IWB Blog

A collection of observations, news and resources on the changing nature of innovation, technology, leadership, and other subjects.

ABOUT

Subscribe to this blog via email

Over the past two years, public discussion of artificial intelligence has focused largely on the technology’s remarkable advances. New foundation models, increasingly capable AI agents, and massive investments in data centers have dominated the headlines. But as AI moves from research labs into workplaces and everyday life, a different issue is emerging: a growing public backlash.

Concerns about jobs, energy consumption, privacy, misinformation, and the growing concentration of power among a handful of technology companies are making AI an increasingly contentious political issue. The question is no longer simply how fast AI can advance, but whether societies will continue to support its deployment.

Writer Cory Doctorow offers an interesting historical perspective in this Medium article published earlier this year, “The AI Bubble Isn’t Like the Internet Bubble.”

“The arrival of internet-based tools — especially ones that ran in browsers — represented a major challenge to IT departments, who had been long accustomed to dictating terms to their users,” Doctorow wrote. Before the internet became widespread in the early 1990s, IT departments exercised tight control over the software employees could use. But that quickly changed once the web reached the corporate desktop. Employees increasingly ignored centralized restrictions and adopted browser-based tools because they helped them do their jobs more effectively.

IT managers understandably worried about security, compliance, and malware. But they were never forced to persuade employees to use the web. Quite the opposite. Workers enthusiastically incorporated internet technologies into their daily work because they immediately recognized their value.

“Today,” Doctorow observes, “you can’t turn around without reading a story about bosses who are threatening to fire workers if they don’t increase their AI usage. Virtually every major company now has a program to force workers into using AI.”

That difference highlights an important distinction between the internet revolution and today’s AI revolution. Historically, workers have eagerly embraced technologies that made them more productive. AI has proven far more complicated.

The June 27 issue of The Economist devoted several articles to what it called the growing AI backlash. “Advances in artificial intelligence have long terrified techies,” noted its lead article. “Lately, voters are feeling the angst, too. AI is unpopular in the West and climbing up the political agenda.”

The backlash is becoming increasingly visible. Protests against proposed data centers have delayed nearly $100 billion worth of projects in the United States. Around 40 percent of American voters tell pollsters they would like AI banned from most industries. Political battles over AI regulation have become increasingly polarized.

“The backlash is only just getting started,” The Economist argued, “because the technology is only just getting started, too.”

The magazine warned that this backlash carries significant risks. AI has the potential to increase productivity, accelerate scientific discovery, improve health care, enhance education, and contribute to economic growth. If governments overreact by starving AI of computing resources or regulating it into ineffectiveness, societies could lose many of these benefits.

The backlash is particularly significant because AI has also become a central element of strategic competition between the United States and China.

As a recent Wall Street Journal article observed, the United States continues investing enormous sums in pursuit of artificial general intelligence (AGI), believing that achieving human-level AI could transform scientific research, military capabilities, medicine, and economic productivity. China appears to be pursuing a somewhat different strategy. Rather than emphasizing AGI, Chinese leader Xi Jinping has encouraged the country’s technology industry to focus on practical applications that improve productivity and can be deployed broadly throughout the economy.

These contrasting approaches represent two different visions for AI’s future. Silicon Valley continues betting heavily on achieving increasingly powerful frontier models, while China appears more focused on extracting economic value from today’s technologies.

Given the uncertainty surrounding how soon — if ever — AGI will be achieved, the practical approach may prove increasingly important. If public resistance slows AI deployment in the United States while China steadily expands practical AI applications across its economy, America could lose some of the competitive advantage that its technological leadership has created.

Doctorow also points to another important difference between the internet and AI. “In the early 1990s,” he notes, “workers were enthusiastic about tools that let them do a better job.” Today’s younger workers are considerably more ambivalent.

A recent Gallup survey found that although more than half of Generation Z regularly uses generative AI, their attitudes toward the technology have become noticeably less positive. The percentage of young people who say they feel hopeful about AI declined sharply over the past year, while feelings of anger and concern increased.

Doctorow argues that the economics of AI also differ fundamentally from those of the early web. Every additional internet user helped make the web more valuable and more profitable. AI, by contrast, remains extraordinarily expensive to build and operate. Each new generation of frontier models requires even larger investments in computing infrastructure, making the economics of AI considerably more challenging than those of earlier digital technologies.

Political attitudes reflect these broader concerns.

Another article in the same issue of The Economist, “Democrats and Republicans Agree: AI Is Scary,” notes that anxiety about AI is one of the few issues that unites both political parties.

Three-quarters of Americans believe AI should be more heavily regulated. Nearly two-thirds think the technology is advancing too quickly. Around 65 percent believe AI will reduce the number of available jobs, while more than 70 percent worry about its effects on mental health and the growing concentration of power among large technology companies. Perhaps most striking, Americans appear more pessimistic about AI than citizens of almost any other advanced country. While most acknowledge AI’s enormous potential, many expect its overall effects to be more negative than positive.

Can governments do anything to address these concerns?

The Economist argues that broad declarations about a new social contract are unlikely to be very helpful while the technology continues evolving so rapidly. Instead, it recommends four practical priorities.

First, spread AI’s benefits as broadly as possible. Communities that host AI infrastructure should share directly in its economic gains, and workers displaced by technological change should receive meaningful support through policies such as wage insurance, retraining, and other transition assistance.

Second, regulate aggressively where genuine risks exist. AI-enabled cyberattacks, biological threats, and other serious dangers deserve careful oversight. Effective regulation of these risks could strengthen public confidence while avoiding unnecessarily broad restrictions on AI innovation.

Third, improve measurement. Much of today’s public debate relies on anecdote rather than evidence. Better data on employment, productivity, energy use, and other impacts would lead to better public policy and help distinguish real problems from exaggerated fears.

Fourth, governments should use AI to improve their own services. AI could simplify tax filing, modernize health-care systems, improve public education, and increase government transparency. Citizens are more likely to trust AI if they experience tangible benefits in their own lives.

“People are less likely to oppose a technology if it is behind their grandmother’s cancer treatment or helping their child’s education. And they are more likely to trust that the state can oversee it if they believe that government works,” notes The Economist, — an observation that captures what may be the central challenge facing AI over the next decade.

The future of AI will depend not only on continued advances in models and computing power, but also on maintaining public trust. Like previous transformative technologies, from electricity to the internet, AI will ultimately be judged less by its technical achievements than by whether people believe its benefits are broadly shared, its risks responsibly managed, and its deployment aligned with the public interest.

The challenge before us, therefore, is not simply to build more powerful AI systems. It is to ensure that society has confidence in how they are being developed, governed, and used. If we succeed, the current backlash is likely to prove temporary. If we fail, public resistance could become one of the biggest obstacles to realizing AI’s extraordinary potential.

AI disclosure: I wrote the initial draft of this blog and used ChatGPT for editorial assistance. I remain responsible for the final content.

Posted in , , , , ,

Leave a Reply