Learn to Struggle with AI: Singapore Educators Push Back Against Cognitive Decline

Singapore's embrace of AI in the classroom has been swift and comprehensive — but a quieter second wave of the debate is now taking hold, one that asks not whether students can use AI, but whether they should. As tools like ChatGPT, Perplexity and Gemini become study companions for a generation that has never known school without them, researchers and educators are drawing a line: the real learning happens in the moments you choose not to ask the machine for the answer.

At the Singapore University of Technology and Design (SUTD), engineering student Puah Joon Leng, 23, has turned this philosophy into a daily ritual. He works through a practice problem fully before photographing his solution and asking AI to identify his mistakes. "When the AI schools me, it will be more impactful," he says. On open-ended assignments, AI acts as a research teammate — mapping the field, surfacing papers, and stress-testing his arguments — but the synthesis and final judgments remain entirely his. Studies increasingly back this instinct: heavy AI reliance measurably reduces brain engagement and exam performance, a pattern researchers call "cognitive offloading."

The concern is not about banning AI but about preserving what educators call "productive struggle" — the deliberate, effortful engagement with a problem that builds lasting understanding. SUTD and other institutions are designing assessment workflows where AI serves as a thinking partner rather than a shortcut, emphasising cycles of drafting, comparing, critiquing and revising. The goal is to close what some see as a new digital divide: not between those who have access to AI and those who don't, but between those who use it to think more deeply and those who use it to think less.

This shift in emphasis arrives as Singapore's education system is increasingly intertwined with AI tools at every level, from primary school adaptive learning platforms to university research labs. The Ministry of Education has invested heavily in AI literacy, but the cognitive dimension adds a new layer of complexity — it is not enough to teach students how to prompt effectively; they must also learn when not to prompt at all. For a system that has long prided itself on rigorous assessment, the challenge is designing evaluation methods that measure thinking, not just the quality of AI-assisted output.

Why it matters for Singapore: Singapore's economic future rests on its ability to produce graduates who can do what AI cannot — exercise judgment, connect disparate ideas, and think critically under pressure. As the country positions itself as a global AI hub, the paradox is sharp: the more AI capable the workforce becomes, the more valuable the distinctly human skills of struggle, synthesis and discernment become. The students who learn to wrestle with problems before turning to the machine will be the ones who end up directing it, not the other way around.