With generative AI tools readily available to students, student assessment was a recurrent topic of discussion at a bilingual, two-day symposium hosted by McGill last month in partnership with Google and the Pôle interordres de Montréal.
EduGenAI: Shaping Tomorrow’s Learning with Generative AI attracted more than 100 higher-education and industry leaders for wide-ranging presentations, panel discussions and workshops exploring the evolving role of generative AI in the classroom.
“The generative AI tools that have proliferated over the past 18 months have raised complex and urgent questions for higher education,” said Eric D. Kolaczyk, Director of McGill’s Computational and Data Systems Initiative (CDSI) and a professor in the Department of Mathematics and Statistics.
“This symposium fostered engaging discussions among stakeholders across disciplines and institutions, underscoring how salient an issue this is in our community, and emphasizing the need for ongoing dialogue.”
Developing critical thinking skills
Professor Jasmin Chahal of the Department of Microbiology and Immunology at McGill shared her experience incorporating the use of generative AI tools into her course assignments.
“I don’t want students to take things at face value, I want them to think about what they’re reading and develop their critical thinking skills,” Chahal said.
In one of her courses, designed for those in the first year of their major, Chahal asked her students to use a prompt she provided to generate a paragraph of text in ChatGPT and provide three in-text citations and references.
“They had to determine whether the references were reliable and why or why not,” she said. “It’s important that students know how to use these tools and understand their benefits as well as their weaknesses.”
Robert Stephens, Professor of Humanities and Philosophy at Dawson College agreed that “it’s worthwhile to bring these tools into the classroom and show students where they are useful and where they are not,” but highlighted the implications of the technology for student assessments.
“That also means it will be important to be more creative with assessments to ensure students still have to use their own non-AI-assisted powers themselves at times. In some cases, this may involve a total reworking of assessments and the way students are evaluated,” he said.
Potential to enrich lives, promote lifelong learning
For Samira Abbasgholizadeh Rahimi, co-director of the McGill Collaborative for AI & Society (McCAIS), Canada Research Chair in Advanced Digital Primary Health Care and Assistant Professor at McGill and Mila-Quebec AI Institute, it’s important not only to ensure that AI is taught to the next generation of students, but also to better educate the public:
“AI should also be promoted to citizens in broader society, so they understand what AI is and what it isn’t and can therefore form better decisions,” she said, noting that “AI has the potential to enrich and promote education and lifelong learning.”
Advocating for what he called a “middle path,” neither overly sanguine nor overly pessimistic, Professor Adam Dubé of the Faculty of Education said that technology, including AI, doesn’t inherently alter learning.
“Generative AI is not a game changer,” he said. “It’s just another piece on the board, and we need to make the right moves.”
Continuing the dialogue
The symposium, an initiative of McGill’s School of Continuing Studies, Computational and Data Systems Initiative (CDSI), and Teaching and Learning Services, included a workshop at the headquarters of Google Montréal. Vincent Lecuyer, a Pre-Sales Engineering Specialist at Google Cloud and a member of the Google for Education program in Quebec, and Rihana Msadek, an AI Customer Engineer at Google, led discussions around generative AI and other key trends driving the future of education. The keynote speech, delivered by leading AI researcher Professor Doina Precup of the McGill School of Computer Science, delved into the development and training of AI models.
The symposium was made possible through the support of the Pôle interordres de Montréal (PIM), explained Nabil Beitinjaneh, program coordinator for Data Science and Data Analytics programs at the School of Continuing Studies and one of the organizers of the event.
“We plan to continue the dialogue with our participating Montréal CEGEPs and universities through the PIM,” Beitinjaneh said. “Generative AI is here, and we will work together to understand its limitations and build on its advantages, incorporating it into our educational toolkit to upskill and reskill both students and educators.”