This post is part of Sci-Fi Scenarios, the foresight series on TeachOnline.ca in which leaders in education and technology respond to five “sci-fi–sounding but plausible” AI futures.

We asked contributors to review five AI-driven scenarios for higher education (2025–2035), pick the one they find most compelling and explain why, and then add one future scenario of their own.

Below is the response from Dr. Stephen Murgatroyd, one of the leaders we invited to comment.

From the list: Dr. Murgatroyd’s #1 pick

The most likely option from the list is the idea of the cognitive twin for every learner, although I would modify this thinking to suggest that it is a constant learning companion capable of social and emotional intelligence as well as cognitive (and metacognitive) support.

Cognitive twins — a persistent AI companion that learns and grows with its student partner — offer a transformative approach to education by providing hyper-personalized support tailored to each learner’s evolving needs.  Here’s why “Cognitive Twins for Every Learner” is likely to become reality by 2035. 

High likelihood and strategic value

The core technologies (AI tutors, personal agents, learning analytics, mental health supports) are already available, making widespread adoption of cognitive twins highly likely. The strategic value lies in boosting student retention, focus, and motivation — key drivers of academic success —  while freeing up faculty to focus on higher-level teaching and mentorship.

Personalized, persistent learning support 

This personalization is achieved through:

• Long-term memory of individual goals, study habits and feedback, allowing the AI to adapt strategies and recommendations uniquely for each student
• Just-in-time nudges and meta-cognitive coaching, which help students develop self-regulation, resilience and effective learning strategies exactly when they need them
• Support for the emotional well-being and mental health of the student, using the emerging capacities of AI to sense and respond to emotions

Integration and scalability

The integration of cognitive twins across courses and co-curricular activities ensures continuity and coherence in a student’s educational journey, breaking down silos and enabling holistic development. Since these AI systems can operate at scale, they provide high-quality, individualized support without overburdening faculty, making advanced mentorship accessible to all students.

Empowerment and data ownership

By giving students ownership of their data with opt-in sharing, cognitive twins respect privacy and empower learners to control their educational narrative, fostering trust and engagement.

A worthy addition to the list: Angela, My 24x7 Research Assistant 

We focus a lot on learning. But as a researcher and writer, I am fascinated by the emergence of tools like Research Rabbit, SciSpace and Undermind, which make finding and reviewing research resources so much easier than it used to be. Seconds after providing an appropriate prompt, there are 100-200 papers or studies that “fit” the prompt and meet my query.

I have developed AI agents using Zapier. They go out to a range of sources twice a day and automatically bring to my attention new studies or research related to my interests and ranks them against the “marker” studies I have used to benchmark my interest (so-called “flagship” papers). Zapier finds material — in any language — and then summarizes them in English and places them in my review files.

My imagination now has AI reviewing the materials and drafts of papers or books I am working on and adding additional materials in draft form to my own drafts for me to review, edit, delete, change or check. I can also use these agents for “mock” peer review to anticipate concerns of journal editors and their reviewers. “Automated drafting” is what we might call this.

Some may see this as a form of research cheating — because aren’t we supposed to do all this laborious work ourselves? My response? No. We are supposed to take responsibility and ownership of our work, but getting help from AI or a human research assistant should be commonplace. Given that there are more than 5.2 million academic papers published each year, I see this as an efficient and effective use of the available resources and my time.

Back in the late 1970s I was awarded a grant to hire my first ever research assistant,  Angela Ffion Williams. Hence the name of my 24x7 assistant: Angela.

Contributor profile: Dr. Stephen Murgatroyd

Contact North | Contact Nord Research Associate Dr. Stephen Murgatroyd is an internationally recognized leader in educational innovation, strategic foresight online learning systems, artificial intelligence in higher education and technology-enabled learning.

The author of 50+ books and more than 200 peer-reviewed papers and policy reports,  his works include The Future of Higher Education in an Age of Artificial Intelligence, Rethinking Education in the Age of Technology, and Beyond Resilience: Thriving in a Time of Disruption. He is also a regular contributor to TeachOnline.ca, EDUCAUSE Review, and academic foresight journals.

Areas of specialization:

• AI in higher education: Assessment, teaching, advising, learner support
• Strategic foresight & scenario planning: Helping institutions imagine and prepare for multiple futures
• Innovation & systems change: Transforming how institutions design, deliver, and scale learning
• Online & open learning: Champion of access, equity and quality in digital education