Will electricity demand rise with AI?

There is one truth that has followed every major technological revolution in human history.

Energy demand always rises to meet technological capability.

When we industrialized, coal consumption exploded.

When we built the modern transportation system, oil demand reshaped global geopolitics.

When we entered the digital age, electricity quietly became the backbone of the global economy.

And now we are entering the AI era.

What most people don’t appreciate is that AI is not just a software revolution. It is an electricity revolution.

Training a single advanced AI model can consume as much electricity as tens of thousands of homes use in an entire year. And once trained, these models continue to run inside data centers filled with specialized hardware operating 24 hours a day.

A single large AI data center can require over 1 gigawatt of power. To put that into perspective, that’s enough electricity to power roughly 700,000 homes. One building consuming the equivalent of a major city.

Now consider that companies like Microsoft, Google, Meta, and Amazon are planning dozens of these facilities.

Suddenly, you begin to see the scale of what’s happening.

Even individual AI queries consume more power than traditional computing tasks meaningfully. One estimate suggests an AI query can use roughly 10 times the electricity of a traditional search query. That difference seems trivial until you multiply it by billions of interactions per day.

This is why, for the first time in decades, electricity demand in the United States is accelerating again.

For nearly 20 years, electricity demand was relatively flat. Efficiency gains offset economic growth. But AI, electrification of transportation, and domestic manufacturing are reversing that trend.

And here’s where the story becomes even more interesting.

China understands this.

China is building power infrastructure at a pace that is difficult to comprehend. They are adding entire national-scale power capacity every few years. In 2023 alone, China added more new coal power capacity than the rest of the world combined. At the same time, they are installing solar and wind at record rates, becoming the global leader in renewable deployment.

They are not choosing one energy source.

They are choosing all of them.

Because they understand that energy availability determines technological leadership.

Meanwhile, in the United States, building new power plants and transmission infrastructure can take a decade or more due to regulatory hurdles, permitting delays, and political resistance.

This creates a very real risk.

The country that can generate the most reliable, scalable energy will have a structural advantage in AI, manufacturing, and economic growth.

Energy is becoming the limiting factor.

And whenever something becomes a bottleneck, investment opportunities emerge.

We are entering a period where trillions of dollars will be spent on power generation, grid modernization, nuclear energy, solar, battery storage, geothermal, and technologies that most people have never even heard of.

Some of the biggest fortunes of the next decade will likely be tied directly or indirectly to solving this energy constraint.

In today’s episode, we explore alternative energy sources, the challenges we face, and the technologies that may power the future.

Because understanding energy is no longer optional if you want to understand where the world is going.

And as investors, those who see these shifts early have the opportunity to position themselves ahead of the crowd.

Watch on YouTube:

https://youtu.be/D0Lpmq0SAvo

Listen on Apple Podcasts:

https://podcasts.apple.com/gb/podcast/548-ai-is-about-to-trigger-an-energy-crisis-most/id718416620?i=1000752299883

Listen on Spotify:

https://open.spotify.com/episode/5l4674hFIJPWkz0spMq4YL

Transcript

Disclaimer: This transcript was generated by AI and may not be 100% accurate. If you notice any errors or corrections, please email us at phil@wealthformula.com.

 Welcome everybody. This is Buck Joffery, the Wealth Formula podcast. And today, before we begin, I wanna remind you as always, there is a website associated with this podcast, wealthformula.com. That’s where you want to go. If you have, uh, an interest in uh, ing more in the community in particular, there is a, a credit investor club.

AKA investor club, which you need to sign up for. Uh, go to wealthformula.com and see some private deal flow at, uh, no cost to you, uh, that, uh, you might have an interest in. Uh, let’s talk about today’s show. It’s a little bit about, uh, something. You know, that is, uh, on I think, a, a major issue, uh, going into the next decade.

Um, you know, there’s one truth that’s followed. Every major technological revolution in human history. Energy demand is always rise, uh, to meet technological capability. You know, when we industrialize, uh, coal consumption exploded, obviously when we built modern transportation system oil. Demand, uh, reshaped global geopolitics.

