Experts Debate Merits of AI in Education

• By Dian Schaffhauser
• 12/20/18

Will artificial intelligence make most people better off over the next decade, or will it redefine what free will means or what a human being is? A new report by the Pew Research Center has weighed in on the topic by conferring with some 979 experts, who have, in summary, predicted that networked AI "will amplify human effectiveness but also threaten human autonomy, agency and capabilities."

When the experts were asked whether AI and related technology will by the year 2030 enhance human capacities or allow them to deteriorate, the majority (63 percent) said most people will be better off.

The opportunities cited in "Artificial Intelligence and the Future of Humans" were far-ranging: "smart systems" built into cities, vehicles and buildings "will save time, money and lives," and AI-driven uses in healthcare will advance diagnosis and treatment of patients or help people who need daily aid to "live full and healthier lives." At the same time, as decision-making is turned over to "black box tools," people will sacrifice their "independence, privacy and power over choice" — an outcome that will "deepen" as automation becomes "more prevalent and complex." Likewise, job loss is inevitable, "widening economic divides and social upheavals."

These same experts also weighed in on the expected changes in formal and informal education systems. Many mentioned seeing "more options for affordable adaptive and individualized learning solutions," such as the use of AI assistants to enhance learning activities and their effectiveness.

Among the respondents was John Laird, a professor of computer science and engineering at the University of Michigan, who said he anticipates "improvements in customized/individualized education and training of humans, and conversely, the customization of AI systems by everyday users."

Lou Gross concurred. This professor of mathematical ecology and expert in grid computing, spatial optimization and modeling of ecological systems at the University of Tennessee, Knoxville, reported that he expects AI to deliver "better adaptive learning" and help teachers personalize education based on students' individual progress.

Guy Levi, chief innovation officer for the Center for Educational Technology, based in Israel, noted that he sees similar benefits, considering the potential for students to use personal AI assistants "a game changer." Among the advantages of their use: They'll "be able to manage diverse methods of learning, such as productive failure, teach-back and other innovating pedagogies."

Kristin Jenkins, executive director of BioQUEST Curriculum Consortium, also fell into this camp: "AI systems are perfect for analyzing students’ progress, providing more practice where needed and moving on to new material when students are ready," she stated. "This allows time with instructors to focus on more complex learning, including 21st-century skills."

Barry Chudakov, founder and principal of Sertain Research, which produces marketing data through emotion recognition, said he expects AI to "finally demolish" the memorization aspects of education. "Knowing is no longer retaining — machine intelligence does that; it is making significant connections. Connect and assimilate becomes the new learning model."

Others weren't as convinced of the efficacy of AI in education. According to Betsy Williams, a researcher in the Center for Digital Society and Data Studies at the University of Arizona, most AI used in education by the year 2030 "will be of middling quality," which for some could be the best alternative. After all, Williams noted, research in K-12 has found that the "typical computer-aided instruction yields better test scores than instruction by the worst teachers." However, she added, for the "rich and powerful," their children won't have AI used on them in school; they'll be "taught to use it."

That was an opinion echoed by Henning Schulzrinne, a professor at Columbia University and co-chair of the Internet Technical Committee of the IEEE Communications Society. "Human-mediated education will become a luxury good," he asserted. "Some high school- and college-level teaching will be conducted partially by video and AI-graded assignments, using similar platforms to the MOOC models today, with no human involvement, to deal with increasing costs for education."

Karen Oates, director of workforce development and financial stability for La Casa de Esperanza, an advocacy organization for providing community opportunities, especially among the Hispanic population, expressed concern about the impact of AI on "working poor and low-to-middle-income people." Those are the populations, she explained, that lose their jobs when "robots and self-operating forklifts" are put in use. Without economic incentives from the government, she doesn't expect most employers to train employees to learn how to program or maintain those machines; and most of the workers on that lower run "won't have the confidence to return to school to develop new knowledge [or] skills" because they've been "unsuccessful in the past" As a result, she concluded, "as the use of AI increases, low-wage workers will lose the small niche they hold in our economy."

David Zubrow, associate director of empirical research at Carnegie Mellon University's Software Engineering Institute, suggested that advances in AI could make the "world better for all," by helping to deliver education to "remote and underserved areas." Even there, however, one fear he expressed is that control will be "consolidated in the hands of the few that seek to exploit people nature and technology for their own gain."

The complete report is openly available on the Pew Research Center website.