Bloom's Taxonomy (AI Edition)

Generative AI is shaking up education on multiple levels. The much-anticipated and much-discussed white paper from APRU, "Generative AI in Higher Education: Current Practices and Ways Forward," (Jan 2025) laid out the challenges and spoke urgently about recommended actions. It was a welcome guidepost. But what is most heartening to me, in the midst of so much uncertainty, is the near-total agreement within that paper and elsewhere about one profound area of opportunity for AI: improved teaching and learning.

When it comes to the all-important interface of teachers and students in a classroom, AI, it seems, is genuinely here to help. Improvements can and will be made using AI, not just in speed of creating lessons, but in the quality of those lessons. "Here's the content. I've got 50 minutes. Use Gagne's Nine Events. Build a quiz." In seconds, you have a lesson that is very good, not perfect, but certainly a better starting place than a blank sheet of paper.

But all that assumes we are using the same approaches in the classroom. The ubiquity of AI, coupled with the assumption that students are often more familiar with it than teachers, necessitates fundamental changes in how we teach those lessons that AI helps us build. We need to move well beyond traditional ideas about delivery of information and assessment of knowledge, and cultivate in students the new skills that empower them to think for themselves while they leverage AI. But how do we do this?

I suggest that we already have a very powerful framework for addressing these new, unprecedented pedagogical demands. It's our old, trusty teaching partner: Bloom's Taxonomy. If we look at Bloom's through an AI lens, and AI through a Bloom's lens, we can start to see how we might build classroom learning experiences that leverage AI's strengths while at the same time cultivating essential human skills for the AI-enhanced future: AI literacy, critical thinking, source evaluation, metacognition to name a few. Let's take a look.

The foundational level of Bloom's is Remembering. While rote memorization has obvious limitations, instant recall of fundamental knowledge remains crucial. As I posit in my post here, it may become even more important in the age of AI. I have actually used AI to build my own flash cards in order to ingrain and retain new knowledge that I want to recall instantly. But the primary shift with AI regarding this level is not about memorization at all. It's about asking. It's about how to formulate effective prompts so AI provides answers efficiently and accurately. In other words, there are times students can and should use AI for "cognitive offloading." To use a common parallel, a lot of rote knowledge can and should become like the slide rule, effectively replaced by the calculator that is AI.

But whether they remember facts or dial them up, students must always have Understanding. And AI excels at explaining complex concepts in different ways and designing associated assessments. Instead of relying on one-size-fits-all textbooks or even classroom explanations, educators can guide students to use AI as their own personal tutors to clarify concepts in ways that resonate with their individual learning styles, and that match their current comprehension levels. Some powerful tutor programs are already available, some of them at no cost. Tutors free up educators to focus on fostering metacognition, helping students understand and improve their own learning processes. And we've known for decades how much metacognition can enhance learning.

When it comes to Applying knowledge, AI can assist with standard problem-solving but it runs aground wherever there are novel or ambiguous situations, in other words, where theory meets reality. Since the real world remains quite mysterious to AI, human guidance at this level is crucial. Learning experiences can involve students collaborating with AI to tackle complex challenges in hypothetical situations, while teachers guide their application in the real world. This way, students can develop through experience the ability to discern when AI's suggestions are effective and when they are lacking, and why.

Which takes us up the ladder to Analyzing, a skill that is particularly critical in this new landscape. While AI can identify patterns and analyze data with remarkable speed and mathematical precision, it often misses contextual, cultural, and ethical dimensions that human intuition can perceive, sometimes instantly. A seemingly minor symbol in literature or art can carry profound cultural weight, or what reads as objective reporting may contain a subtle yet powerful framing bias. We need to guide students to rigorously analyze AI's output, and in doing so we can teach them a skill that is applicable to all of life.

Evaluating in an AI-driven world requires that we intensify our focus on teaching ethical reasoning and making value judgments. AI can assess options based on predefined criteria, but it cannot inherently determine what is important, or desirable, or right, or good. It cannot inspire. It cannot lead. Learning activities aimed at developing students' moral reasoning and their capacity to be compassionate, trustworthy, and impassioned citizens will inherently involve critically assessing AI-generated content. In fact, just having the open conversation about what AI can and cannot do is a valuable approach to developing a cohesive view of what it means to be human.

Creating, then, carries the most significant opportunity of all. An AI image generator I was using recently popped up this cheery message: "Your imagination is the only limitation!" At first this struck me as just another marketing line, but after some reflection I realized that the GenAI was making a profound statement about itself. It was plainly stating that the creative capacity of AI is fundamentally and forever dependent on human ideas. In a GenAI world, educators will need to intentionally cultivate creativity, helping students experiment, overcome creative blocks, generate and rapidly prototype new ideas, with the goal of expanding their own realms of possibility. AI can serve as a powerful partner in bringing innovative ideas to life, allowing learning to focus on real solutions, on purpose and meaning, and on consequential impact in the real world.

We are rapidly entering what I call the "idea economy," where information is valuable only as fuel for the fire, and the fire is all about generating ideas: new approaches, better methods, improved products, new avenues of exploration -- and then collaborating with AI to develop and test and implement them for the betterment of all. The APRU White Paper astutely recognizes the urgent need for pedagogical adaptation that incorporates AI. I'm suggesting that the sturdy workhorse of Bloom's can pull and steady our wagons as we pioneer the wild and uncharted territories of AI-enhanced learning ahead.