And when he entered the digital age, electricity became the backbone of the global economy, and now we’re entering

the era of artificial intelligence. Now, what most people don’t appreciate is

that AI is not just a software revolution, it’s an electricity revolution. Uh, training a single advanced AI model can consume as much electricity as literally tens of thousands of homes in an entire year.

And once trained, these models continue to run inside data centers filled with specialized hardware operating 24 hours a day. A single large AI data center can require what’s called a entire one gigawatt of power. Now, what’s a gigawatt? Well, to put this all into perspective, that’s enough electricity to power.

Roughly 700,000 homes, one building consuming the equivalent of a major city. Now, consider that companies like Microsoft, Google Meta, Amazon, they’re applying to build dozens of these facilities, and suddenly you begin to see the scale of what’s happening. Uh, even individual AI queries when you do them, they consume a lot more power than traditional computing tasks.

Um, there’s an estimate that suggests that an AI query. Can use roughly 10 times the electricity of a traditional, uh, search query. The difference seems trivial until you multiply that by like billions of these interactions per day. And that is why for the first time in decades, electricity demand in the United States is accelerating again and doing so quickly.

Now you might ask, well, you know, what’s been happening for the last 20 years? Well, electricity demand was actually relatively. Flat. And a lot of that is because of efficiency gains, offsetting economic growth, but ai, electrification of transportation, domestic manufacturing, they’re all gonna reverse that trend.

And, and here’s where the story becomes even more interesting, because we know that China already understands this. China’s building power infrastructure at a pace that’s difficult to really even comprehend. They’re adding entire national skill, power, capacity every few years. In 2023 alone, China added more new coal power capacity than the rest of the world combined.

And at the same time, they’re installing solar, wind, all these things at record rates becoming really the global leader in re renewable deployment. So you don’t think of China is that way, but they are. They’re not choosing one energy source. They’re choosing all of them. And because they understand that energy availability will determine technological leadership.

Meanwhile, in the US things are kind of slower. Building a, a new power plant and transmissions infrastructure can take a decade or more. We got lots of regulatory hurdles and permitting delays in political resistance that the Chinese don’t have, and that creates a lot of risk. The country that can generate the most reliable, scalable energy, we’ll have a structural advantage in AI manufacturing and economic growth.

And that is a big, big deal because energy at the end of the day is becoming. The limiting factor for growth, and whenever something becomes a bottleneck, you also get investment opportunities that emerge. So we’re entering a period where trillions of dollars will be spent on power generation, grid modernization, nuclear energy, solar battery, geothermal, you name it.

And a lot of those things you’ve never heard of. Some of the biggest fortunes of the next decades will be tied directly or indirectly to solving these energy constraints. That is why in today’s episodes we’re gonna explore these alternative energy sources, kind of get an idea of what’s going on with them.

I know it doesn’t sound super exciting or sexy, but understanding energy right now is, is not optional. If you wanna understand where the world is going, and as investors, those who see these shifts early are gonna have an opportunity to position themselves ahead of the crowd, and we’re gonna have. A conversation to highlight all of that right after these messages.

Wealth formula banking is an ingenious concept powered by whole life insurance, but instead of acting just as a safety net, the strategy supercharges your investments. First, you create a personal financial reservoir that grows at a compounding interest rate much higher than any bank savings account. As your money accumulates, you borrow from your own.

Bank to invest in other cash flowing investments. Here’s the key. Even though you’ve borrowed money at a simple interest rate, your insurance company keeps paying. You compound interest on that money even though you’ve borrowed it at result, you make money in two places at the same time. That’s why your investments get supercharged.

This isn’t a new technique, it’s a refined strategy used by some of the wealthiest families in history, and it uses century old rock solid insurance companies as its back. Turbocharge your investments. Visit wealthformulabanking.com. Again, that’s wealthformulabanking.com.

Welcome back to the short rewind, uh, energy demand is, uh, rising, not just from ai but from electrification.

Population growth, economic activity itself. At the same time, we’re trying to transition how energy’s produced, which creates, uh, real trade-offs around cost, reliability, and scale. Today’s conversation isn’t about, uh, ideology necessarily, but it’s about the economics of energy and what’s realistic as demand continues to grow.

