This is a delayed post from NCTE 2025 – where I was fortunate enough to deliver four presentations/roundtables (see list below) as well as launch my new, co-authored book, Teaching Writing in the Age of AI – as I am now reflecting on work last semester with 1) my own students in a writing-intensive course as well as 2) a group of K-12 colleagues in a sustained series of PD experiences focused on AI and 3) a higher education faculty learning group focused on the intentional integration of AI.
The work related to AI integration in all aspects of the teaching and professional learning work that I have been doing this fall has been rewarding and challenging with the overall results being, well… sporadic.
While overall I feel that it has been good, I am finding nearly as many stumbling blocks as successes. Briefly, I want to offer some reflections on limited use cases of AI in my teaching, through professional learning in K-12 and higher ed settings, a particular instance where I worked through disciplinary ways of thinking with a librarian colleague in an effort to build an “agent” in Co-Pilot, and as part of an adminsitrative assignment I was tasked with through an interdisciplinary group.
Each of these uses has come with some insights about when, why, and how I would want to encourage students and colleagues to use AI, and when I would otherwise encourage them to take pause. In particular, of all the AI-related frameworks that are out there to help one make these kinds of decisions, I am finding that I keep returning to the WestEd Digital Fluency group‘s “Friction by Design: A Framework for Centering Learning in the Age of AI.” This is one of many frameworks that are out there, and it focuses less on specific AI literacy skills and more on elements of learning design more broadly.
They contend that:
AI has the power to reduce both kinds of friction. This is what makes its use in education so powerful, and also so risky. Without careful design, AI can just as easily remove the thinking as it can remove the noise. Using AI in learning contexts requires more than permission or enthusiasm, it demands intentionality.
In this framework, the WestEd Digital Fluency group describes how “friction” can be both positive and “productive friction,” encouraging purposeful effort from a learner as well as removing “unproductive friction,” thus slowing a learner down through hesitation or frustration. While the WestEd Digital Fluency group doesn’t position these two types of friction on a continuum or even necessarily in opposition to one another, I am beginning to think about the use of AI in terms of “turning up” or “turning on” either kind of friction in some cases while, in others, using AI for “turning down” or “turning off” friction.
For instance, in the process of “increasing” friction, there are times when I might ask learners to engage in a conversation with AI chatbot I’ve designed to encourage them to slow down, explain themselves more fully, and think through ideas in a more nuanced manner. Rather than reducing the “unproductive” kind of friction related to “activation energy,” instead I am intentionally trying to get them to slow down and activate even more energy. By doing so, I am also hoping that this would increase, to use the West Ed terms, their “cognitive ownership” of the task and engage in an AI-infused round of “social sense making.”
On the other hand, when I “reduce” friction, there are times that I might ask learners to offload some cognitive work to AI, intentionally, so we can move ahead to other work. This might be cognitive work that they already know how to do and I simply want to skip ahead to deeper, more substantive work. Or it could be work that I want them to use AI for so they can see a process unfold (with an AI output that they can then adopt wholesale, adapt in part, or push back against). In whatever way I am choosing to reduce friction, I am making a choice that acknowledges how all students will be moving past some of the typical friction points that we would typically (whether explicitly or inadvertently), and that I will need to work with some students to consider what they may be missing by not going through the entire process of unpacking the assignment and engaging the thinking process that would ensue.
No matter how I am choosing to use AI with students and colleagues, these are intentional, instructional design decisions that I – and all educators – must consider at each moment during an instructional cycle, and as we craft meaningful, authentic assignments. Over the next semester, my goal will be to share some ideas that I have been trying out, and, as I always do, with the intent of keeping the core goal of using any digital tool to support writing and learning. For now, my hope is that the links to my NCTE presentations can be helpful in outling some of my ideas (and acknowledging the sources that I am drawing from in doing so).
Troy Hicks – Presentation Slide Decks from the National Council of Teachers of English 2025 Annual Convention (Denver, CO)
- AI in Education: Problems, Potentials, and Possible Futures
- Introducing, Interpreting, and Integrating Primary Sources in the English Language Arts Classroom
- Crafting Prompts and Building Bots (part of a panel presentation, Expanding the Conversation: Dreaming Boldly in Considering GenAI’s Promises and Pitfalls for Teaching and Learning in ELA Education)
- AI & Uncertainty: Exploring Principled, Nuanced, and Authentic Approaches to Media Literacy in a Changing World
This work is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-sa/4.0/
