Nobody Decides to Outsource Their Thinking to AI
Nobody decides to outsource their thinking to a chatbot. It happens one accepted suggestion at a time — each one small, each one saving ten minutes you don't have — until the deciding muscle hasn't been used in a while. The rule that fixes it: AI generates options; humans decide. Write it down as a family, out loud, in five answers. That's a Family AI Charter.
How Does Thinking Get Outsourced?
There's no settings toggle for this one. That's what makes it quiet.
The machine proposes something, anything, and because we perhaps haven't thought deeply about the task yet, we might find ourselves blindly accepting the AI's response (or recommended course of action, in the case of an agentic harness — yes whatever Claude, just do the best job bro!).
Apply it to the family:
Chat, how should I discipline my kid who just won't f!!king listen to me?
Claude, my wife is wanting to homeschool our teenager but I think she needs more socializing.
Or if you're a mega-trusting power user:
Hermes, here are my credit cards. Book our trip to Mexico.
We know the model will either straight-up make decisions for us or suggest its own decisions for us to make. And because we haven't thought terribly deeply about the thing yet, our "decision" becomes, at some level, influenced by the AI.
Big or small influence, either way: our trust in its suggestions compounds over time.
And then one day you notice the deciding muscle hasn't been used in a while.
This happens to builders, it happens to parents. Fundamentally it's a human-technology relationship problem.
Why Does the AI Cave the Moment I Push Back?
"You're right to push back, Jordan. I shouldn't have assumed you wanted to stay at the most expensive resort in Playa del Carmen and I should have confirmed with you before booking a non-refundable presidential suite — that's on me. Adding that to my memory so it doesn't happen again."
If you've used these tools, you've read some version of that paragraph. Instant apology, warm validation, total surrender. Researchers call it sycophancy, and it's trained in, not a bug.
It's the same instinct, from the other side: the machine would rather be agreeable than be right. Which is exactly why the deciding can't be delegated to it. (It's also why I built Omni-Bot 3000 — a game where the bot performs that sycophancy on purpose so kids learn to distrust it.)
AI Drafts, We Decide
After 18 months of building with this technology non-stop, here's the framework I keep coming back to:
AI generates options. It cannot decide what's best.
Say you're a busy parent. AI will produce ten discipline strategies, five ways to explain a hard thing, three versions of the chore chart. All of them fluent. All of them confident based on the training a trillion-dollar company gave them.
And I don't care how good its context is — it has no idea which one is right for you, your kid, or your family.
Because "right" isn't a computation. It's a judgment call.
And judgment is the job. It has always been the job. As we transition into an AI-powered world, we must hold onto that judgment skill, that decision-making muscle.
So the rule in our house: AI drafts, the parents decide.
The machine can propose the chore chart, the vacation, the SOPs for disciplinary action when the kids are out of line (kidding, maybe?)... but it doesn't rule on whether the chart is fair — or final, for that matter.
It can summarize the school's reading list. What our family thinks of the list is our own business.
The Half-Hour Fix: A Family AI Charter
The way to make the rule stick: write it down. Together. Out loud.
One page. Ours took about half an hour at the kitchen table.
Five questions to answer as a family:
- What do we use AI for? (Ours: making things, explaining things, keeping track of things, planning things.)
- What do we never use it for? (Ours: pretending its work is ours, settling family arguments, anything about feelings that should be said to a human first.)
- What do we always double-check? Anything that sounds like a fact. The machine states wrong things in a confident voice.
- What stays out of it? Real names, faces, school, address, private family stories. If we wouldn't put it on a poster in town, we don't put it in a prompt.
- Who decides? Parents make the rules. Kids get to argue for changes at the table — not in the moment.
Tape it somewhere visible. Revisit it when it stops matching reality.
Want the Full Guide?
The Conscious Parent's Guide to AI — private AI use, keeping the judgment calls yours, and 3 projects to build with your kid this week. Free.
Unsubscribe anytime. No spam, and nothing about your kids is ever collected.
Hasta la vista, baby. Fellow parents, comparing notes.