And to help us think this through. I’m joined by Dr. Ga Hockman, professor of Environmental and Resource Economics, with the PhD from Columbia University Gall. Welcome to the show.

Good morning.

So let’s just start very basic here. In your view, why does economic growth almost always translate into higher energy demand?

Because production is very dependent on energy. And so whenever you wanna expand production, you wanna expand food, you need more energy. And this is actually what we’re trying to decouple, to create production processes that are less energy intensive. So as we grow, as we become happier, more viable, we don’t necessarily need more energy.

So, uh, setting, uh, ai, artificial intelligence aside for a second, are we already in a path where electricity demand has to rise, you know, meaningfully over the next decade? I mean, what, what kind of projections do we look at there?

We need to decouple growth from energy. We didn’t do that yet. As long as we don’t do it.

Uh, growth will be associated with an increase in energy demand, not as much as AI has been introducing. And that is, uh, uh, uh, jumping to a higher step. Right. Now,

you’ve mentioned this a couple times in the decoupling idea how in the big picture, like how do you do that?

Uh, does the low hanging fruit that the US implemented from the 1980s, 1990s, and that is energy efficiency.

It, which creates a win-win. Uh, it just changed the light bulbs in your, in your house. You save electricity, but you also save money ’cause these bulbs last much longer. Assuming their cost is not high enough. Is not too high. Uh, industry is the same thing. Introducing more efficient processes. Can result endless need for energy, but we need to go a step further to make it more meaningful and to introduce production processes that simply depend less on energy or depend less on energy that is polluting.

Give us another example. I mean, the light bulb is an easy one, but, um, I mean, what are some large scale ideas for that energy efficiency issue? That you’ll think about when you think about these kind of decoupling ideas.

Uh, another thing, just, uh, the appliances at home, uh, you want them to, uh, be more energy efficient and the windows you put on your houses, you want it to be double blast, maybe even triple in some cases that blocks the sun and helps I, uh, isolate the house better so you don’t need to heat it as much.

Insulation is very important. Uh, very similar things exist in the commercial sector. Uh, if you look at the big retail stores, they’re using a lot of light bulbs. They’re using a lot of insulation to reduce their, uh, heating costs. If they are wanting to become more energy efficient. So these are not very complicated things that can really make a change in residential, in commercial.

And you can then expand it further into production process in the manufacturing. And there are different examples also there.

There’s also this big driver of energy in the next couple of decades, uh, which, you know, people talk about how many more terabytes we’re gonna need just to support the artificial intelligence revolution.

Do you think it’s realistic, you know, just to focus on these efficient levels? Is that enough for, for how much energy we need?

No, no. And we need to expand the energy. Uh, it’s important to expand it in ways that is cleaner energy, so it does not create harm. So you don’t create a good with a bad, uh, you wanna introduce energy that is cleaner so you don’t increase, uh, pollution.

Uh, impact greenhouse gases. Um, so it is also the fuel mix that you’re using. The fuel sources. Will you use solar? Will you use hydro? Will you use, uh, wind, uh, bio bioenergy, same thing. Bioenergy crops. So you wanna exp expand, you wanna. Introduce a more diverse set of feedstocks that many of them are much more, uh, cleaner than the existing one.

Uh, so the movement to renewable is important. Uh, and again, you don’t need to decrease the existing infrastructure, but the new infrastructure at least needs to come from a cleaner sources. You need to improve our use of batteries.

Yeah. Let, let’s break down some of the things that you’ve talked about. So, solar, okay.

Um, what did, what does solar do well and where does it struggle?

Solar, people forget, in 2005 it was $10. Now it’s below $1. So we need to understand that there is a transition in the transition. Many times costly, but we need to learn and bring it down that. Learning came in terms of installation. The installation became much more efficient, uh, much less costly, much faster, and that brought the price of solar down.

Uh, solar has been performing very well in many places. Uh, eh, solar today is cheaper than many of the most polluting, uh, infrastructure for power in the world. If I remember correctly, the number, it’s around 500 gigawatts, which is a big number. Uh, they can, that solar can outcompete the existing, uh, energy sources.

Uh, where it’s struggling is that, um. Silicon will be is is in high demand and that is a creating a floor that prevents solar from going even lower, but it can also create a constraint in the future as you expand it further.

