The AI study buddy any parent can set up in 15 minutes

A practical way to turn free AI tools into a calm, structured homework and exam-prep routine at home.

Article link: https://open.substack.com/pub/johanosteyn/p/the-ai-study-buddy-any-parent-can?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Audio summary: https://youtu.be/PHRA_DlFEDA

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

Over the past year, I’ve been using large language models such as ChatGPT, Google Gemini and Google’s NotebookLM to help my 12-year-old son work through homework and prepare for exams. Not to do the work for him. Not to “generate answers”. But to create something many parents wish they could afford: a calm, patient tutor that can explain a concept five different ways, then ask questions until understanding clicks. 

Most of these tools have free tiers, and the biggest barrier is not money or technical skill. It is knowing how to set the rules. The trick is a master prompt that forces a Socratic style: teach in small chunks, ask one question at a time, wait for the child’s answer, then guide them closer instead of revealing the solution.

CONTEXT AND BACKGROUND

Children are already using AI tools for schoolwork, often without adults fully realising how frequently, and why. Recent reporting has highlighted how normal it has become for kids to turn to chatbots for homework support, explanations, and even companionship. That reality should change the conversation at home. The choice is no longer “AI or no AI”. It is whether your child uses it randomly and privately, or with structure, oversight, and clear boundaries.

At the same time, the tools are getting better at working with your child’s actual study material, not just the open internet. Notebook-based systems can be particularly useful because you can load a worksheet, notes, or a textbook extract and keep the AI grounded in those sources. This matters because AI can be confidently wrong, and “grounding” reduces that risk. As NotebookLM expands its study-focused capabilities, including richer ways to work with large sets of source material, it’s becoming more practical for learning workflows at home.

INSIGHT AND ANALYSIS

Here’s the key idea: you don’t need an AI that is clever. You need an AI that is disciplined. Without guardrails, children will naturally ask for answers. That’s not because they are dishonest; it’s because they are under pressure and they want relief. So your first job as a parent is to design the interaction so that learning is the default outcome.

My master prompt does three things. First, it instructs the AI to analyse the study material and produce a short, structured lesson in plain language. Secondly, it forces a Socratic sequence: “Ask one question, wait for my child’s reply, then respond with a hint and ask the next question.” Thirdly, it sets rules for integrity: “Do not write the final answer. Do not complete homework submissions. Focus on understanding and exam practice.”

Voice is the second accelerator. Many children think out loud far better than they type. When your child answers verbally, you can hear uncertainty, guesswork, and misconceptions immediately. It also makes the interaction feel like tutoring rather than Googling.

The third accelerator is building practice materials from the child’s own work. Google has been adding education-friendly features, such as automatically generating flashcards and study guides from class materials in Gemini for Education environments. You don’t need every feature to benefit from the idea: turn notes into retrieval practice. Learning sticks when children recall, explain, and apply, not when they re-read.

IMPLICATIONS

For parents, the path forward is surprisingly simple. Start with one subject and one worksheet. Set up your master prompt once. Then run 15-minute sessions: explain, question, correct, repeat. Keep it light, consistent, and honest. If the child gets frustrated, shorten the session rather than escalating. Small wins compound.

For schools and policymakers, the shift is cultural as much as technical. Blanket bans ignore reality, while laissez-faire adoption invites chaos. Guidance is increasingly focusing on responsible use and child protection, including the risks of over-reliance and the importance of supervision. The question is not whether children will encounter AI, but whether adults will equip them to use it wisely.

For children, the long-term benefit is bigger than marks. When used properly, AI can teach a child how to learn: how to ask better questions, check sources, spot mistakes, and persist through confusion without shame. But it must remain a tool, not a substitute for thinking. As more families introduce AI into daily life, experts are also warning about blurred boundaries and emotional attachment, which is another reason to keep the focus firmly on learning tasks and healthy limits.

CLOSING TAKEAWAY

The most powerful thing about AI for education is not automation. It is access to guided practice, at home, for millions of families who will never have a private tutor on call. But it only works when parents set the tone: curiosity over shortcuts, understanding over answers, and integrity over convenience. If we get this right, we don’t just help children pass exams. We help them build the habits they’ll need for a world where AI is always present: judgment, accountability, and the confidence to think for themselves.

Author Bio: Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net