The AI arms race begins at age 4

Key points:

In a brightly lit preschool classroom in Shenzhen, 4-year-olds are gathered around an AI-enabled robot named Doubao. With practiced ease, they issue voice commands, identify image patterns, and experiment with rudimentary machine learning games. These children are not simply digital natives. They are something newer–and potentially far more consequential: AI natives.

In China, the state is systematically rewiring its education system to raise a generation fluent in artificial intelligence. Beginning in kindergarten, children are exposed to age-appropriate AI tools, taught how to interact with large language models, and trained to think computationally in ways designed to mimic how AI “thinks.” In pilot programs rolling out this year, primary and secondary students in Beijing will receive a minimum of eight hours of AI instruction per academic year, building cumulatively over time.

The goal, articulated by China’s Minister of Education Huai Jinpeng, is sweeping: to integrate AI into every layer of learning and to create a national workforce primed not only to use AI, but to lead it. In Huai’s words, AI is the “golden key” to the nation’s educational transformation. A forthcoming white paper will cement this policy for the rest of the country, formalizing a framework to ensure China’s global AI leadership by 2030.

Meanwhile, in the United States, AI in K-12 classrooms is more often seen as a threat than a tool.

A 2023 Pew Research Center survey found that only 6 percent of U.S. public school teachers believed AI does more good than harm in education. Roughly one-quarter said it does more harm than good. Many districts have responded by restricting its use entirely. Rather than preparing children to thrive in an AI-driven world, the prevailing American impulse is to fence it off–especially in schools.

This divergence could prove existential.

Learning AI like a language

Neuroscientists have long known that language acquisition is far more effective in early childhood. A child who starts speaking French at age 4 may grow up without an accent. An adult who begins learning at 30 won’t. The same cognitive principles apply to AI.

“Kids who start young will develop intuitive fluency,” says Weipeng Yang, a researcher in AI and early childhood education. “They won’t just know how to use AI tools–they’ll understand how AI thinks.” Pilot programs in mainland China already show children as young as 4 successfully interacting with conversational agents, story-generating apps, and sensor-based robots.

Experts like Yang compare early AI literacy to musical improvisation: You need to start early to develop “automaticity”–the ability to make split-second decisions without conscious effort. “It’s jazz, not classical,” explains one educational technologist. “AI fluency is improvisational, intuitive, and cognitive. It’s not about memorizing facts–it’s about navigating ambiguity.”

The strategic divide

To some American observers, China’s advantage is not just technical. It’s cultural–and systemic. China sees education as a strategic asset in the AI arms race.  

When it comes to AI literacy, the U.S. trails its global peers. South Korea and Singapore have already begun integrating AI across grade levels, training teachers en masse, and building AI-customized learning platforms. Finland offers free national AI courses for all citizens. In contrast, most U.S. AI education remains confined to pilot grants, ad hoc workshops, or optional electives. 

One independent-minded 10-year-old U.S. student (full disclosure–she’s my granddaughter) said, “On a lot of my assignments I do use AI even though it’s not allowed. I do it because I know it’s the future. Isn’t school supposed to prepare you for your future?”

New Yorker Liz Ngonzi, founder of The International Social Impact Institute, warns: “This isn’t just a digital divide–it’s a digital chasm. Every month a student isn’t on board, they fall a year behind.” She likens current resistance to early internet fears, arguing that lack of AI literacy risks not only economic stagnation, but societal instability.

The consequences of inaction go beyond education. The nations that lead in AI will dominate in economic productivity, cybersecurity, and military innovation. If China raises a generation that thinks in algorithms and neural nets, while the U.S. raises one that fears them, the geopolitical implications are stark.

And this isn’t speculation–in China, it’s policy. China’s 2017 “New Generation AI Development Plan” explicitly names talent cultivation as key to its global ambitions. The education system is its primary tool. The U.S., by contrast, lacks a national AI curriculum, and remains mired in debate about whether students should even use ChatGPT to write essays.

If AI is the new literacy–and the foundation of the Fourth Industrial Revolution–the future may be written by those who learned it first.