Can you explain for, for us just the silicon issue? ’cause is that. So it’s just a, a silicon is a major component and we don’t have enough, is that what you’re saying?

Yes. Yes,

exactly.

And then doesn’t that drive up the price of silicon?

Yes, but we, we didn’t hit that. We, we we’re, we’re, uh, but there are actually various entities working on alternatives. From MIT to companies, uh, that are offering interesting solutions. Yes.

You mentioned storage as well. Um, energy storage.

Um, how close are we to storage being really viable at scale? I mean, this is, um, you know, we certainly, battery technology has improved, but, you know, how, how, how close are we to it? Becoming something that is, is really, really helping the issues. Uh,

it’s challenging ’cause right now it makes it more expensive.

But if the more we use it, the more we learn, the more we understand, the more, uh, efficient and cost efficient we can introduce it. Cost will go down. So it’s like the, how do you push it forward? How do you adopt these technologies? Now, we should always remember that there are, in some places, it is already very viable.

But it demands certain, uh, uh, circumstances. For example, uh, the Southwest has a location where it has, uh, underground water and solar. The solar heats the underground water. So the underground water becomes the storage that, uh, then the steam becomes the electricity in the night. And that is a very viable process.

Hydro with wind goes also very well, and again, uh, they manage to store, uh, use the wind to bring water upstream, and then when there’s no wind, the water flows downstream and through hydro creates electricity. Batteries, it’s technology. Uh, will a breakthrough come one day? I believe so, but again, I, I can’t predict it.

Um, we can talk about, um, you know, natural gas, right? I mean, natural gas doesn’t get much attention, uh, in the transition narrative, but how important is it today in maintaining grid stability in supporting renewables?

Reliability is more important than prices to many of us. No one likes blackout and if you talk with the, those that monitor and and manage the electricity markets, that’s their top priority, not the price.

Uh, we don’t like it when we don’t have electricity. We we’re very dependent on it. So reliability is definitely be, uh, uh, uh, a must before you even move towards renewables. Absolutely. Before prices even, uh, uh, for anyone in the us. Um, so NA Gas has the potential, uh, it has less. CO2. The problem with NA gas is that the infrastructure is leaking.

That means that the pipeline are emitting and methane because of leaks. Uh, I believe that needs to be addressed. Uh, uh, natural gas has the potential to be used, but. You need to not use it with an infrastructure that is, uh, resulting in more damage than good. It kind of defeats the purpose of it.

What would do you look at natural gas as a short term bridge or something that, you know, the, the system may rely on, you know, in, in a much longer, uh, timeframe, even with other renewables.

I would be careful in creating a bridge because that this infrastructure is very expensive. Once you put the amount of money needed to create infrastructure, it’s very hard to change it. Having said that, you will have solutions that will use fossil fuels, which includes natural gas, even in the long run, simply because the cost and the benefits will add up in a way that.

It won’t make any sense moving away from fossils. In my opinion, not everyone will agree with me. Yeah, but, and, and you do have technologies that can make fossil fuels much, much cleaner. Like carbon capture used in storage. Uh, that technology has a huge potential. You can recycle the hydrogen and recycle other components in the refinery process that results in a cleaner fuel.

But it’s something that we need to incentivize the companies to do. Uh, a company will not do it independently ’cause it’s more costly and that’s important.

How about nuclear? I mean, nuclear. Offers reliable carbon free, you know, power. Yet it hasn’t scaled the way many people expected. Um. Why is that

people are afraid of nuclear.

Look at the three Mile Island and, and look at Fukushima and Chernobyl for that matter. People remember those stories and that really resonates with them badly. And there’s also a problem in the accounting of nuclear. Even the most safest countries in the world like Japan will everyone considered super safe.

Even they have an accounting problem. So there is the concern that. Even small amounts get leaked out to the wrong hands. That can be a very bad outcome. Eh? Having said that, there is, I don’t know. I don’t follow it too much, but I do know there is a drive to create small nuclear plants, mobile plants, eh, from my recollection for two, three years ago, the company that I heard of was very successful at that.

