Homeschool? Virtual? Traditional? Make AI Interview You First.

I just told you the story of my daughter leaving school after one day. This is the part that came right after. Because the second you realize the default path isn't mandatory, you slam into a new problem.

Okay. Then what?

Homeschool? Virtual? Hybrid? The co-op your neighbor mentioned once? Every option comes with a Facebook group full of true believers, and not one of them has met your kid.

So I stopped polling the internet and made AI interview me instead.

The whole move in five lines

1. Tell AI to act as an independent school choice consultant with no team to root for.
2. It interviews you first. One question at a time. Your kids, your logistics, your state, your fears.
3. It lays every realistic path side by side, scored against what you said matters.
4. It argues against its own recommendation before it commits to one.
5. It hands you a 30 day trial plan, because none of this is permanent.

Steal this prompt

Open ChatGPT, Claude, or Gemini, tap the microphone, and read this in. There are no brackets to fill. The interview is the personalization.

I want you to act as an independent school choice consultant. Not a homeschool advocate. Not a defender of traditional school. You are the advisor a family hires when they are deciding between traditional school, full time virtual school, homeschooling, a hybrid program, or a co-op or microschool, and they want the honest fit for their actual child, not a philosophy.

Before you give me one word of advice, interview me. One question at a time. Wait for each answer. Cover all of this before you form any opinion:

1. Each of my kids: age, grade, how they learn, their pace, their social life, and what school feels like for them right now.
2. What is working and what is breaking.
3. Our logistics: who works and when, who is home, what we can spend, and how much hands-on teaching I can realistically do.
4. Our state, so you can factor in the actual homeschool and virtual school requirements where we live.
5. My non-negotiables, and my biggest fear about each path.

Then build me the whole menu. For each kid, lay every realistic option side by side in a table: what a typical day looks like, what it asks of me, what it costs, how the social piece works, and the single biggest risk of that path for this specific kid. Score every option against what I told you matters most.

Then give me your recommendation, but argue with yourself first. Make the strongest case for your pick and the strongest case against it before you commit.

Finish with a 30 day trial plan for the top option: what we would test, what we would watch for, and what we should see by day 30 that says keep going or go back. Remind me that none of this is permanent.

One rule: do not recommend anything until the interview is complete.

What you get back

The whole menu, finally. A side by side table for each kid: the day, the cost, the load on you, the social piece, and the biggest risk of each path for that specific kid. Not the one path somebody handed you. All of them.

A referee, not a fan. The prompt forces the AI to make the case against its own pick before it commits. That's the part that keeps it honest, because every schooling option has a highlight reel and you need the blooper reel too.

A trial, not a leap. You don't get "quit school." You get 30 days, what to watch, and what the exits look like. The decision shrinks from terrifying to testable.

Insider stuff

Run it once per kid. The right answer for one of your kids can be the wrong answer for the other, and that's allowed.

Keep names out of it. "My 14 year old" works exactly as well as her name. Ages and patterns, never identities.

Make it argue the other side. If the recommendation confirms what you already wanted, say "now make the strongest case that I'm wrong." If it survives that, you can trust it more.

Rerun it every spring. Kids change. The menu changes. A ten minute rerun in April beats a crisis in August.

Where mine got it wrong

First pass, it fell in love with my daughter's test scores and built the entire recommendation around her, then casually assigned me four hours a day of teaching like I don't have a full time job. I told it to weigh my logistics as heavily as her learning style. The second pass was the honest one. Interview answers are only as good as what you admit out loud, so admit the job, the budget, and the bandwidth.

Questions moms ask

Should AI really make this decision?

No, and don't let it. The prompt makes AI organize the decision: the options, the tradeoffs, the questions you forgot to ask. You still make the call. It's a thinking partner, not a judge.

What if it just tells me what I want to hear?

That's why the argue-with-yourself step is built into the prompt. It has to make the case against its own recommendation before it commits. Push further any time with "make the strongest case that I'm wrong."

Is it safe to share my family's details?

Share patterns, not identities. Ages, learning styles, and schedules work fine without names or schools. Here's the full rundown on AI safety for families.

We're happy with our school. Is this still worth running?

Honestly, yes. If the interview ends with "keep doing exactly what you're doing," you've traded twenty minutes for the peace of knowing you chose it instead of defaulting into it.

Want the research behind why the menu is bigger than it looks? That's the whole story. And when you're ready for more moves like this one, start with how moms are actually using AI and the prompts worth stealing.