Not long ago, I participated in an exercise that asked educators to define thinking and learning. It was a familiar prompt, one we have returned to countless times over the past decade.
This time, it felt different. The task was to triangulate, even pinpoint, what these concepts mean in today’s educational landscape.
The conversation was thoughtful and wide-ranging. Educators from varied contexts shared perspectives shaped by their classrooms, their students, and their lived professional realities. As the discussion unfolded, a shared realization emerged: Our understanding of thinking and learning is becoming increasingly abstract.
As a chief academic officer, I spend much of my time thinking about how learning is designed and measured. Yet, in that moment, listening to educators wrestle with the meaning of thinking itself, I knew the challenge was no longer to define thinking, but to work in a world where its definition is constantly shifting.
The Shift We Didn’t Plan For
Education has always adapted to new tools, but rarely at this pace. In a matter of months, technologies capable of summarizing texts, generating essays, and mimicking academic voice have become widely accessible in classrooms. What once required sustained cognitive effort can now be produced in seconds (Doss et al., 2025).
The result is not merely a new instructional challenge; it is a fundamental shift in what it means to learn.
For generations, schools treated knowledge acquisition as the central hurdle. If students could read closely, recall accurately, and write coherently, they were considered prepared. Tasks that once demonstrated understanding now signal access.
This does not make learning easier; it makes it different. And it forces us to confront an uncomfortable question: If machines can do much of what we once taught students to do, what should learning now require?
Literacy Beyond the Page
Bloom’s Taxonomy (Krathwohl, 2002) has long articulated cognitive rigor. Remembering led to understanding; understanding enabled application; application supported analysis, evaluation, and creation.
But artificial intelligence is flattening that progression.
What once represented higher-order thinking—summarizing a text, drafting an essay, explaining a concept—is now executable at the push of a button. These tasks no longer serve as reliable indicators of mastery (Walker & Vorvoreanu, 2025). They have become baseline capabilities within the learning environment.
Artificial intelligence does not invalidate Bloom’s premise; it reframes it. In an AI-rich world, the lower levels of the taxonomy are no longer destinations; they are starting points.
The true measures of learning now lie above them. Can students interpret nuance rather than extract information? Can they evaluate credibility instead of repeating content? Can they connect ideas across disciplines and explain why something matters?
These are not extensions of literacy. They are new definitions of it. In this sense, literacy is no longer merely technical. It is interpretive. Ethical. Strategic.
This kind of literacy cannot be automated. Automation can, however, support its development.
Designing for Thought, Not Just Performance
To meet this moment, we must rethink how learning experiences are designed: framing tasks that require judgment, designing assessments that foster analysis, and valuing ambiguity and intellectual risk.
When applied intentionally, automation through AI can strengthen, not dilute, this kind of learning. For students, its greatest value lies in responsiveness. Research shows that AI can adapt in real time, offering targeted practice when gaps emerge; enrichment when mastery is demonstrated; and prompts that ask learners to explain their reasoning, compare approaches, or revise claims as their thinking develops (Hariyanto et al., 2025). It can also support deeper engagement through simulations, branching scenarios, and feedback loops that respond to student choices without turning learning into a race for completion.
Just as important, automation can protect student focus. By reducing cognitive clutter, streamlining navigation, pacing tasks, and offering timely hints, AI keeps learners in productive struggle rather than frustration or disengagement.
For teachers, the benefit is leverage. Used well, AI functions as an instructional partner in the invisible work that consumes time but does not require uniquely human judgment (Ash, 2025). It can draft lesson variants, surface patterns across student work, suggest groupings, and prepare concise summaries that help teachers see which students need support and why.
The result is not automation of teaching, but an expansion of a teacher’s capacity to teach well.
Practically, this means automating what can be standardized and continuously improved, collecting evidence of learning, tagging misconceptions, generating formative checks, and organizing instructional options—all while preserving teacher judgment as the final authority. The teacher remains the editor-in-chief: approving, revising, and applying professional discernment, while the system does the work of noticing more and preparing faster.
This is the promise of AI in education: not accelerating answers but amplifying reflection; not replacing judgment but making room for it.
