Related Articles
Over the past year or so, you’ve probably had conversations with friends, family, and coworkers about the rise of generative AI capable of making convincing text and imagery—but perhaps also about the hype and fear swirling around the technology. A poll out this week finds that worry over harmful effects of AI is outpacing the wow of helpful AI.
A majority of Americans say their concern about artificial intelligence in daily life outweighs their excitement about it, according to a Pew Research Center survey of more than 11,000 US adults. The results come at a time when a growing number of people are paying attention to news about AI in their daily lives. Pew has run this survey twice before and reports that the number of people more concerned than excited about AI jumped from 37 percent in 2021 to 52 percent this month.
The balance of concern and excitement people reported varied between different use cases for AI.
When asked how they felt about the police using AI for public safety, roughly half of respondents said they weren’t sure, with the rest evenly split between saying the technology would help or hurt. Many more people believed that AI would help doctors to provide quality care to patients, but it’s likely people would have different feelings about some specific applications of medical AI. Many would probably feel uncomfortable with a triaging algorithm making life-or-death decisions about who receives what treatment.
Pew found the largest swing towards concern about hurtful AI when asking what impact the technology would have on the ability to keep their information private. That fits with how US activists, policy experts, and researchers who want to protect civil rights and hold businesses and governments using AI accountable often call for comprehensive data privacy protections. So far, Congress is yet to pass a privacy and data protection law.
One impact of AI on daily life the survey didn’t ask about is the technology’s potential to help or hurt discrimination. Years of evidence show that AI systems can reinforce or amplify racism, sexism, or discrimination against the poor and people who identify as queer. But AI can also detect bias and prevent discrimination. Sennay Ghebreab, director of an AI lab at the University of Amsterdam, told me last year, “I’ve been working on this topic for a decade, and although it can be harmful to people, AI presents an opportunity to uncover hidden biases in society.”
Pew's findings raise the question of how people not working on AI themselves can retain any feeling of autonomy as the technology becomes more visible and powerful. I was struck by remarks earlier this month by former US secretary of state Condoleeza Rice, who at a recent Stanford event on AI described meeting a group of students visiting from Latin America who told her that AI feels like something that’s happening to them rather than technology they’re playing a role in shaping.
That feeling, Rice said, may be more pronounced for people outside China, Europe, and the US. But plenty of people in those countries feel they don’t have enough agency in their own lives. And even people active in the fight against AI that enables human rights abuses can feel helpless or lose hope.
Most PopularGearThe Top New Features Coming to Apple’s iOS 18 and iPadOS 18By Julian ChokkattuCultureConfessions of a Hinge Power UserBy Jason ParhamSecurityWhat You Need to Know About Grok AI and Your PrivacyBy Kate O'FlahertyGearHow Do You Solve a Problem Like Polestar?By Carlton Reid
Although such feelings are understandable, there are ways that citizens can gain more control over the impact of AI on their lives. Here are a few things to keep in mind.
Question the motives of people who talk about superintelligence and artificial general intelligence. The fear about the future these terms can stoke helps the economic interest of a small group of companies that possess the unique combination of compute power, AI talent, and money necessary to build the most powerful AI systems. Debate about the hypothetical power of future AI technology only centers more attention on the dominant players of today.How AI influences society should be a society-wide conversation. It’s fine that major AI company executives visited the White House in May and then again in July and agreed to voluntary measures, but relying on them for answers is, as I said in May, “like asking a group of arsonists how to put out a fire.”We can take lessons from history, which show that predictions about powerful AI can be wrong, and regulation can work. Despite widespread claims that the advent of more powerful machine learning in the 2010s would kill jobs, mass unemployment did not occur. And while tech companies warn that regulation can stifle innovation, a study out this week found that US antitrust enforcement from 1971 to 2018 increased jobs and business formation.Understanding the role of AI in daily life is, as Microsoft CTO Kevin Scott said in 2019, part of citizenship in the 21st century—so educate yourself on what accountability looks like to you and tell your elected officials. Some commonly suggested solutions I’ve heard while reporting stories about AI and accountability over the better part of the last decade: Require businesses and governments to disclose when AI influences a decision like whether you can rent an apartment or get labeled a fraud, to provide the option to opt out from automated decisionmaking; and to perform audits or testing on AI systems they use.
