China, AI Kids, and the New Power Race
China just made “AI class” mandatory in school. Not in college. From grade one.
Beijing has ordered every primary and secondary school student to receive at least eight hours of AI lessons per year starting in 2025, as part of a national push to build AI talent across the entire population.
On top of that, the Ministry of Education has laid out a plan to make AI education universal in primary and secondary schools by 2030, and fully embedded in textbooks and exams by 2035.
That is not a “maybe this AI thing is real” posture.
That is “we are raising a generation of power users” as national strategy.
And here is the part most investors miss:
You cannot turn an entire country into AI natives without a grid that can feed them.
China is building the grid for the world they see coming
While the U.S. has been debating permits and fighting over transmission lines, China has been quietly overbuilding its power system on purpose.
In 2022 alone, China invested about $166 billion in its transmission grid, more than the rest of the world combined. (Energy Monitor)
In the second half of the 2010s, China built 80 times more high-voltage transmission than the U.S., and in the 2020s has completed over 8,200 miles of ultra-high-voltage lines, compared with roughly 375 miles in the U.S. (Energy Platform News)
Experts now describe China’s “quiet electricity dominance” as the result of decades of deliberate overbuilding. Estimates suggest reserve margins in many regions are 80–100%, meaning the grid often has roughly double the capacity it needs, which lets it absorb the surge from hundreds of new data centers. (Tom's Hardware)
In other words, China is preparing for AI by building the power system first and then training the population to lean into it.
Taken together, the message to investors is clear. AI is being positioned as a core infrastructure layer, with energy capacity built ahead of demand.
The U.S. is doing the opposite
Now contrast that with the U.S. picture.
The International Energy Agency projects that global electricity demand from data centers will more than double to around 945 TWh by 2030, roughly equal to Japan’s entire current electricity consumption. AI-heavy data centers are expected to be the biggest driver of that growth. (IEA)
In the U.S. specifically, data centers are on track to account for nearly half of all electricity demand growth through 2030. (Reuters)
Meanwhile, our grid has been patched instead of rebuilt. In the last email, we looked at PG&E and the Camp Fire as an extreme case of what neglect looks like when a for-profit monopoly is in charge: shareholder returns stay protected, and the cost of failure lands on homeowners and communities.
Now layer AI on top of that same structure.
Utilities are already warning they need massive new investment to handle data center loads, EVs, and electrification. The capital they spend goes into the rate base. Their regulated return is built on that spend. And in practice, the people who ultimately fund that build-out are not the AI companies. They are the ratepayers.
So ask yourself: when AI demand forces billions in new grid upgrades in an already fragile system, who is going to pay for it?
Exactly the same people who paid for wildfire settlements and deferred maintenance: homeowners and small businesses.
Even the AI guys are saying it out loud
Sam Altman, CEO of OpenAI, has been unusually blunt about this.
At Davos, he said that AI will consume “vastly more power than people have expected” and that an energy breakthrough is necessary for the future of AI. He pointed directly to nuclear fusion and cheaper solar plus storage as the way forward. (Reuters)
From someone in Altman’s position, the takeaway is hard to ignore. The constraint on AI’s growth is power availability.
China’s response has been: build more grid, build more clean power, train more AI-literate citizens.
The U.S. response so far has been: send more interconnection requests to an already strained grid and hope utilities can catch up in time.
For markets, this split is consequential. It defines the contours of opportunity and risk over the next ten years.
Why this matters if you care about renewables and real assets
China making AI class mandatory is a signal. It says AI is not a passing fad. It is a foundational technology, important enough to be treated like math and language.
When you combine that with:
A global forecast that AI and data centers will double electricity demand in just a few years, and
A U.S. grid model where investor-owned utilities pass most big costs through to customers,
You get a simple conclusion:
Energy is about to become one of the defining investment stories of the AI era, and ordinary households are going to feel the pain first.
That is where renewables, storage, and especially distributed systems come in. They are not just a “green” play. They are a resilience play. They are a way for end users to claw back a little control in a world where someone else controls the wires and the bill.
The question is which companies in this space are actually structured to benefit from that shift instead of being crushed by it.
Want to see how we’re thinking about it?
In our upcoming live investor briefing, GigaWatt’s CEO Deep Patel will connect these dots directly:
What China’s AI-and-grid strategy is signaling about the future
How AI demand interacts with a neglected U.S. grid and investor-owned utility incentives
Where decentralized solar and storage fit into that picture, especially after the ITC phase-out
Deep has spent 19 years building in this industry, across multiple cycles, while advising policymakers at the local, state, and federal levels, including engagement with the Obama administration. He now leads the Los Angeles chapter of the California Solar & Storage Association, which gives him a front row seat to how this transition is unfolding in real time.
If you want more than buzzwords and hype, and you actually want to understand how AI, geopolitics, and the grid intersect in the real world, this session is worth your time.
