The Promise of Personalized Learning, Enabled by AI

From introductory gen-ed classes to advanced graduate seminars, wherever classes online or on campus include more than a couple of students, we have struggled with finding ways to assure that all students are given personalized attention to meet their learning needs.

This has led to differentiated learning models in which students are presented materials based on assessments conducted prior to the class. But that approach too often fails to adapt to progress during the semester and misses opportunities for exchanges and synergies among all learners. It is also most practical only when there are enough classes to support multiple sections at the differentiated levels or multiple groups within a single class.

As expert systems and AI technologies have developed, the promise of personalized learning is now being tested. Matthew Lynch, the "Tech Edvocate," describes one model:

• First, learning is guided by the interests of the student. Teachers will guide students to select materials, projects and products that reflect student interests.
• Second, students have more choice in virtually every aspect of the process, including where, when and how they learn the material.
• Third, teachers take on the role of coaches instead of the role of information purveyors.
• Fourth, the pace is determined by the learning process of the student.
• Fifth, ed-tech tools are used to manage the multiplicity of learning experiences.

This model utilizes best practices in engaging the learners in class design and adopting materials tied to interests of students. It acknowledges different learning preferences of students. And it uses adaptive learning to accommodate the pace of the students. Managing the adaptive release and assessment utilizes smart technologies.

Georgia Tech's Commission on Creating the Next in Education notes that:

Based on the successes of these preliminary experiments, the Commission recommends pilot projects on appropriate adaptive learning platforms that could be customized by faculty who would insert the topical content. Some of these experiments may include interactive books and interactive videos, as well as AI agents like “Jill Watson” for many Georgia Tech classes, especially large, remedial, and/or online classes. Some of these adaptive learning platforms can also be transferred to K-12 education as well as many graduate classes (both online and in-person). Pedagogical experiments might be considered that examine where and how this personalized learning is effective. Besides being an integral part of a course, personalized learning modules can be used to support students of varying backgrounds and abilities, or to streamline a curriculum.

We are on the cusp of a new era in which learning is personalized to the needs and interests of the students. With the advent of the lifelong, 60-year learner, we will see more heutagogical approaches to meet the expectations of the self-determined adult learner.

These will require that the learner is adept at designing the learning goals and outcomes. The faculty member will help to guide the learner to relevant resources and suggest paths to reach the desired outcomes.

This may be a far different approach from the standardized lockstep chapter by chapter, weekly quiz and objective final exam of years gone by.

How are you preparing faculty members and students to make the transition to personalized learning? Have you seen expectations shifting from pedagogical approaches to heutagogical practices?

Perhaps this is worthy of a departmentwide or collegewide discussion at your institution.