"Will AI Make My Child Intellectually Lazy?" — An Answer from Cognitive Load Theory

It is the most common question parents ask, and the most legitimate. Essays polished by AI, maths explained by AI, first drafts generated by AI — over the years, does a child’s own thinking atrophy?

Cognitive science gives an answer more honest than yes or no: it depends on how. Cognitive load theory holds that working memory is strictly limited, and that learning happens when that limited capacity is spent on generative processing — organising for oneself, inferring for oneself, erring and correcting for oneself. Seen this way, a tool can enter cognition in two very different modes.

The first is the exoskeleton: the tool absorbs low-level load — searching, formatting, translating, repetitive calculation — freeing capacity for higher-order thought: framing the problem, structuring the argument, making the calls. The learner remains the one deciding. The second is outsourcing: understanding and judgement handed over together, generative processing skipped entirely. An answer appears; no learning has occurred. The two modes can involve the same AI tool and the same interface — the whole difference lies in the user’s awareness.

This is why UAICS writes cognitive load and AI-collaboration awareness into the standard and makes it the core module of teacher training. The content unfolds in three steps: first distinguishing — after AI-assisted work, marking which ideas were one’s own; then knowing one’s load — what to think through oneself, what to delegate; finally writing one’s own rules — is AI, for me, exoskeleton or outsourcing? A child who walks this path develops what we call Interwoven Intelligence: AI as a genuine extension of the mind, with authority remaining at home.

For parents, one immediately usable practice: when your child shows you AI-assisted work, do not ask "did the AI do this?" — ask "which part was your idea?" The first question teaches concealment. The second teaches awareness.