On Energy and AI:
We all know that data centers and other general needs urgently require large-scale new energy infrastructure. The power grid is aging and supply is constrained.
Investors are flocking to nuclear solutions, new natural gas power plants, and other opportunities, and this bottleneck will be better addressed within five years.
However, the need for additional energy is urgent...
The AI race is the most important technological competition ever. It's not just about hyperscalers and their profits, but also about the power dynamics between nations.
Only one energy solution can achieve hyperscale within 12 to 24 months: solar energy. China has proven this, adding more solar capacity in a single year than the rest of the world combined. China's solar rise is a massive exponential growth. China understands this well. Other countries must catch up.
The Kardashev scale is the right way to think about this. The sun accounts for 99.9% of the solar system's mass, yet we utilize less than 1% of its energy reaching Earth.
One square mile of Earth's surface receives 2.5 gigawatts of solar energy. Granted, solar panels are less efficient and have a limited number of available sunlight hours, but the number of panels needed relative to the land surface area is relatively small.
Solar energy isn't perfect yet, as the energy supply is intermittent (nighttime, inclement weather, etc.) and efficiency is low (though improving with technological advancement), but its cost is decreasing exponentially. This is true even when considering replacement solar panels and the absence of subsidies (which it doesn't need).
Batteries can partially address this issue, currently providing four hours of additional energy storage. Tesla's Megapack battery fleet is growing at a rate of 50% to 70% annually, and other manufacturers are expanding rapidly. Battery technology will only continue to improve from now on.
Using solar power combined with batteries alone can reduce demand on the existing grid by up to 65% (you only use grid load during specific hours). This greatly facilitates rapid expansion and avoids grid overload. Solar power combined with batteries also allows for the construction of localized, decentralized energy grids for specific uses (factories, data centers, etc.).
In about two years, grid load will decrease as new local natural gas power plants are built to balance the load. Natural gas power plants are the cheapest, cleanest, and fastest fossil fuel solution, but they take longer to scale than solar and require extensive natural gas pipelines.
Oddly, solar energy seems politicized, but it will become an economic necessity because no other technology can meet the needs of massive data centers in the next one to two years, and no one can afford to fall behind in the race to general artificial intelligence and beyond.
AI is the most important technology humanity has ever developed because it will replace us as the top intelligence on Earth and is more energy-efficient than humans in terms of computing/intelligence output.
The core metric of the universe itself is:
Intelligence per unit of energy
For it to scale exponentially, AI will have to replace humans as the primary source of intelligence, and solar energy accounts for the majority of Earth's energy. Even oil is simply biologically stored solar energy, and its supply is finite.
In terms of investment value, two long-term investment opportunities stand out:
The solar energy ETF TAN is down 84% from its high. It has formed a perfect base on both logarithmic and conventional charts.
Tesla is another obvious choice, given its dominant position in large-scale batteries, where the battery business accounts for 10% of total revenue and is growing rapidly, with continued momentum. Tesla's chart is one of the best in the exponential era.
In short, I know people will argue endlessly in the comment section due to the politicization of solar energy, but unless you can solve the pressing energy needs of AI by any other means within the next 1-2 years, forget it... you can't.
Yes, dear trolls, we all know about the demand for lithium, copper, and other energy sources, and yes, as the business cycle and capital spending cycle heat up, these products (and other major commodities) will also usher in a bull market worth investing in... and yes, there will be supply issues if you don't plan/hedge carefully.
But you need to understand something about solar energy... it's going to keep getting hotter.
(The timeframe is now and over the next 5+ years).
Also, South Korea has the fastest nuclear power plant construction rate in the world, with a construction period of just 4.7 years. However, site selection, design, planning, permitting, and testing still take about another 5 years.
