Ten Questions to Ask Before Enrolling in Any AI Program

Quick answer

Before you enrol your teenager in any AI program, ask the provider ten questions: what will my child be able to do at the end, do you teach tools or judgement, how much time is spent building, how do you teach students to check AI output, who teaches, how big are the groups, how does feedback work, who is this program wrong for, what does it cost all-in, and what can past students show for it. A good provider answers every one specifically, with numbers and examples. A weak one retreats into buzzwords. This is the enrolment-call script, with what a good answer sounds like for each question, so you can put any provider on the spot. Including us.

Why the phone call beats the brochure

AI education is a young market with no regulator checking the claims. Anyone can put AI in a course title, and plenty have. The brochure was written by whoever writes brochures; the person on the enrolment call has to answer follow-up questions in real time, and that is where thin programs come apart.

You are not being difficult by asking. You are about to hand over money and, more importantly, a term of your child's time and confidence. Ten minutes of pointed questions is the cheapest due diligence you will ever do. For the wider picture of what quality looks like in this market, start with AI education for teenagers in Australia.

Questions one to four: what your child will actually learn

1. What will my child be able to do at the end that they cannot do now? A good answer names capabilities and artefacts: they will have built a working project, presented it to a real audience, and be able to direct an AI tool on a problem they have never seen. A weak answer lists topics covered. Coverage is not capability.

2. Do you teach tools, or judgement? Tools change monthly. A good answer puts judgement first - knowing when to use AI, how to evaluate what it produces, where the ethical lines sit - with specific tools as the vehicle rather than the destination. If the syllabus reads like a parade of app names, keep looking. The sensible ordering is laid out in what teenagers should learn first in AI.

3. How much of the time is my child building, versus watching? You want a ratio, and you want building to win. Watching demonstrations feels like progress and produces nothing your child can keep.

4. How do you teach students to check AI output? This is the question most providers have never been asked, which makes it useful. A good answer describes explicit habits: verifying claims against real sources, probing what the model gets wrong, learning that these systems state falsehoods with total confidence. If the answer is a vague nod to critical thinking, ask how, and watch what happens.

Questions five to seven: who teaches, and with whom

5. Who teaches this, and what have they built? You are not looking for celebrity. You are looking for instructors who have used these tools for real work and can field a sharp Year 10 student's question honestly, including with an occasional "I don't know, let's find out."

6. How big are the groups? A number, please. Feedback does not survive a group of forty. Small cohorts are expensive to run, which is exactly why this answer tells you something about where the provider spends its money.

7. How does my child get feedback on their work? A good answer describes structured critique: work shown to instructors and peers, specific suggestions, revision expected. A weak answer describes a certificate.

Questions eight to ten: fit, money and proof

8. Who is this program wrong for? Honest providers answer instantly, because they think about fit all the time. A provider who says it suits everyone is telling you the intake filter is a payment form.

9. What does it cost all-in, and what happens if it is not working? You want a published price with no surprise extras, and a straight answer about what happens if your child wants out after week two. For context on what different price bands actually buy, see what AI education costs in Australia.

10. What can past students show for it? Not testimonials. Work. Ask to see projects, showcase presentations or portfolios from previous cohorts. If nothing exists that a stranger can look at, the outcomes probably do not either.

Red flags at a glance

How Edison answers them, since you will ask

Fair is fair. Our entry program, the Generalist AI Bootcamp, is open entry for ages 13 to 18, runs four or eight weeks in Sydney, Melbourne or online, keeps cohorts to 12 to 16 students, and ends with a showcase where every student presents what they built. Fees are published: $2,400 for four weeks, $4,500 for eight. The flagship year is selective, with a four-step admissions pathway that does turn students away.

And question eight, honestly answered: the Bootcamp is wrong for a teenager who wants to watch content quietly and collect a certificate. Everything in it points toward building and presenting, and a student who dreads both will not enjoy it.

The recommendation: print the ten questions, ask all of them on the enrolment call, and choose the provider whose answers are the most specific, not the most soothing. If nobody you speak to passes cleanly, a short open-entry course is a low-risk way to test one provider's actual teaching before committing to anything longer.