Edutainment, Gamification, and the Right to Struggle

We do not often frame ‘struggle’ as a right in education. Yet increasingly, I find myself wondering whether it should be. By struggle, I do not mean frustration for its own sake, but the productive kind that asks learners to think, wrestle, and make meaning before answers are given.

Recently, I was helping my ten-year-old son, Saleh, with a school research assignment. His task was to compare the human digestive system with that of a camel. It is a wonderfully niche topic, exactly the kind that does not easily lend itself to a single, ready-made explainer.

When I was his age, this kind of assignment would have unfolded slowly. I would have gone to a library, pulled multiple books off a shelf, skimmed and re-skimmed chapters, taken notes, underlined passages, and tried – often clumsily – to make sense of two systems before comparing them. The struggle was not incidental to the learning; it was the learning. Reading, synthesizing, sequencing ideas, writing and rewriting were inseparable from the task itself.

Saleh’s school, like many today, has a Bring Your Own Device policy and actively encourages students to use digital tools for research and creation. His end product would likely live on Book Creator or Canva. When he began, he instinctively turned to a child-friendly search engine, only to realize that there was no pre-packaged article neatly comparing humans and camels.

I suggested a first step: research each digestive system separately, then compare. Before he could even begin, his face lit up. He had a better idea—more efficient, more direct.

“Why don’t I just use Copilot?”

In seconds, an age-appropriate comparison could appear on his screen, bypassing the very processes his assignment was designed to cultivate. The reading; The critical thinking; The comparative reasoning; The drafting; The editing; The productive struggle.

If learners are always entertained and supported by frictionless tools, are they ever truly learning?

To be clear, Copilot is not designed for ten-year-olds, and with careful prompting it can be used in more scaffolded, intentional ways. Through our work running an AI Testbed, I have seen EdTech companies design tools that meaningfully support learning rather than short-circuit it. But the moment stayed with me because it captured something larger: learners today will always have easy shortcuts available to them, and increasingly, systems are being designed to offer those shortcuts by default.

Research from the learning sciences helps explain why this matters. In Make It Stick, the authors show that learning is strengthened when it feels effortful. Tasks that require retrieval, comparison, and synthesis create more durable understanding than information that is simply delivered. Similarly, Productive Failure demonstrates that learners who struggle with complex problems before receiving instruction often develop deeper conceptual understanding than those given immediate solutions.

Struggle, in this sense, is not about making learning unnecessarily hard. It is about preserving the cognitive work that leads to deep understanding. As the National Academy of Sciences note, knowledge is not simply absorbed; it is constructued and organized into mental schemas through effortful thinking, enabling it to be retained and retrieved over time.  When that effort is removed—when answers arrive before learners have had to think, compare, or make sense of ideas—we risk short-circuiting the very processes that make learning stick.

This raises a deeper question: if learners are always entertained and supported by frictionless tools, are they ever truly learning?

When “Engagement” Becomes the Goal

Few concepts in education are invoked as frequently—or as loosely—as engagement. We rightly want learning to be relevant, meaningful, and interesting. But somewhere along the way, engagement has become conflated with a very particular set of indicators: time on task, clicks, streaks, badges, daily usage, and my personal favorite, “fun”.

These are not neutral measures. They come directly from the logic of consumer technology, where engagement is optimized because it drives retention, growth, and profit. In that world, more engagement is always better.

Deep learning requires sustained attention, cognitive effort, moments of confusion, and often discomfort. It also carries its own reward: the quiet satisfaction – and yes, even the dopamine hit – of finally understanding something hard.

Education, however, has different aims. Learning is not a frictionless process. Deep learning requires sustained attention, cognitive effort, moments of confusion, and often discomfort. It also carries its own reward: the quiet satisfaction – and yes, even the dopamine hit – of finally understanding something hard.

International bodies such as the OECD have explicitly warned against equating surface engagement metrics with meaningful learning. Their work on student engagement cautions that observable participation or time-on-task can mask shallow interaction rather than genuine cognitive investment.

Evidence from practice reinforces this point. The Education Endowment Foundation Teaching and Learning Toolkit consistently shows that approaches with the strongest impact on learning—such as metacognition and self-regulated learning—are those that deepen thinking, not those that simply increase activity or interaction.

This is not an argument against making learning enjoyable. It is an argument against mistaking enjoyment for understanding.

