Women in the United States are more skeptical than men about some applications of artificial intelligence (AI), especially the possible widespread use of driverless passenger cars, according to a new analysis of Pew Research Center survey data collected in November 2021.
The analysis also finds gender differences in views on the overall impact technology has on society and some security issues related to AI applications, as well as the importance of including different groups in the AI design process.
Pew Research Center conducted this study to understand gender differences in Americans’ beliefs about artificial intelligence and people-improvement technologies. For this analysis, we surveyed 10,260 U.S. adults from November 1-7, 2021.
Everyone who participated in the survey is a member of the Center’s American Trends Panel (ATP), an online survey panel that is recruited through random sampling of residential addresses nationwide. In this way, almost all American adults have a chance at selection. The survey was weighted to be representative of the US adult population based on gender, race, ethnicity, partisan belief, education, and other categories. Read more about the ATP method.
These are the questions used for this analysis, along with the answers and methodology.
In general, women in the US fewer more likely than men to say that technology has had a mostly positive effect on society (42% vs. 54%) and Lake likely to say that technology has had as positive as negative effects (45% vs. 37%). In addition, women are less likely than men to say they are more excited than concerned about the increased use of AI computer programs in everyday life (13% vs. 22%). Gender continues to be a factor in perceptions of the impact of AI and technology when other variables are taken into account, such as political bias, education and respondents’ race and ethnicity.
In addition, in a range of potential AI applications asked in the November survey, women are consistently more likely than men to be concerned about computer programs performing these tasks. For example, 43% of women say they would be very or somewhat concerned if AI programs could diagnose medical problems, while 27% of men say the same. Gender gaps also feature in the amount of concern Americans express that AI programs can perform repetitive tasks in the workplace, make important life decisions for people, and know people’s thoughts and behavior.
Women feel more negatively than men about driverless passenger cars
In addition to gender differences on AI in general, women and men express different views on autonomous cars in particular.
Men are more likely than women to respond positively to various questions about these vehicles. About four in ten men (37%) say driverless cars are a good idea for society, while 17% of women say the same. Women, in turn, are slightly more likely than men to say that they are not sure whether the widespread use of self-driving vehicles is a good or a bad idea (32% versus 25%).
Similarly, 46% of men say they would definitely or probably personally drive a driverless passenger car if given the chance, compared to 27% of women. Most women (72%) report that they definitely or probably would not want to do this. Also, a majority of women (54%) say they would not feel comfortable sharing the road with a self-driving passenger car if its use became widespread. Only 35% of men say the same thing.
These differences may be related to women’s greater doubts about the safety of autonomous vehicles. When asked about the effect the widespread use of these cars would have on the number of deaths or injuries in road accidents, about half of men (49%) say it would reduce the number of deaths or injuries, compared to three in ten women. who say the same. By a margin of 33% to 20%, women are more likely than men to think the number of accidents would be increase.
Women’s concerns about the safety of autonomous passenger cars extend to other applications of driverless systems as well: 51% of women say they are against the use of this technology on public transport buses, compared to 35% of men. Women are also about 10 percentage points more likely to oppose the use of driverless technology in 18-wheel trucks, vans and taxis, and ride-sharing vehicles.
While these figures cannot be directly compared to previous studies due to changes in the questioning, the current findings are consistent with a 2017 study by the Center which found that women are less likely than men to say they would drive in a driverless vehicle. want to drive. Previous research from the Center has also shown that women are sometimes more pessimistic than men about technological change in general.
Women and men are different from other applications of artificial intelligence
In addition to exploring public opinion on the potential use of self-driving passenger cars, the Center’s November research also covered two other specific AI applications: the use of facial recognition technology by police and the use of algorithms by social media companies to target false information on their behalf. to find sites.
There are differences between women and men on some questions related to these applications. While a majority of both men and women have heard of any use of AI, a clear pattern is that men are more likely than women to say they have heard or read at least something about each of the three technologies: driverless cars (93% vs. 83%), use of facial recognition by police (86% vs. 74%) and social media algorithms to find false information (81% vs. 70%).
Women are also more likely than men to say that they are not sure whether certain AI applications are a good or bad idea for society. About 34% of women aren’t sure whether social media algorithms to find false information are a good or bad idea, compared to 26% of men. When it comes to police use of facial recognition, 31% of women aren’t sure whether it’s a good or bad idea, compared to 22% of men.
Gender differences in the design of AI technologies
Women are more likely to be in favor of including a wider variety of groups in AI design. For example, two-thirds of women (67%) say it’s extremely or very important for social media companies to involve people of different genders in designing social media algorithms to find false information, compared to 58% of women. men. Women are also more likely to say it is important for different racial and ethnic groups to be involved in the same AI design process (71% vs. 63%).
In addition, women doubt more than men that it is possible to design AI computer programs that can consistently make fair decisions in complex situations. Only about two in ten women (22%) think it is possible to design AI programs that can consistently make fair decisions, while a higher proportion of men (38%) say the same. A multiple of women (46%) say they are not sure whether this is possible, compared to 35% of men.
Risa Gelles Watnick is a research assistant focused on Internet and technology research at Pew Research Center.