Optimism that AI will make human work easier is lowest in developed countries, particularly in the US, Canada, the UK, France and Australia.
This is one of several insights from Peter John Lowen, director of the arts and sciences at Harold Tannerdeen, who shared takeout in “What do Americans think about AI” from a decade-long AI study on July 9th. Lectures at Warrenhall were the first in a series of summer events in Cornell University Continuing Education.
Credit: Lindsay France/Cornell University
Peter John Lowen
“This pattern repeats many times. In some developing countries there is more positiveness about the impact of technology and artificial intelligence on the future and life than in the rich democratic West,” Lowen told more than 60 participants.
He explained how he and his colleagues viewed this pattern in a survey of public attitudes about AI, including around 24,000 people, about 1,000 people from each of 21 countries in 12 different languages, representing the majority of the world's population.
In another study starting this spring, Lowen examined offshoring work and AI attitudes, focusing on changes in work and price. Research participants said they were most concerned about price changes and prefer AI over offshoring, but the favor is more pronounced among Democrats, suggesting new partisan disparities, Lowen said.
Loewen has spent a lot of time and effort working with the government to understand how AI can be used within public services. In a paper published last year in the Journal of Public Policy, he looked into why people may or may not want AI in government services.
“A series of reasons have something to do with fairness,” Lowen said. “We recognize that humans are actually very arbitrary. We have all kinds of bias. We have tiredness, we make bad decisions. Inconsistent, Cetera. And we can design automated decision systems that do not have those biases or conflicts.”
Another reason is that AI is more efficient. According to Lowen, it works faster, saves money and can detect fraud. Researchers found that more conservatives emphasize fairness over efficiency, but found that they are liberal to use automation and algorithms to make governments more equitable.
“Therefore, politicians are trying to build public consent to use artificial intelligence, so people have different reasons to support it,” Lowen said.
However, he added that governments who want to use AI face many obstacles, from innate resistance to changing to mistrust of AI that makes moral judgments.
In another study, Lowen said he and his colleagues found that the way people think about artificial intelligence in the United States and elsewhere maps to many of the ways they already think about politics and the economy.
“Those who think that artificial intelligence is likely to replace work are more likely to support policies that can be characterized as populists around immigration and labor,” Lowen said. In contrast, AI strengthens those who think that governments should respond and complement and strengthen their work by increasing social spending, training workers for new careers, and investing in education.
“Our main problem is not technical – they are sociology,” Lowen said. “And it gives you a sense of why artificial intelligence is difficult to use.”
The Cornell SCE Summer Events series is free and open to everyone, but includes Wednesday lectures by the renowned Cornellian and a Friday concert at the AG Quad.
Linda B. Glaser is the News and Media Relations Manager at the University of Arts and Sciences.