Eh, Japan went back to nuclear different than Germany. By the way. Germany did not try to, uh, divest from nuclear. So there are some places that nuclear becomes very important. I think it’s also becomes important in some areas that work in ai. So it has been

introduced as a source of electricity. Can you tell us a little bit about small modular reactors?

There’s a lot of buzz about that. What, what exactly are they? I mean, how small are they? You know, safety wise,

uh, they’re mobile, they’re not very big. And, uh, that makes them, uh, much more easier to manage and control as opposed to the very big nuclear plans. Nuclear is a base load. So you use it, you, once you turn it on, you don’t want to turn it off.

It’s too expensive. The on and off, it takes it a long time to, to uh, ramp up. Uh, and, uh, mobile, uh, nuclear plants are addressing many of these concerns that exist with the big plants. So they are solving it in, in what I saw pretty well in some

circumstances. How small are they? I mean, are they, so would you.

Would a, you know, one of these AI data centers, or what would they just, would they have one small modular react or they’ll

need more than that? They’ll need more than that.

Oh, they need more, more than one. Yeah.

Yeah,

yeah. So they’re, they’re pretty small or they like, you know, the size of a car or they. How, how small are these things?

No, they’re bigger than the car, but they’re not too big. If you know of a nuclear plant, the old one, you see these big round, uh, domes, uh, they’re, they’re not that big. They’re, they’re much smaller, but they’re not as small as a car.

Yeah. And so you could run maybe, uh, a, an AI center with a couple of those or something like that.

Is that the idea?

They have, you can see some of them. There are examples in Texas where you have the, the center basically is surrounded by small units.

Are they generally safer to use, and if so, why is that?

Uh, I’m not a nuclear guy. I’m not a physic. I should be careful in it, but I, I, what I understood, they’re safer to use.

Also, the material i, i I is not reaching, uh, levels that safer levels than you would need for, for example, for bumps and, and stuff like that. So they’re keeping everything at a safer level.

When you step back and look at the whole system and think about. What’s gonna happen in the future? Do you think it’s more likely to be dominated by one energy source or like a diversified mix as we’ve been going through?

I believe a diversified mix. I also believe that in some places you will always have fossil fuels. In some places you’ll have a very quick transition to renewables. Uh. Uh, we need to look at the system view. In some places it’s easier to clean the dirty fuel. In some places it’s just easier to introduce the, the clean fuel.

Uh, some places I do believe you see, for example, developing world does not have the capacity to electrify. We talk about electrification and some people are very enthusiastic about it. You don’t see it in the development world. They don’t, they lack even the US And there is a study in Princeton that came, I think three years ago.

Um, if you electrify the whole US today, you need to almost triple the grid capacity. Just understand what the magnitude of money that needs to be invested to get there. Is huge. Now developing countries definitely don’t have it. Even the US doesn’t have that capacity. So, uh, developing countries, I think you might see a lot more biofuels, a lot more, uh, other, uh, substitutes that exist that are easier for them to manage.

And then a system view or a more complete view is needed ’cause it’s not. What is the most efficient process? Is what process fits best in a certain area, and, and that will create a lot of heterogeneity, I think.

Do you have a sense in the us I mean, what, what do you think ends up being? There’s gotta probably be one, you know, dominant source that it will, will kind of come to friction based on our own.

Economics in our own situation. Do you think that’s in the, in the near future? Is that solar, you think? I mean, what, what dominates in the future here?

I don’t think you’ll dominate, even in the us you won’t dominate, uh uh. You have regions in the US that are very, uh, windy. Wind farms will be the optimal path.

There are places that don’t have any clouds, 350 days a YA year. So solar is perfect there. Solar also creates employment and live view for certain communities so that the employment component is an important part. So you create. Income and, and, and, uh, in, in, in life, in, in economic variability in regions with the renewables, there are other regions that have, uh, a lot of supply of, uh, excess biomass or the capacity to produce a lot of biomass, and that creates them an alternative to use biomass ’cause that’s what brings them.

Again, income, which is always important, but it also brings them a feedstock that might be of a, a lot of benefits. Um, and you will have regions that are heavily so heavily invested in fossils that it will never make sense to move away from fossils, but it will make sense to create cleaner fossils through carbon capture and storage in other ways.

