on a massive scale; today, however, most manufacturing plants are clean, bright and, in some cases, virtually spotless.
Anyone familiar with the history of American politics knows that whether a problem is genuinely a problem is sometimes beside the point. What matters is the political contest involving different forces over such issues and the distribution of political interests behind it. In this sense, AI could well become, in the United States, a new dividing line in the political spectrum, much like gun control, abortion and immigration—an issue that politicians will debate for years to come.
