DeepMind's CEO on AI-native kids (and what he didn't say)
Was this newsletter forwarded to you? Sign up to get it in your inbox
"The most productive people might be 10x more productive if they are native with these tools... Kids should immerse themselves in these new systems."
Coming from a Nobel laureate building AGI, that's worth listening to.
And there's something crucial I’d like to add:
Before kids can be 10x productive with AI, they need to know what they want to be productive at.
AI can help you build anything, but it can't tell you what's worth building. It can answer any question, but it can't tell you which questions matter to you.
I call this the direction problem.
The Sequence Matters
I first noticed this pattern 8 years ago when I introduced my kids to Brainpop, a platform built around exploration and discovery that became a major inspiration for what we're building at Kubrio today.
What changed wasn't that they learned more facts. It was something deeper:
They developed confidence they could figure out any topic, no matter how unfamiliar. They had the courage to jump into subjects they knew nothing about. Most importantly, they stopped seeing learning as something they did for someone else, for grades, for parents, for college applications, and started seeing it as something they did for themselves.
Over the years, we ran events where hundreds of education experts said variations of the same thing: kids who discover what pulls their attention before being pushed into specific skills end up going further in those skills.
The sequence isn't arbitrary: Direction → Tools → Mastery.
Skip the first step, and the other two never happen, or they happen slowly, painfully, and only for external rewards.
But most parents are starting at step 2 and wondering why kids just use AI to do their homework faster instead of building something that matters to them.
What Direction Actually Looks Like
Direction isn't a career choice. It's not "I want to be a doctor" or "I'm interested in coding."
Direction is knowing:
- What kinds of problems pull your attention
- What intellectual challenges you lean into, not away from
- What you're willing to struggle with because the struggle itself is interesting
- What questions keep surfacing in your mind
For adults, we call this "intrinsic motivation" or "intellectual curiosity." For kids, it's simpler: it's discovering what makes their brain light up.
And here's the problem: the legacy school model doesn't optimize for this discovery. It optimizes for standardization, compliance, and grade-maximization.
I had this conversation recently with a brilliant kid, fresh out of a prestigious university, did everything "right”, who shared something they couldn't tell their parents:
"I don't know what to do next."
Not in a "which job offer should I take" way. In a deeper, scarier way: I don't know what I want. I don't know what problems interest me. It feels like I never chose the race.
This was a brilliant, capable person who did exactly what the system asked, but the system never asked them to develop an internal compass.
Sitting with someone who crushed every external metric but had no internal direction is what convinced me to restart Kubrio.
Why This Matters More in an AI Era
In a world where AI handles execution, the only scarce resource is knowing what's worth executing.
The kids who will thrive aren't the ones who can prompt ChatGPT the fastest. They're the ones who have:
- Clarity about what problems interest them
- Experience exploring different intellectual domains
- Practice finishing projects that they chose
The "AI-native" part comes naturally once they have direction. Give a curious 12-year-old a clear goal and access to AI tools, and they'll figure out the tools in a weekend.
But give them the tools without the goal? They'll just use it like everyone else, to make school slightly easier.
A Different Frame for Parents
Instead of asking "How do I use AI to help my kid with math or homework?", try asking:
"How can I use AI to help my kid discover what they're passionate about?"
Because curiosity isn't passive. It's not "expose them to stuff and see what sticks." It's active exploration across domains, with tight feedback loops and real stakes.
Find ways to use AI to help them explore what makes their brain light up.
Some practical approaches:
- Projects over consumption: Building something small weekly beats watching educational content
- Cross-domain exposure: Music → design thinking → storytelling → science. Let them see patterns across fields
- Real audiences: Kids learn faster when making things for people who'll actually use them
- Finished > perfect: One completed project beats five abandoned masterpieces
The goal isn't to find their "passion" at age 9. It's to give them enough reps making decisions about their own learning that by 15, they have strong intuition about what they want to dive deep on.
Tools Second, Not First
Here's what we're testing at Kubrio: use AI to support discovery first, then go deeper on the tools themselves.
During the exploration phase (ages 6-13), AI acts as a guide—helping kids build projects faster, explore new domains, and get unstuck. But the focus isn't on learning AI. It's on discovering what intellectually stimulates them.
Once they have direction—once they know what problems interest them—then introduce them to the deeper mechanics: agents, different models, how these systems actually work.
Because at that point, AI isn't abstract. It's leverage for something they already care about.
They'll learn prompting and system design naturally because those skills are now in service of a goal that matters to them.
But the framework isn't proprietary. Any parent can apply it:
Direction (exploration + feedback loops) → Tools (when ready) → Mastery (through building)
This is why I shared Demis Hassabis's perspective at the start—because when a Nobel laureate building AGI says kids should be AI-native, parents need to pay attention.
Kids should become AI-native.
But here's what that actually means:
Being AI-native without direction is like being fluent in a language you have no interest in. The syntax is perfect, the grammar is flawless—but there's no passion to talk with anyone about it.
Direction + AI = 10x leverage on something that matters.No direction + AI = 10x efficiency at something you don't care about.
And here's the uncomfortable truth: This transformation is happening fast. Unfairly fast. The gap between kids who are introduced to these concepts early and those who aren't is about to become exponentially wider.
I don't have all the answers.
But I do know these questions matter more than most parents realize, and things are moving much faster than we're used to.
—Vlad