Start Free for 14 Days Under $7/mo billed annually Cancel anytime

AI Isn't Good or Bad for Kids and Parents Need a Plan

April 29, 2026

The False Binary

Parents right now are given an impossible choice: give their kids access to AI and hope for the best OR prohibit access and hope for the best. The social media echo-chamber I am currently living in seems to be organizing itself into two neat camps on opposite sides of the AI opinion spectrum. On one end, there are those who believe this is all hype and fleeting and nothing to concern themselves with. On the other are those who have stopped using sunscreen because they believe AI will cure their cancer when they get it.1

Ethan Mollick LinkedIn Post [h/t to Ethan Mollick]

There’s plenty of evidence available to both camps to support their claims. How can this be, though? How can it be that two groups at opposite ends of the argument are both simultaneously correct (and, if we flip it, both simultaneously wrong)? And if you are trying to figure this out for you and your family, what the hell do you do?

The Jagged Frontier (and Why It Matters at Home)

Back to Ethan Mollick, whose AI prompt created the cartoon above… He calls this phenomenon the “Jagged Frontier”. What it means is that generative AI capability is better in some areas than it is in others. It is actually human-level (maybe better) in some domains, and not in others. Even more confusing, because these are complex machines, things can veer on and off the tracks in a single conversation.

The easy thing is to stick your head in the sand. The hard thing is to understand the nuance and come up with a policy that is unique to you, your parenting style, and your family.

Jake's New Yorker Cartoon [my mediocre attempt at creating a similar image using ChatGPT]

Here are some facts:

  • AI is here to stay and getting more powerful. It has actually been here for a long time; what we call “AI” now is actually just a new, more capable flavor of a thing we’ve had for decades. What’s the next thing? I don’t know… but we are 100% not putting this thing back in the proverbial box.
  • Humans are not obsolete. While AI has made me much more productive and capable, it is still nothing without me. It didn’t have the idea for my business or the cartoon that inspired this post (it is not a replacement for creativity). I still review [almost]2 everything it creates. I own its output and I am the reason it outputs anything at all. When we allow it to do everything for us, we start to get very homogeneous output. (See your LinkedIn feed right now.)
  • AI is an amazing and powerful tool that truly requires hands-on experience to understand well. You need to try it first in low-stakes ways; need to learn it to get the most benefit from it.

What This Means for Your Kid

To say it is only a tool, though, is to be a bit glib with the concerns. Other tools, like knives and guns, are dangerous, too, but in an easily identifiable way: sharp edges, loud bangs, and clear injuries from mistakes. AI chatbots, which are always available, overly agreeable, and very capable of mimicking human emotions, are dangerous, too, just not as obviously dangerous. This combination of capability, a child’s developing mind, and a lack of parental oversight can be catastrophic.

This is the gap I’m trying to close with MyDD.ai. Kids get access. You get visibility into what they’re actually saying. See how it works.

So. Here’s the complicated truth. You and your child are now in a world that demands they understand AI in order to be successful in life. Simultaneously, we need to teach both parents and kids the dangers and how to use these tools effectively. We’d never just hand a knife to our child; we’d show them how to use it, supervise, then, gradually, allow them independence. The alternative, cutting our children’s chicken for the rest of their lives, is not a viable option either.

In that vein, we must give access to AI so our children can learn. We must show them how to use it properly. And we must keep an eye on how they are using it. If we don’t all learn that way, we’ll keep paying the price.

This is exactly why I’m building MyDD.ai… to give parents the access AI requires while keeping eyes on what their kids are actually doing. If you want to see it in action, try it free for 14 days, then under $7/month billed annually.


Footnotes

  1. Heard this anecdote on a recent episode of the Hard Fork podcast.

  2. I used Claude to create supplemental exercises for my class, which it did well, and I sent out after a full review 3 times in a row. After that, and in a rush, I did it a fourth time, Claude created a really bad assignment, which I didn’t review, and I sent it out. After it thoroughly confused my students, I had to come clean and take responsibility. Lesson learned (again) and I won’t be doing that anytime soon.