A comparison of five recent, high-quality US surveys finds that negative views of artificial intelligence outweigh positive ones, even though question wording affects the size of the gap. When respondents can choose a middle option, many do: in Pew’s 2026 survey, 37% were neither more concerned nor more excited. Still, 52% said they were more concerned than excited, compared with just 9% who were more excited. That concern has grown since 2021, when Pew found roughly twice as many people concerned as excited; by 2026, the ratio exceeded five to one. Across other surveys, Americans were also more likely to say AI would do more harm than good, and strong opinions tended to be especially negative.
Concern spans political parties: 56% of Democrats and 49% of Republicans said they were more concerned than excited, while only about one in ten in either party were more excited. Worries include severe risks: 73% of respondents in a Quinnipiac poll said they were at least somewhat concerned that future AI could threaten human survival. The main exception came when people were asked about AI’s effect on them personally: Ipsos found views nearly evenly split, with slightly more calling it good than bad. Even in that case, intense personal judgments were more often negative. The surveys used probability-based samples and population weighting, lending greater credibility than opt-in online polls. The findings suggest that earning public trust will require taking widespread concerns seriously.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.