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 Professor Don A. Klinger, one of the leaders we invited to comment.
The Scenario Dr. Don A. Klinger Selected From our Five Futures Is “Cognitive Twins for Every Learner”
For a measurement researcher such as me, the most intriguing scenario is the concept of a cognitive twin for every learner (Scenario 1) in combination with Scenario 3 (autonomous feedback engines). The foundations for both of these scenarios are already in place. The combination of these two scenarios meets important aspects of formative assessment and self-regulation, providing tailored feedback and direction for an individual in a learning situation. In order to maximize the value of such a cognitive twin, the twin will need to be more than a companion and cheerleader. This future AI twin will serve as a Learning Director, challenging the individual and asking questions that requires the learner to engage in Learning Dialogues with the AI, other students and the teacher. These learning dialogues with peers and teachers have already been shown to be critical methods for concept development, understanding, application of knowledge and skills, self-evaluation and criticality (key foundations for self-regulated learning). Effective dialogues can challenge one’s thinking, creating cognitive dissonance that leads to deeper conceptual understanding.
High Likelihood and Future Value
The foundations for these scenarios are already in place (Personal AI assistants (agents), assessment scoring and feedback, AI generated tutoring) are already available. Currently, it is possible to direct an AI “companion” (twin) to access one’s “digital profile” to complete a variety of tasks and interactions. There are AI systems that instantly score a student’s work, provide feedback, and even direct the student to an AI generated tutorial specific to the learning concepts not yet mastered. AI systems will become critical for foundational learning and assessment, ensuring learners have “mastered the basics”, allowing educators and learners to focus on application, evaluation, and creation of new ideas and solutions.
The future value of this scenario will be fully realized when it provides opportunities to connect people together, whether it be with a teacher for individualized instruction, or with a group of peers to work or learn together on a common problem.
How do We Get There
There are issues to overcome. In terms of assessment and feedback, it is critical the data sources are well defined and “clean”, and additions to these data sources will need to be shown to be accurate, without being limiting. Further, current AI companions (twins) tend to be overly supportive, rather than serving as a “critical friend” or “digital conscience” (the vision I keep having is Jiminy Cricket who served as a conscience for Pinnochio). They need to challenge the individual or attempt to create cognitive dissonance, which is needed for deeper learning. There is also a likelihood that these digital twins will rely on personal data that the individual has long since discarded. Such shifts and transitions are already occurring. VR systems and agents can be instructed to ask questions and probe for more complete explanations. This same structure can be incorporated into AI twins.
The New Future Scenario Don Klinger Envisions Himself Is The Use of Using AI Twins to Connect People and Community AI
Expanding on the scenario above, one of my real interests in the use of AI for the purposes of learning dialogues is the potential for AI-generated assessment and AI twins to connect people together to solve complex or “wicked problems.” Integrated digital twins have the potential to find a diversity of perspectives amongst individuals who may then be best able to work on these problems. An addition to this is the creation of “Community AIs” that oversee a group of individuals and their digital twins. This community could be a classroom of learners, a company or organisation, a group of researchers, etc.
How Community AI Would Work
Imagine every teacher having access to a personalized AI agent, a co-pilot, which is accessible at the teacher’s beck and call, not to take over or share their workload, but as a mentor, guide, and sounding board. There is nothing untoward about this. Every professional needs a mentor, someone with the same or additional skills that a professional can chat with, confide in, and seek feedback from, as and when necessary. Most professionals and artists do this all the time. For example, they ask people to read what they may have written or composed, observe their practice and offer critical observations, or help with something with which they may be struggling. Most tertiary educators are teachers first, along with being researchers. While they possess advanced subject matter knowledge, many will lack the requisite technological and pedagogical knowledge, or the time and resources to acquire and keep abreast with it. It would be immensely helpful to explore how AI can be used to support teachers in their professional practice.
Strategic Value
- Connects people within the community to increase learning, productivity, and problem solving.
- Maintains the community knowledge, skills, and values.
- Avoids (removes) extraneous and flawed data sources that currently impact AI effectiveness, whether it be AI led searches, assessment and feedback.
- Helps to monitor and maximize the expanding community knowledge, and incorporates this to further enhance the effectiveness of members of the community.
