Seeing Learning Differently: What Generative AI Reveals About Human Capability Development

Across higher education, conversations among faculty, instructors, and other teaching and learning professionals about generative artificial intelligence (GenAI) often begin with practical questions. How should courses change? What belongs in assessment? How do we as educators discourage inappropriate AI use while preserving academic integrity?

Yet as those conversations unfolded, something deeper began to emerge. What began as concerns about AI technology shifted to discussions about how learners think, make decisions, and develop the human capabilities needed to navigate a volatile, uncertain, complex, and ambiguous (VUCA) world. At the heart of the shift was a more fundamental educational question: What should we pay attention to if we want to understand how learners develop the capabilities we value most? This question shifts our attention away from whether AI can produce a particular piece of work and toward what learners reveal through their interactions with ideas, complexity, feedback, peers, and, increasingly, AI itself. The capabilities that colleges and universities seek to cultivate have not changed. What has changed is how we can recognize evidence of their development.

Recognizing Evidence of Learning

Learning itself cannot be observed directly.^Footnote1 It becomes visible through questioning assumptions, explaining reasoning, revising understanding, justifying decisions, and applying knowledge in new contexts. Essays, presentations, projects, portfolios, discussions, and AI-supported learning activities each reveal different aspects of learner development. Together, they provide a richer body of work for us to recognize how learners are developing over time.

These activities are a window into learning. Windows allow us to see inside, but they never reveal the entire room. A polished final product can demonstrate what a learner achieved while leaving much of the questioning, uncertainty, revision, decision-making, and reflection that shaped it outside our view.

GenAI has not pulled down the window shade. Instead, it has created new opportunities for learners to externalize aspects of their thinking while reminding us to look more intentionally at what those interactions reveal.

A single demonstration provides valuable evidence of learning but cannot fully reveal how it develops over time. Through experience, learners develop the capabilities they need to frame problems, evaluate information critically, adapt to change, collaborate thoughtfully with AI, and continue learning in unfamiliar and increasingly complex situations. In a world where GenAI is becoming part of learning, work, and professional practice, the enduring contribution of colleges and universities is designing learning environments that support the development of those capabilities while creating opportunities for that development to become visible and recognized.^Footnote2 Here, "capabilities" refer to learners' capacity to navigate complexity by thinking critically, reflecting on their own thinking, exercising sound judgment, adapting to changing circumstances, generating new possibilities, and making informed decisions when answers are incomplete. It is sustained by a willingness to continue learning, particularly when knowledge is evolving, contexts are changing, and uncertainty remains.

Like learning itself, capability development unfolds over time. It is shaped through repeated opportunities to frame problems before solving them, question assumptions, reflect on one's own thinking, adapt to unfamiliar situations, communicate across different perspectives, integrate ideas across disciplines, exercise ethical reasoning, and make decisions when certainty is unavailable.^Footnote3 Final products remain valuable evidence of learning, but they tell a richer story when interpreted alongside the thinking, interactions, revisions, and decisions that contributed to them.

Perhaps the more useful question is no longer how GenAI changes teaching or assessment. Instead, we might ask how learning environments can create opportunities for learners to make the development of these capabilities more visible. That question creates an opportunity to see learning differently.

Inviting Learning to Become More Recognizable

What does this shift in focus look like in practice? One place to begin making this shift is by paying closer attention to the interactions that most strongly contribute to the learning process rather than looking only at the final product. Questions learners ask, ideas they reconsider, explanations they refine, decisions they make, and moments of uncertainty all provide valuable evidence of capability development. Together, these interactions provide a fuller picture of capability development than a finished product alone.

GenAI can make more of these moments visible. As learners use AI to explore ideas, compare perspectives, challenge assumptions, or test emerging thinking, they leave behind traces of how their understanding develops. The educational value lies not in the AI response itself but in how learners engage with it: what they question, refine, reject, and ultimately use to construct their own understanding.

A simple example illustrates what changes when our attention shifts from evaluating a final product to understanding the learning that shaped it. Imagine two people completing the same assignment. At first glance, they appear to have reached the same destination. Both submit thoughtful work demonstrating a high level of achievement. Yet looking only at the finished submission, however, tells us little about the different paths they took to get there.

For the first person, the final submission is the primary evidence available. They engage with course materials, use generative AI to clarify ideas, organize information, and refine their work, ultimately producing a strong final submission. The work demonstrates what was accomplished, but it offers only limited insight into how understanding developed along the way.

The second person approaches the assignment with a different purpose, treating the process as a critical part of their learning journey. Before engaging with AI, they clarify what they are trying to understand, identify what they already know, recognize where uncertainty exists, and consider what information is still needed. As they work with AI, they compare perspectives, test assumptions, revisit course materials, revise their thinking, and explain why some suggestions are accepted while others are set aside. Throughout the process, they document questions, reflections, and key decisions that shape their understanding before arriving at the final submission.

