AI Tutors for Key Stage 3 is a full-length, illustrated book for UK parents of children aged 11–14 and for the students themselves. Written by Duke Harewood and published by Innovatorly Ltd in 2026, it comes with roughly forty line-art diagrams and no assumed technical knowledge. A Simplified Chinese edition — 《英国中学的AI私教》 — is available for Chinese-speaking families alongside the English text.
The book introduces eight subject tutors, explains the ideas behind them, and gives parents and students the practical knowledge to use them well. This guide walks through each of the six parts so you know what to expect before you begin.
Part I — The Trend
The opening section makes the case for why one-to-one tutoring has always outperformed classroom teaching, and why most edtech has failed to close that gap. Rather than citing fashionable claims, it traces the evidence back to Benjamin Bloom's classic research on mastery learning and to the Socratic tradition that has underpinned the best tutoring for two and a half thousand years.
Part I also introduces the one rule that every chapter returns to: a good tutor never gives the final answer. That single principle turns out to be surprisingly difficult to honour — for human tutors under time pressure, and for AI systems that are designed to be helpful. The rest of the book is, in many ways, an extended explanation of why that rule matters and how the eight tutors are built to keep it.
If you want to understand the thinking before meeting the tutors, start here. If you are already convinced and want to get started, you can skip ahead and return later.
Part II — Understanding the Machine
This is the AI-literacy section, written for readers who have never studied computer science. It explains what a large language model is, what it cannot do on its own, and why a harness of tools, memory, skills, hooks, and workflows is needed to turn a language model into a tutor with a consistent personality.
It covers the concepts behind agents, MCP (the protocol that lets Claude connect to the aitutors.me tutors), and slash commands, without asking you to read a line of code. The goal is to give parents enough understanding to have honest conversations with their children about what the tutor is and is not. It is not magic, it is not a search engine, and it is not a person — understanding the difference matters for using it well.
The section closes with an explanation of why each of the eight tutors has a distinct personality rather than being one generic assistant. Personality is not decoration; it is pedagogy.
Part III — The Faculty
Eight chapters, one per tutor. Each chapter introduces the tutor's subject, explains the particular challenges KS3 students face in that area, and describes how the tutor's approach is shaped by those challenges.
The eight are: Mentor (wellbeing and daily energy), Pi (mathematics), Quill (English language and literature), Darwin (biology), Curie (chemistry), Newton (physics), Harari (history), and Mercator (geography). Mentor is different from the subject tutors: its role is to check in with the student first, handle anything pastoral, and route to the right subject tutor. The chapter on Mentor explains why wellbeing comes before curriculum.
For subject tutors, each chapter includes a short annotated exchange — what the tutor says, what it is doing, and why it is doing it rather than simply giving the answer. These are among the most useful pages in the book for students who want to understand what is happening when they are being taught.
Part IV — The Methods
This section goes deeper into the pedagogy. It describes the four-level hint ladder in detail: the tutor begins with a question, then a nudge, then a partial example, and only as a last resort walks through a worked solution. Each rung is deliberate. The aim is to keep the student doing the cognitive work for as long as possible, because that work is the learning.
Part IV also explains the weekly energy system — the idea that a student's capacity to learn varies across the week and across the day, and that a good tutor adjusts to that rather than delivering the same session regardless of how the student is feeling. Alongside this is the mastery loop: how the tutors decide when a student has genuinely understood a concept rather than just answered one question correctly.
The final chapter in this section covers safeguarding. It explains the hard rule: if a student shows any sign of distress, the tutor stops and directs them to a parent or to Childline (0800 1111). This is not an optional behaviour — it is built into every session.
Part V — The Parent Playbook
Part V is addressed directly to parents. It takes the questions that come up most often — is this safe, how much is too much, what happens when the AI gets something wrong, should it replace a human tutor — and gives honest, practical answers without overselling.
The section on daily rhythm is particularly concrete. It suggests how to introduce the tutors to a child who is sceptical, how to notice early signs of over-reliance, and how to keep the relationship with a child's school teacher at the centre of their education rather than letting an AI tutor become a parallel system that the school never knows about.
There is also a chapter on the distinction between a friend and a tutor. Children sometimes form a strong attachment to a patient, consistent AI. The book does not treat this as a problem to be eliminated, but it does explain why maintaining the tutor's role — rather than letting it drift into general chat — is in the child's long-term interest.
Part VI — Evidence and Curriculum
The final section is reference material. It opens with a review of the educational research that informs the methods used throughout the book, keeping citations accessible rather than academic. This is followed by a full annotated tutoring session — a complete exchange from energy check to problem solved, with margin notes explaining each move.
The KS3 curriculum map covers the subjects handled by the eight tutors against the National Curriculum for England, so parents and students can see how the tutors relate to what is being taught in school. The section closes with a FAQ and an appendix written for educators who want to understand how AI tutoring fits into a classroom context rather than replacing it.
Where to find the book
Both editions are listed at aitutors.me/books. The companion service — where the eight tutors are live and available — is at aitutors.me.
If you are curious about Mentor specifically, the Meet the Mentor piece is a good place to start. For the thinking behind the core teaching approach, The Socratic Method in AI Tutoring explains the one rule in detail.