How Could Artificial Intelligence Shape the Future of Higher Education? #DLNchat

Do anthropomorphized forms of artificial intelligence serve students better than hidden bots? Will students receive AI-manifested messages through wearable devices in the future? How do we create frameworks for the ethical creation of artificial intelligence?

There were a lot of big questions and a lot of big ideas at #DLNchat on Tuesday, March 27 when special guest Bryan Fendley, Director of Academic Computing at University of Arkansas, Monticello, guided our discussion: How Could Artificial Intelligence Shape the Future of Higher Education?

The chat started with a healthy dose of skepticism about the potential misuses of artificial intelligence. #DLNchat-ters voiced the need to create algorithms and other AI systems within carefully construed ethical frameworks. Fendley, our special guest, suggested that such frameworks may need to be government regulated. How do we keep algorithms in check? Or, as Aneesa Davenport, Social Media & Analytics Manager for EdSurge, suggested, perhaps give them a periodic “check-up?” Cali Morrison believes we need to start with a diverse team who can build the code and then review it for biases. “And intentional investigation,” she adds. “Not just assuming that because you have a diverse team that all your problems will be solved.” In the case of education, Tom Pantazes thinks we need to strive for including students in algorithm development. Students should also be involved in decision-making efforts around what data is used and how that data is used, said Megan Raymond.

There is a lot of data to sort, and Fendley, our special guest reminded us there’s only more to come. Kent Darr suggested that “Parsing large data sets could help the instructor and instructional designer make quicker, more informed decisions about course content, tasks, and assessments,” work that “may otherwise prove to be time-consuming for the instructor/ researcher.” That is particularly true when institutions want to include data from other fields and organizations. Joshua K. Farrar imagined for us “a world where aggregated statistics across the industry of education can help direct conversations on local, state, and federal education budget priorities.”

Register now for the EdSurge webinar, How Analytics Can Support Student Success in Higher Ed, April 25, 2018 at 12 pm PT | 3 pm ET. Sponsored by Salesforce.org.

In addition to behind-the-scenes algorithms, the #DLNchat community also discussed artificial intelligence that interacts with students, faculty and other individuals. How should these bots communicate? What should they sound like, and look like? Dr. Ryan Straight pointed out that “our perception of that AI (the extent to which it's anthropomorphized) can seriously alter how we use and react to it (and, theoretically, how it reacts to us)... Sometimes an artificial persona (call it AI, virtual assistant, robot like ASIMO, whatever) can, in its failed attempt to be human, be truly tragic. The result can be jarring.” As an alternative, Sean proposed these interactions occur in a virtual reality space, and Fendley positied AI-manifested messages buzzing or activating a wearable device. Most #DLNchat-ters agreed, however, it seems best if we leave the human interaction to other humans, no matter where the alert originally comes from. As Patrice Torcivia tweeted, it’s “critical to combine this data with human interaction. Students need someone to help them make meaning of the data. We don't want a student to see the are on track to get a C and quit, rather than use the information about how to change track.” Kent Darr put it this way: “AI can quickly interpret qualitative data to show us large trends in classes, colleges, or entire student bodies. Humans can more easily assign context and meaning.”

Whether through front-end interactions, or back-end operations, it seems that we will continue to see the use of artificial intelligence grow on campus. The #DLNchat community agreed that higher ed institutions have a responsibility to prepare students who will create, communicate and cohabitate with different forms of artificial intelligence. Part of this education can be crafted into the curriculum. Lindsey Downes reminded us to, “Bring questions and encourage discussions about emerging technologies in all types of classes (not just computer science). Ask students to bring examples from the real world into the classroom.” Another part of this education is to continue to support student growth in the arts, in creativity, in philosophy. As Dr. Straight put it, to “instill a deep love for and definition of "humanity" and the difference between people and constructs.”

What do you think? How can institutions help students prepare for lives and careers in a world infused with artificial intelligence? Tweet us your thoughts and don’t forget the hashtag #DLNchat. For other topics, you can also check out our index of past #DLNchats.

Join the Digital Learning Network to stay up to date on all events and the latest news for highered digital learning leaders! RSVP for our next #DLNchat: How Have MOOCs Impacted Approaches to Student Learning?We’ll be joined by special guest Dhawal Shah, founder of Class Central, on Tuesday, April 10 at 1pm PT/ 4pm ET.#DLNchat is co-hosted by theOnline Learning Consortium,WCET andTyton Partners.