The MIT AgeLab hosted a roundtable event to discuss the results of its AI and Longevity study, which examined the implications of artificial intelligence for older adults and individuals across the lifespan. The event featured participants from the AgeLab’s November 2020 symposium on AI, “Charting the Longevity Economy’s Endless Frontier: How AI Will Enable Us to Live Longer, Better.”
Bank of America’s Surya Kolluri and AgeLab Research Scientist Chaiwoo Lee started off the discussion by providing an overview of the AI and Longevity study, which surveyed consumers and experts on their attitudes toward AI and beliefs about its impacts in five key longevity-related domains. Many areas of peoples’ lives may be affected by implementation of AI; a literature review solidified the researchers’ attention to the following areas: finance and retirement planning; healthcare and caregiving; social interaction and communication; infrastructure and community; and workplace, career and benefits.
Dr. Lee broadly described some key findings from the study: Both experts and consumers generally thought that AI would be broadly beneficial, but there were some stark differences in the responses between the consumer and expert cohorts. The experts were more confident in AI’s ability to perform certain tasks than the consumers; and the experts predicted a higher rate of consumer adoption of AI than the consumers themselves.
Within the consumer sample, differences in trust and confidence in AI emerged by age, comfort with technology, gender, education, income. Higher-income individuals, individuals with more education, men, younger people, and people who were comfortable with technology tended to be more confident in AI’s performance in different applications. One demographic exception arose among members of Gen Z, who showed levels of confidence closer to the oldest generations than to Gen X and Millennials. AgeLab Research Associate Luke Yoquinto suggested that Gen Z’s attitudes may be attributable to the young cohort’s greater native understanding of technology and awareness of its unproductive or even exploitative tendencies.
For the roundtable’s second segment, AgeLab Research Scientist Shabnam FakhrHosseni led a discussion on the implications of the gap between consumers and experts in their estimation of consumer readiness to adopt AI. Steve Ewell of the Consumer Technical Association suggested that the difference in confidence may be related to differences in knowledge— experts have more specialized understanding of the technology than consumers, and consumers receive much of their information about AI applications from the news media, which tends to focus on technological failures such as autonomous vehicle crashes. Focusing as well on knowledge and understanding Kevin Crain of Bank of America predicted that increasing consumers’ exposure to AI should gradually increase their acceptance of the technology.
AgeLab Technical Associate John Rudnik moderated a dialogue on the issues of bias, discrimination, and inequality in the application of AI. He distinguished between bias—which refers to “systemic error” in a decision-making process or algorithm—and discrimination, which is defined as performing a differentiation between groups. Rudnik also highlighted that the experts who were surveyed in the AI and Longevity study generally believed that groups with more access, power, and status are more likely to benefit from applications of AI.
On the topic of discrimination, Surya Kolluri noted legal restrictions for corporations on beneficent discrimination toward marginalized or underserved groups, which may lead to greater inequalities in the distribution of services and interventions. AgeLab Research Scientist Julie Miller suggested that AI’s use and benefits may be more concentrated among groups with more access in its early phases, but may become more equitable as it develops and is disseminated more widely.
The roundtable’s concluding segment was a lightning round for participants to explore implications of AI in business, research, and policy. AgeLab Director Joseph Coughlin suggested the need for a special agency to guide regulation and policy for AI due to the technology’s complexity. Michael Finke of the American College of Financial Services argued for the need for robust regulations in order to prevent bad actors deploying AI for exploitative and manipulative practices to extract wealth from individuals.
Chaiwoo Lee concluded the discussion by advocating for government, business, and other stakeholders to engage in a collective educational effort to ensure that the concept of AI, as well as a proper recognition of when AI is and is not present in a technological application, is grasped by those who will use and be affected by it. Such an education will be necessary for a willing adoption of AI by individuals, as well as for meaningful discussion of the implications of the technology for people of all ages.
The roundtable also showcased a report on the results of the AI and Longevity study, which breaks down findings across the five domains and by the expert and consumer samples. The report is soon to be released.