Architecting the Path of AI into Education

Author, Dawn Booth, Ph.D

Christopher Alexander, a pioneering architect and design theorist, believed that good structures emerge from careful observation of how people naturally live and move, and essentially, how they create meaning. He encouraged designers to let patterns reveal themselves — to let the path first show itself — before imposing solutions: “the form of a place must come from the life that is lived there.” His philosophy of architectural design offers a powerful analogy for higher education as it experiences one of the most disruptive periods in modern history, certainly in the last quarter of a century that I have been teaching.

The use of AI in English-language teaching and learning in higher education has caused major debate among academics, not least among colleagues at the well-respected STEM university where I work. Rarely have I encountered discussions where views conflict so sharply. Debates, at times, become understandably heated. For some, plain and simple, the use of AI to complete a written assignment is cheating — the thought of opening the floodgates to AI signals the end of authentic learning, critical thinking, and, for many, perhaps, job security. For others, the use of AI in higher education brings us closer to creating an equal playing field. For years, students with strong critical-thinking skills and creativity have been limited by their English level. What more can these students do with powerful machine-learning tools at their disposal?

Ethical and philosophical arguments aside, sophisticated machine-learning tools are now readily available to students — and they are eager to use them. It has become highly impractical, and often impossible, to detect or prove their use. As a result, the conversation is shifting: rather than debating whether students should be allowed to use these tools, educators are increasingly asking what the best teaching practices are for integrating AI effectively into higher learning. And importantly, how must our assessment practices evolve in response to these new tools?

This is where I return to the design principles of Christopher Alexander. In many universities,

the fear of failing or losing grades for using AI has eased, and open and honest dialogue about the use of AI with students has begun. As a result, we can start to understand how students are actually using AI to complete their assignments. What paths are emerging? Which ones lead to higher-quality, audience-focused writing? What constitutes high-quality AI-assisted writing? Which patterns of AI use support higher-order critical thinking, and which limit engagement and learning? What forms of assessment might guide students toward deeper interaction and more thoughtful use of AI? Which areas of language or content must still be assessed in controlled environments where AI is inaccessible? And what balance between these approaches best supports teaching and learning? There is much to observe.

Answering these questions will require extensive research across a wide range of students and contexts — and experts are quickly responding. Still, much is uncertain, and AI is rapidly

adapting. What is clear to me, however, is that in order for AI use to be both educationally

meaningful and culturally impactful, we need to collaborate with students to reveal the paths

forward rather than quickly imposing solutions, much as Alexander urged designers to do.