World News

China is rapidly surpassing the U.S. in nuclear energy, building more reactors at a faster pace and developing advanced technologies like small modular reactors and high-temperature gas-cooled units.

Okay, yes, very broadly-speaking, I agree that US nuclear power generation capability relative to China is something to keep an eye on. As well as ability to construct nuclear power generation capacity. There might be a way that China could leverage that in some scenario. However.

At least some of that is tied to population; China has over four times our population. One would expect energy usage per-capita to tend to converge. And for that to happen, China pretty much has to significantly outbuild the US in generation capacity.

If we in the US constrain ourselves to outpace China in expanding generation capacity, then we're constraining ourselves to have multiple times the per-capita energy generation capacity.

Now, okay, yes, there is usage that is decoupled from population size. AI stuff is in the news, and at least in theory -- if maybe not with today's systems, but somewhere along the road to AGI -- I can imagine productivity there becoming decoupled from population size. If you have more electrical generation capacity, you can make effective use of that electricity, convert it to productive capacity.

But a lot of it is going to be tied to population. Electrical heating and cooling. EV use. You'd have to have a staggering amount of datacenter or other non-tied-to-population power use to dominate that.

These statistics aren't from the same year, but they have a residential-industrial-commercial breakdown, and then a breakdown for each of those sectors.

https://www.eia.gov/energyexplained/electricity/use-of-electricity.php

Commercial use, residential use, and industrial use are, on that chart, each about a third of US electrical power consumption. Of the commercial category, computers and office equipment are 11%. So you're talking maybe 3% of total US power consumption going to the most critical thing that I can think of that represents productive capacity and is potentially decoupled from population. And that's all computer and office equipment use, not just stuff like AI. A lot of that is going to be tied to per capita usage, too.

About half of commercial use of electricity is space cooling. Almost everything else is either cooling, lighting, or ventilation. Those are gonna be tied to population when it comes to productive capacity.

If you look at residential stuff, about half of it is cooling, heating, or lighting, and my bet is that nothing in the residential category is going to massively increase productive capacity. Up until a point, on a per-capita basis, air conditioning increases productivity. Maybe it could provide an advantage in terms of quality of life, ability to attract immigration. But I don't think that if, tomorrow, China had twice our per-capita residential electrical power generation capacity, that it'd provide some enormous advantage. And it definitely seems like it'd all be per-capita stuff.

In industry, you have some big electricity consumers. Machinery, process heating and cooling, electrochemical processes. And with sufficient automation, the productive capacity of those can be decoupled from population size. Given enough electricity, you could run a vast array of, say, electric arc furnaces. But I think that "American industrial capacity vis-a-vis Chinese industrial capacity" is a whole different story, that it's probably better-examined at a finer-grained level, and I think that there are plenty of eyeballs already on that. Hypothetically, you could constrain residential or other use, pour power capacity dedicated to it into industrial capacity in a national emergency, but I can't think of any immediately-obvious area of industry where exploiting that is going to buy that much. Unless we expect some massively-important form of new heavy industry to emerge that is dependent upon massive use of electricity -- like, throw enough electricity into a machine and you can get unobtanium -- I'm probably not going to lose sleep over that.

If your concern is that there might be ways in which China can leverage its population and so per-capita statistics matter, then sure, I get that, but again, I think that that's probably better considered in terms of metrics of human capital rather than in terms of just energy generation capability. And I think that the constraining factors there, if you're talking ability to increase existing capacity in percentage terms, are probably (a) fertility rate, (b) immigration rate. I am pretty sure that if we wanted to get power capacity built and tied into the grid, it could be done on a shorter timescale than we could get people to have children and then raise those children and provide them with a necessary skillset, so I don't think that existing electrical generation capacity or ability to increase it in the short run is the bounding factor. Maybe we could do immigration at a higher rate than we could expand generation capacity, making electrical generation capacity the bounding factor, though there are -- looking at popular irritation that drove voters to support Trump -- some political limitations. The last time we were seriously looking at going balls-to-the-wall against another country was World War II versus principally Germany, and war plans included, after mobilizing large portions of the American population, hiring huge chunks of population out of Latin America to fill in the now-absent farm labor need in the US to keep US productive capacity ramping up. As World War II played out, Germany ultimately didn't conquer the UK and then initiated a fight with the Soviet Union, so a lot of the levers never needed to be pulled. We ultimately only used it in a considerably-scaled-down form.

https://en.wikipedia.org/wiki/Bracero_Program

The Bracero Program (from the Spanish term bracero [bɾaˈse.ɾo], meaning "manual laborer" or "one who works using his arms") was a U.S. Government-sponsored program that imported Mexican farm and railroad workers into the United States between the years 1942 and 1964.

The program, which was designed to fill agriculture shortages during World War II, offered employment contracts to 5 million braceros in 24 U.S. states. It was the largest guest worker program in U.S. history.[1]

But I think that that maybe provides insight into what the US would be willing to do in another situation where we wind up in a serious power struggle with another country. If we had to pull tens of millions of people from abroad into the US in short order in a balls-to-the-wall situation, I'm pretty sure that we would.

So, okay. Maybe, if you think that you can make use of extremely-high-rate immigration capacity, you might want to have a certain amount of electrical generation capacity available or ability to ramp it up very quickly.

However.

China could do the same to some degree. But it's also more-difficult for China due to her larger size relative to the pools abroad from which she might draw -- if she wants to scale production proportionally to her population -- and I suspect what China would be able to offer in terms of environment, if a contest were predicated on our respective abilities to draw labor from abroad.

So, in summary:

I'm not sure that I'd be concerned about "what China could do in the short run in terms of dramatically increasing her capabilities in a way that threatens the US if China had large amounts of electrical generation capacity in absolute terms, and then started a massive immigration program". I don't expect that that sort of contest would play to China's strengths.

In the sense that China could make use of more electricity to produce more industrial output, sure. China has significantly more steel production capacity today. That's not really new, and I would expect that it's been taken into account, that one doesn't expect steel production capacity to be some sort of bounding factor that's of special concern. Going back to World War II again, steel production over an extended period of time mattered there...but I'm skeptical that we'd find ourselves in some kind of sustained conflict with China where steel production capacity mattered. It's too easy to knock out steel production infrastructure or the like in 2025. Maybe someone could identify some kind of concern there, but I don't think that one would express it in terms of electricity. I don't think that there's some sort of way in which a country can translate steel into productivity in a peacetime environment to the degree that available steel is the limiting factor, either, where we'd say "Oh, no, China pulled ahead in steel capacity and steel is now mainly determining a country's economic or military strength, and we cannot catch up."

Having electrical capacity might matter if it's the bounding factor for something like AI, which potentially has productive capacity decoupled from the size of the labor pool. I think that keeping an eye on the critical resources governing AI capacity is going to be something to do moving forward in the years and decades to come. But as things exist today, usage there is a very small portion of electricity consumption. I don't think that we're looking at the limits imposed by electrical generation. Maybe if technology advances and we do enough buildout of capacity, that would change. But I think that we're also some ways away from electricity being a serious constraint there.