People with expertise working with artificial intelligence tools are fairly confident that AI will have a positive effect in the next 20 years, but relatively few US adults feel the same way, according to the results of a Pew Research Center survey.
The surge of generative AI tools like OpenAI’s ChatGPT in the last few years has pushed the subject into the mainstream. Pew said the share of the public who say they’re more concerned has risen since a similar survey in 2021.
Among AI experts, 56% said they expected AI’s impact on the US to be positive in the next 20 years. In contrast, only 17% of the public had that perspective, with 35% feeling AI will have a negative effect. (Among AI experts, 15% expected a negative impact.)
A big reason for the gap is that ordinary Americans are far less optimistic about the effect that AI will have on important aspects of their lives, like the workplace, health care and education. While 73% of experts said they expected AI would improve how people do their jobs, only 23% of US adults surveyed felt the same, for example.
Not all artificial intelligence works like the generative AI models that have risen in prominence since 2022, but the popularity of ChatGPT and other large language models like Google’s Gemini has led to a race among tech companies and what can feel like a saturation campaign of AI features across everyday apps and devices. That’s garnered mixed reactions from consumers. A 2024 CNET survey found many smartphone users unenthusiastic about AI integrations like Apple Intelligence.
The Pew report released this week covers the results of a pair of surveys. US adults were polled as part of Pew’s American Trends Panel, a survey of more than 5,000 randomly selected adults conducted in August 2024. The experts’ opinions came from a panel of 1,013 people living in the US whose work or research relates to AI and who presented at or published at any of 21 different AI-focused conferences in 2023 and 2024.
«These surveys reveal both deep divides and common ground on AI,» the report said. «AI experts are far more positive than the public about AI’s potential, including on jobs. Yet both groups want more personal control of AI and worry about lax government oversight.»
Worries about AI in the workplace
Neither group was optimistic that AI would lead to more jobs in the US in the next 20 years. Among AI experts, 19% expected more jobs, while 39% expected fewer. The skew was much greater among the general public — 64% expect fewer jobs and only 5% expect more.
As to which jobs were most at risk, both groups agreed that cashiers and journalists were vulnerable. Neither group had large numbers expecting job losses for mental health therapists, despite advances in the use of AI chatbots for therapy.
A majority of AI experts, 61%, said they expected job losses for truck drivers, compared with just 33% for the general public. Driverless trucks have been a dream for the autonomous vehicle industry for far longer than ChatGPT has been a household word, although the survey may show that the idea hasn’t caught on yet with the general public. One AI expert quoted anonymously in the report particularly highlighted truck drivers, saying, «They’ll be gone in 10, 20 years probably.»
Even for jobs that continue to exist despite AI, the work done by humans might change. Nickle LaMoreaux, chief human resources officer at IBM, told an audience at South by Southwest last month that it will change the qualities companies look for in workers, forcing them to focus more on critical thinking and human qualities.
Despite the hype, AI tools have yet to make much of a dent in many workplaces. A separate Pew survey released in February found only a sixth of US workers were using AI in their jobs and a majority said they don’t use chatbots at all or use them rarely.
Americans are more concerned than excited about AI
The two groups were virtually mirror images on the question of whether people were more concerned or excited about the increased use of AI. Among the general public, 51% were more concerned, while 47% of AI experts said they were more excited.
Experts quoted in the report pointed to things like the automation of repetitive tasks and the potential improvements in medicine as reasons for hope. «Most excited about the positive impact that it could have in the health industry,» one expert said.
The concerns, however, are significant. Issues like inaccurate information, impersonation (or «deepfakes») and the misuse of personal information were cited by both the public and the experts. The public reported being far more concerned about job loss than experts.
