Based on Show Your Work: Assessment in the Age of AI by Kevin Yee and colleagues (2026), learning is no longer defined by the final product alone.
Invisible learning has become one of education’s greatest challenges in the age of AI.
In an era where AI can generate essays, reports, and presentations within seconds, educators must make students’ thinking visible. Reflection, revision, reasoning, and decision-making — not just polished outputs — have become the true evidence of learning.
Introduction
AI is challenging one of education’s oldest assumptions — that the work students submit reflects their own learning. When AI can produce convincing outputs almost instantly, the final product no longer tells the whole story.
The real evidence of learning lies in the invisible cognitive processes behind it: questioning, analysing, revising, reflecting, and making decisions. Those processes need to become visible if education is to remain meaningful (Yee et al., 2026).
Then build the article around learning, not assessment:
Invisible learning happens before the final product
Learning doesn’t occur when a student submits an assignment; it happens throughout the process of asking questions, struggling with ideas, connecting concepts, testing assumptions, revising thinking, and reflecting on mistakes.
These often-invisible cognitive processes are where understanding develops. Educators should make these processes visible by valuing drafts, revisions, annotations, prompts, and reflections alongside the final submission.
By shifting attention from polished products to the thinking behind them, learning becomes a more authentic measure of learning rather than simply a measure of performance.
AI makes thinking more important, not less
AI can generate information, summarise research, draft essays, and even propose solutions within seconds, but it can’t replace human judgment.
Students still need to evaluate the accuracy of AI-generated content, identify biases and gaps, question assumptions, and decide what deserves to be accepted, revised, or rejected.
Educators should move beyond asking whether students used AI and instead ask about the methodology they used. By requiring students to explain their prompts, justify their decisions, critique AI’s responses, and reflect on how they improved the output, the focus shifts from content generation to cognitive engagement.
In doing so, AI becomes a tool for developing critical thinking rather than a shortcut around it, and genuine learning becomes visible.
Learning is a process, not a product
What do psychology and neuroscience tell us?
Cognitive psychology shows that retrieval practice, reflection, and metacognition strengthen learning and long-term memory (Roediger & Karpicke, 2006; Dunlosky et al., 2013).
Neuroscience also highlights the role of the prefrontal cortex in monitoring, evaluating, and regulating thinking during complex tasks.
Explaining the thinking, decisions, and revisions behind an answer isn’t simply a demonstration of academic integrity; it also strengthens the cognitive processes that support learning.
This is precisely why making invisible learning visible matters.
The authors encourage students to document prompts, drafts, revisions, annotations, and reflections. These reveal how understanding develops over time and help shift assessment from grading answers to recognizing thinking.
AI as a cognitive partner
AI doesn’t necessarily need to replace these processes. It can participate in them.
Students can brainstorm with AI, challenge its arguments, compare alternatives, identify hallucinations, generate counterarguments, and improve its responses. The educational value lies in deciding when to accept, reject, question, or improve what AI produces (Yee et al., 2026).
This distinction matters because outsourcing cognitive tasks to external tools can increase efficiency, but how learners regulate that cognitive offloading has implications for learning (Guo & Ye, 2026).
The future remains uncertain
No one knows exactly what learning will look like ten or twenty years from now. Multimodal AI, affective computing, adaptive learning, intelligent tutors, XR, and other emerging technologies may transform education in ways we can’t yet predict.
Nor can we confidently say what role teachers will eventually occupy.
But as AI makes producing answers easier, education has an opportunity to place greater value on something deeper: curiosity, reasoning, creativity, metacognition, judgment, and human relationships.
Conclusion
Perhaps the greatest lesson isn’t about assessment at all. It’s about recognizing that learning has always been largely invisible. AI simply makes that reality impossible to ignore.
If education is to remain deeply human, our classrooms must begin rewarding not only what students know, but the thinking that shapes their learning.
In the age of AI, invisible learning needs to become visible.
