Learn about Crystal’s approach to AI with our US Director of Teaching and Learning, Julia Rubin ‘00. When innovation moves as quickly as AI, it is tempting for schools to chase the newest tools or trends. At Crystal, however, we’re taking a different approach: leading with our mission. Through transformative teaching and meaningful connections, we empower kind, curious learners to thrive together. This commitment reminds us that even as technology evolves, our core purpose doesn’t.
But what does that truly mean, to stay “mission true,” and how is that achieved? At Crystal, it means leaning into the concept,
Transformative Teaching, understanding its definition and critically evaluating the ways in which transformative teaching can be enhanced by using AI, so that students’ learning is enriched and deepened, not bypassed. The
Transformative Teaching Framework at Crystal is composed of five core components: Active Learning, Student-Centered Learning, Collaborative Learning, Problem-Based Learning, and Experiential Learning. None of these modalities is mutually exclusive; they co-exist and complement each other, and, of course, can be used with or without AI.
In a recent Crystal AI Student Survey, over 140 students reported that using AI deepened their learning. This finding, of course, highlights the dilemma AI presents. We respect our students’ insight that AI can heighten the learning experience, but we want to avoid another reason people may be attracted to AI, that it merely shortcuts the effort and work one must put into a project. Thus, our challenge is how to harness the tremendous capabilities of AI without discouraging a student’s honest engagement with a subject that is necessary to master it.
How is this achieved, and what does it look like in practice? Here are some examples: In Ms. Agoff and Dr. Jones’ 9th Grade Biology class, students took a creative approach to this year’s cell organelle project and were provided balanced and appropriate guidance: they were allowed to use Flint AI only to generate ideas for the project, but all analysis had to be written by the student. The instructors also provided just-in-time guidance concerning how to leverage AI. For example, they suggested that AI be consulted regarding the project’s presentation. AI was then used to generate an idea of how to represent a three-dimensional cell in a written report–taking a picture of a three-dimensional model and including that picture in the report. In this way, the integrity of students’ engagement with the academic material was maintained, while they also experienced using AI to maximize their creativity.
Using AI to serve as personalized language tutors in Mandarin, French, and Spanish courses is another way we have deepened students’ learning. By using AI as language tutors, we further students’ mastery of the material on a level that is not readily duplicated without such a tool. AI serves a similar tutorial function in many classes, for example, in teaching basic programming in Computer and Data Science, as a writing coach in English and History classes, and in studying deep research in Cognitive Neuroscience. Even more challenging is the use of AI to fact-check nutritional information found on TikTok. This, of course, serves a dual purpose: (1) teaching students about nutrition and (2) teaching students how to fact-check AI and avoid AI “hallucinations.”
Of course, over time, AI, our familiarity with it, and the ways we use it will evolve in unforeseen ways. What can clearly be foreseen, however, is our continuing commitment to use it in ways that promote Transformative Teaching in order to foster our students’ understanding and mastery of new skills.

Julia Rubin '00,
US Director of Teaching and Learning