I get asked this a lot, "Are you against AI?" No. I'm not totally against it. I'm also not for it just because it's new technology with much capabilities. I think most of us are stuck picking a side when we don't actually need to.
AI is here. Our kids are going to grow up and use it, whether we say yes or no. The real work is figuring out how much, when, and for what. So let me just walk through what I actually think, plainly, without picking a team.
What AI is doing to how kids learn
Here's the honest truth. A child can now ask a chatbot to solve their math problem, write their essay, or explain a concept, and get an answer in seconds. That's powerful. It's also a shortcut that skips the part where learning actually happens.
Learning happens in the struggle. The part where a child sits with a problem they don't understand yet, tries something, gets it wrong, and figures out why. If AI removes that struggle every time, the child gets an answer but doesn't build the thinking muscle that produces answers on their own next time.
I'm not saying AI makes kids dumb. Though it does it to an extent. But what I'm saying is it removes friction, and friction is where the learning was happening. If we hand kids a tool that removes all the friction, we have to be intentional about putting some kind of thinking work back in, on purpose.
The confidence trap: why AI sounds right even when it's wrong
One thing about AI that catches children and adults off guard is how confident it sounds. It doesn't say "maybe" or "I think." It answers in full sentences, with examples, and acts like the answer is settled. The problem is, confidence is not the same thing as correctness. AI can be wrong about facts, wrong about math, and wrong about history, while sounding completely sure of itself.
For a child who is still learning how to think critically, this is a big deal. If the machine always looks certain, why would they question it? But that is exactly the habit they need: stopping to ask, "Does this actually make sense?" before trusting the answer. Not all answers should be accepted, especially in this AI era. Fact-checking is not extra work; it is the main skill now.
This is why I don't think the solution is to keep AI away from kids. The solution is to teach them that speed and confidence are not proof. Real understanding means slowing down, comparing the answer to what you already know, and being willing to say, "I don't think this is right." That is the muscle Glitch Detectives is built to exercise.
What it's doing to their wellbeing
This part worries me more than the academic side. Kids are starting to lean on AI not just for homework, but for company. To talk to. To ask "what should I do" about things a parent, a sibling, or a friend used to be asked. That's a quiet shift and it's easy to miss because it looks harmless in the moment.
There's also the blank page problem. I'll get to that in a second because it's actually the reason I started building something for it. But the short version is, kids are starting to feel like they can't start anything without AI. Not because they're lazy, but because the habit of sitting with nothing and creating something is fading, and it's atrophying fast when AI is one tap away every time.
And there's comparison and speed anxiety too. When a machine can produce something polished in three seconds, a child's slower, messier first attempt can start to feel like it isn't worth making at all. That's the feeling we have to protect them from.
What it's doing to the environment
I don't think most parents think about this part, but it's worth knowing. Every time a chatbot answers a question, it runs on a data center somewhere, and those data centers use real electricity and real water to keep running and cool. Some estimates put a single AI chat exchange at using a small cup's worth of water for cooling and about as much electricity as a few minutes of a lightbulb being on, and that scales up fast across millions of people using it every day.
I'm not saying don't let your kid use AI because of the environment. I'm saying it's not a free, invisible resource. It has a cost somewhere, even if we never see it. Worth knowing, worth teaching your kids too, honestly.
So should homeschool parents use it or not?
My honest answer: use it if you wish, but with limits you set on purpose, not limits you figure out after something's already gone wrong.
Here's how I'd think about moderation, in plain terms:
- Let AI be the second step, not the first. Kid gets stuck on a problem, they try it themselves first, mess it up, then you bring in AI to check it or explain where it went wrong. Not the other way around.
- Set specific jobs for it, not open access. "You can use it to check your answer" is different from "you can use it whenever." Kids do better with a clear boundary than a vague one.
- Sit with them when they use it, especially at first. Not hovering forever, just enough to see what it's actually doing for them and how they're leaning on it.
- Watch for the avoidance pattern. If your child reaches for AI the second something feels hard, before they've even tried, that's the moment to pull back and have them sit with the hard part a little longer.
- Talk about it like a tool, not a friend. It can be useful. It's not a person. Kids need to hear that distinction said out loud, more than once.
None of this means banning it. It means being the adult in the room instead of letting the tool set the rules by default.
Why I'm building Glitch Detectives
This is the part that ties it all together for me. A lot of what I described above, the blank page fear, the reaching for AI before even trying, comes from one place: kids don't trust their own first attempt anymore. Staring at a blank page and being expected to produce something from nothing feels scary, so the easiest move is to skip it and let AI produce the something instead.
Glitch Detectives works differently. Instead of handing a child a blank page, we hand them something that's already there, a piece of math reasoning with a mistake hiding in it, and ask them to find it. That's a much less scary place to start than "create something from nothing."
And it happens to build the exact skill this AI era actually rewards: knowing when an answer, human or machine, is wrong, and being able to explain why. Kids who grow up practicing that don't need to fear AI, and they don't need to lean on it blindly either. They'll be the ones who can actually use it well, because they'll know how to catch it when it's wrong.
That's not picking a side on AI. That's just raising a kid who can think for themselves either way.