Our perspective
In the AI era, education should begin with each child
AI models could offer additional explanations to support teaching. The purpose is learning at a pace that makes sense to each child.
An idea borrowed from distillation
In AI, knowledge distillation commonly means a student model learning from a teacher model. As an analogy, we imagine foundational education drawing on AI models as teaching support to bring knowledge, explanations and feedback to each child. The child is the learner, gradually building their own understanding.
Where personalisation should happen
It should go beyond putting a name into a story. What does the child already understand? Where are they stuck? Which explanation makes sense? When should we take a smaller step, or leave the screen and try something in the real world? A shared learning goal can allow different routes and rhythms.
Teaching support needs more than a test score
A teacher model suitable for children also needs reliable content, appropriate language, understandable feedback and boundaries for when adult help is needed. Parents and teachers still judge whether it is suitable for a particular child.
Children have directions of their own
The analogy has limits. Children are not collections of parameters to optimise, and growth is not copying model answers. Curiosity, choices, relationships and real-world experiences belong at the centre. A model should serve a child’s development.
From current activities to a future direction
Today, Little English offers vocabulary, phonics, prepared speaking activities and maths with English. A graded reading app is in preparation. These products do not already constitute a general AI teacher or a complete adaptive learning system. This essay describes a direction; released capabilities are described on each product page.
The idea we want to share
Let AI support teaching, and respect each child’s learning path.