Dopamine, Design, and Ethical Blind Spots

We now know that many digital platforms are intentionally designed to capture attention through dopamine-driven feedback loops. Streaks, variable rewards, and notifications are not accidental features; they are engineered mechanisms.

What is less discussed is how much of this design logic has migrated into educational technology. In edtech, these mechanisms are often justified by good intentions. After all, if learners are more motivated and persistent, isn’t that a good thing?

Here, contemporary critiques of digital design are instructive. In The Anxious Generation, social psychologist Jonathan Haidt argues that many digital platforms exploit reward systems in ways that fragment attention and reduce opportunities for deep, sustained thought—particularly among children and adolescents. His argument is not a rejection of technology itself, but a warning about design incentives that prioritize engagement over wellbeing and cognitive development.

When similar logics are imported into education, ethical questions emerge. Are we comfortable normalizing addictive design patterns because the context is supposedly learning? And if the tools used to “engage” learners reduce their tolerance for effort and ambiguity, what are we optimizing for in the long run?

Beyond ethics, there are pedagogical implications.

Why Deep Learning Requires Struggle

Decades of research in the learning sciences point to a consistent conclusion: deep learning is effortful. It requires learners to actively construct understanding, test ideas, and revise their thinking over time.

The National Academies of Sciences synthesis in How People Learn makes this clear. For knowledge to be transferable and enduring, learners must engage in sense-making, explanation, reflection, and practice—processes that are cognitively demanding by nature.

When educational technologies are designed primarily for ease, efficiency, and immediacy, they risk undermining these processes. Systems optimized for smooth progression and instant feedback can unintentionally remove the friction that signals learning is actually taking place.

Edtech, borrowing engagement models from Big Tech, often prioritizes what is immediately visible and measurable. Yet the most important dimensions of learning—conceptual change, persistence, metacognition—are not easily captured by dashboards.

This raises a difficult question for the field: can tools designed for surface engagement truly support deep, sustained learning at scale?

And then, AI Enters the Classroom

All of these tensions exist in educational technology more broadly. With AI, they intensify.

AI does not simply gamify learning; it personalizes it. It adapts in real time, anticipates needs, and increasingly resolves uncertainty before learners have a chance to sit with it. In doing so, AI has the capacity to personalize the struggle away.

At WISE’s AI-focused Summit in 2023, there was palpable pressure to act—to adopt, pilot, and integrate AI quickly. Two years later, at our 2025 Summit, that pressure had only intensified. Schools and systems are under strain to demonstrate innovation, justify investments, and show quick wins. Everyone is being asked to accelerate and optimize.

In that context, AI’s promise of efficiency is deeply seductive. But without careful design and thoughtful implementation, we risk creating learning environments where thinking itself becomes optional.

This does not mean AI has no place in education. Our Testbed work shows its immense potential when deployed intentionally. But it does mean we must be clear-eyed about the trade-offs we are making.

An Invitation, Not a Verdict

This is not a rejection of technology, edutainment, or AI. Nor is it a call to return to pre-digital classrooms. It is an invitation to pause and reflect.

As educators, designers, policymakers, and researchers, we need to reclaim struggle as a legitimate and essential part of learning.

If engagement has become our dominant metric, it is worth asking: engagement in service of what? If efficiency is our guiding principle, what are we optimizing away? And if learning is always supposed to be fun, what happens to the learner’s right to struggle?

As educators, designers, policymakers, and researchers, we need to reclaim struggle as a legitimate and essential part of learning. Not as an obstacle to be eliminated, but as a condition to be protected.

In the months ahead, we need deeper conversations about what it would mean to design for struggle rather than against it. That is not a question with easy answers. But it is one we can no longer afford to avoid.

References

• Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make It Stick: The Science of Successful Learning.
• Kapur, M. (2016). Productive Failure: Unlocking Deeper Learning Through the Science of Failing Productively.
• (2017–2023). Student Engagement and Learning Outcomes.
• Education Endowment Foundation. Teaching and Learning Toolkit.
• Haidt, J. (2024). The Anxious Generation.
• National Academies of Sciences, Engineering, and Medicine. (2018). How People Learn II.
• Stanford Institute for Human-Centered Artificial Intelligence. (2023–2024). AI in Education briefs and policy guidance.