Both submissions demonstrate what was learned. However, looking only at the final submission tells us what was accomplished, while looking beyond it reveals how understanding developed through questioning, revision, reflection, and informed decision-making. Taken together, these moments provide a more complete picture of how human capability developed over time. Reviewing all of the interactions does not diminish the importance of the final submission; it broadens our understanding of learning by revealing how each person's understanding developed and where meaningful feedback would best support continued growth, not just evaluate a finished product.

Another practical way to make learner development more recognizable is to invite them to articulate their thinking throughout the learning process. Simple questions such as these can help:

• What evidence or counter-argument changed your thinking? This requires learners to track the evolution of their own reasoning.
• Which suggestion did you reject and why? This question probes their critical evaluation and requires them to justify their own reasoning.
• What evidence became most important? This asks them to prioritize and synthesize their research and thought.
• What uncertainty remains? This question encourages honest self-reflection and points directly to future areas of growth.

Questions like these invite learners to make their reasoning explicit. They build on longstanding educational practices that use dialogue, reflection, and guided practice to better understand how learning develops over time and to inform meaningful feedback.^Footnote4

This perspective also broadens what counts as evidence of learning and capability development. Final products remain important, but feedback and assessment can be enriched when we as educators also consider learners' reasoning, revisions, and decisions throughout the process. As learners reveal more of their thinking, we are better able to understand how capability develops alongside demonstrated achievement. Taken together, these insights provide a fuller picture of both achievement and growth.

Beginning Where We Are

If human capability is the focus, the opportunity may not be to redesign every course or rethink every assessment. It may be to begin paying attention differently.

Many opportunities to make learning more visible already exist. People explain ideas during discussions, revise work in response to feedback, justify decisions, compare alternative approaches, and reflect on how their thinking has changed. GenAI introduces additional opportunities for these interactions, but the educational purpose remains the same: creating learning experiences that help learners develop and demonstrate the capabilities they will carry into professional, civic, and personal life.

Faculty, instructional designers, academic leaders, and others who support teaching and learning may find it useful to begin with the following questions:

• What distinctly human capabilities are important for learners to develop?
• In what assignments and learning activities do learners already have opportunities to reveal the development of those capabilities?
• What additional opportunity could be created within a learning environment for learners to explain, question, reflect, revise, or demonstrate how their thinking is evolving?

These questions do not require us to abandon familiar educational practices. In many cases, they simply invite us to build on what already exists. A reflective conversation following a presentation, a comparison of alternative solutions, an explanation of why one approach was chosen over another, or a structured opportunity to revise work after feedback can reveal dimensions of learning that might otherwise remain hidden.

GenAI has prompted colleges and universities to revisit important questions about teaching, learning, and assessment. One of the most valuable effects of the emergence of AI in education may be that it has brought renewed attention to a question that has always been present: How do we know learners are becoming capable? Becoming capable means more than demonstrating what has been learned. It means developing the capacity to frame problems thoughtfully, think critically about information, exercise discernment, adapt to changing circumstances, communicate across differences, and continue learning in new and unfamiliar contexts. That question shifts our attention beyond what learners produce toward how they are becoming capable of navigating complexities over time.

Seeing learning differently does not begin with technology. It begins with creating opportunities for learners to reveal how their thinking, decisions, and understanding develop over time. When those opportunities exist, assessment is informed by richer evidence of both achievement and development. Teaching becomes more intentional because learning experiences are designed not only to evaluate what learners know, but also to understand how learners are becoming more capable. The lasting contribution of GenAI may lie not in what learners produce, but in how it invites those involved in teaching and learning to better recognize, support, and cultivate human capability.

Author's Note

OpenAI ChatGPT (GPT-5.5) was used to support editorial refinement and language revision. All ideas, interpretations, examples, and final editorial decisions were developed, reviewed, revised, and verified by the author.

Notes

1. James W. Pellegrino et al., Knowing What Students Know: The Science and Design of Educational Assessment (National Academies Press, 2001).Jump back to footnote 1 in the text. ↩
2. UNESCO, Guidance for Generative AI in Education and Research (UNESCO, 2023). Jump back to footnote 2 in the text. ↩
3. Valerie J. Shute, "Focus on Formative Feedback," Review of Educational Research 78, no. 1 (2008): 153–189.Jump back to footnote 3 in the text. ↩
4. Allan Collins et al.,"Cognitive Apprenticeship: Making Thinking Visible," American Educator 15, no. 3 (Winter 1991): 6–11, 38–46. Jump back to footnote 4 in the text.↩

---

Tope Onitiri is Learning Designer at Harvard T.H. Chan School of Public Health.