Do-it-yourself game plans

What is the simplest responsible way to do this yourself?

Define one learning goal, check the proposed mentor and tools, buy a small trial, then commission a bounded project with evidence of independent learning. Do not begin with an annual package, a public application or a promise of an impressive portfolio.

This guide concerns a human teaching AI, not a chatbot marketed as an “AI tutor.” Its recommended purchasing sequence is:

  1. Agree a learner brief and spending ceiling.
  2. Source four to six plausible candidates or providers.
  3. Resolve safeguarding, identity, tool eligibility and data-handling questions.
  4. Interview the actual proposed mentor and observe a short teaching trial.
  5. Compare total costs and sign a limited agreement.
  6. Run a six-session project with a contribution record.
  7. Review what the child can explain, change and reproduce before renewing.

Suggested planning allowance: three to five calendar weeks and approximately 12–19 combined adult working hours to reach an informed selection. Follow this with six weekly project sessions and a short follow-up check. These are desk-designed planning estimates—not measured market averages, supplier turnaround promises or validated educational dosages.

The objective is broader than using fashionable software. UNESCO’s student AI framework covers human-centred thinking, ethics, technical applications and system design, with progression from understanding to application and creation. That is a useful corrective to a programme consisting mainly of prompts and demonstrations. (unesco.org)

What does the evidence establish—and what does it not?

Stanford’s November 2026 evidence review identifies promising educational uses of generative AI but also substantial uncertainty. Much of the research concerns postsecondary learners, and the authors explicitly caution against generalising results across tools, populations and settings. It does not establish that a particular premium private mentor will improve your child’s capabilities. (scale.stanford.edu)

Research method: this is public-page desk research, not mystery shopping, a safeguarding audit or a review of confidential supplier files. Prices are published examples, not personalised quotations. Unless attributed to a source, the schedules, worksheets, interview exercises and decision rules below are this desk’s proposed operating method, not a certified assessment instrument.

Sources cited on this page

  1. AI competency framework for students | UNESCO ↗unesco.org
  2. The Evidence Base on AI in K-12: A November 2026 Review ↗scale.stanford.edu