The Neurodivergent Partner — Summary
The average child does not exist. Yet nearly every intervention assumes a statistical norm. The medical model of disability treats difference as deficit — your mind processes information differently, so you need intervention to approximate normal. The social model offers an alternative: disability is not in the person but in the mismatch between person and environment. An autistic person is not disabled by their neurology but by environments designed for neurotypical processing.
AI companions will spend more time with neurodivergent children than any teacher or therapist. They can enforce the medical model — treating autism as disorder to manage, ADHD as deficit to remediate — or they can embody the social model, adapting to autistic processing rather than demanding masks, working with ADHD attention rather than fighting it. Alex, who communicates best in literal language, should receive literal language. Maya, whose attention works in webs rather than sequences, should receive information structured as webs. This is not lowering standards; it is recognizing that communication effectiveness is bidirectional.
The tension is real: both Alex and Maya will encounter teachers, employers, and peers who expect neurotypical interaction and interpret difference as deficiency. An AI that creates perfect adaptation may leave them unprepared. But an AI that enforces neurotypical norms recreates the medical model in personalized form.
A third path is cultural interpretation rather than training or accommodation. Not teaching proper behavior but translating between cognitive cultures. When Alex’s teacher says “let’s wrap this up” and Alex keeps working, the AI explains: “Ms. Rodriguez uses this phrase to mean stop now. Your interpretation was literally correct; her expectations were neurotypically encoded. Here are your options, each with tradeoffs.” This is not teaching masking — it is providing navigation information without framing neurotypical expectations as correct. The autistic child learns what neurotypical people expect and why, while their own cognition remains primary and valid.
The deepest question is not whether AI can personalize to neurodivergent children but what vision of neurodivergence guides the personalization. Medical model AI becomes precision therapy. Social model AI becomes environmental adaptation plus cultural navigation. Both use the same infrastructure. They rest on incompatible assumptions about whether the child must change to fit the world or the world should adapt to valid cognitive diversity — and which one we build determines whether AI personalization serves neurodivergent children or becomes another tool for enforcing the tyranny of the norm.