So I don’t think the US will move into one place or another.

Yeah. Um, you know, you often hear discussions about, in the US about, um, our grid being outdated. Tell us sort of at, at a high level, if you wouldn’t mind explaining the issues with the grid and, you know, what, what kind of issues that brings up as we need more energy sources.

Just look at the power plants. They were, look at their ages, the age of power plants. Look at and, and then there are a few that were supposed to be retired and now have been extended, but just. That by itself is sufficient to create problems whenever you encounter a natural, uh, extreme event that, uh, stresses the system.

Uh, we saw with Sandy in the northeast. The northeast was, a lot of the infrastructure was outdated. Sandy came, the system collapsed. They fixed it now, so they upgraded it. There is, uh, uh. Some of the utility. Again, I’m not, I’m following anecdotal evidence and news, not beyond that, but some of the companies are striving to improve their grid and they are trying to, uh, introduce a more sustainable and reliable system again, ’cause reliability is so important.

What does, what does it mean really to even update the grid? I mean, just for people who are not in this space, what does that even mean to upgrade it?

You, you, you change the equipment, you upgrade the equipment, you better manage the inter, uh, interaction of trees and, and, and the electricity lines. Uh, you bring electricity lines underground.

You also improve a lot of the infrastructure, uh, of the power plants and how they distribute the energy. So this whole infrastructure is being upgraded so it can support. For example, the ai. And that actually is something that the AI might bring as a very positive thing. So it will force the system to, uh, upgrade, to introduce more efficient processes, uh, distribution mechanisms that are more resilient, which I think is important.

I hear we’re kind of behind when it comes to this, when you compare it to China. Can you talk a little bit about that? China has a different structure of, or economic structure. So a lot of the, uh, driver, the driver in China is the government and money that the government allocates to these alternative technologies, and that creates a very strong drive for renewables.

Eh, China is also a big driver in coal in China, so. It’s basically where the government decides to put the money, and that’s where you see the industry flourish. If you look at the numbers, the investment numbers, China outpaces any country in the world in terms of the value invested per year in the recent years,

and, and they’re producing a lot more, a lot more energy than us too.

Isn’t that correct? I mean, I, I’ve just been, just in terms of following the AI news, I keep hearing about it. China has no. So many more terabytes than us, uh, of energy, uh, ability. Is is that true?

Uh, that I don’t know. I don’t know exactly ’cause, uh, I know they’re producing a lot. I know they are expanding a lot, and I know that in the solar space, for example, they dominate because of that.

They’re already, they’re also starting to dominate in the electric vehicle space. Uh, they’re becoming to leaders in those areas. Yes.

Um, big picture, I think if you wanted to sort of sum up some of the, you know, major issues that you think that, you know, people like us who are. Investors or you know, just people wanna know what’s happening in the future.

Like what, what’s, what’s the message for, for people?

I would, I would try to make my house more efficient. I would try to, uh, and it’s important to understand this is not only about, it is about greenhouse gases, but it’s also about if your house is more efficient, you are also paying less money. And that has a lot of benefits to it.

Similar logic can follow to the industries and how they work, how, and, and conserving energy is not necessarily coming at the cost of being more or less productive. That’s what we need to understand. You can conserve energy and still produce more. You can become more efficient and you can still, and you can reduce your dependencies on, uh, energy, which I think is important.

Dr. Ga Hoffman, thank you so much for being on Wealth Formula Podcast today.

Thank you for inviting me.

You make a lot of money but are still worried about retirement. Maybe you didn’t start earning until your thirties. Now you’re trying to catch up. Meanwhile, you’ve got a mortgage private school to pay for, and you feel like you’re getting further and further behind.

A good news. If you need to catch up on retirement, check out a program put off by some of the oldest and most prestigious life insurance companies in the world. It’s called Wealth Accelerator, and it can help you amplify your returns quickly, protect your. And money from creditors and provide financial protection to your family if something happens to you.

The concepts here are used by some of the wealthiest families in the world, and there’s no reason why they can’t be used by you. Check it out for yourself by going to wealthformulabanking.com.

Welcome back to the show everyone. Hope you enjoyed it. And, uh, yeah, again, you know, the goal of this show is really to give you, you know, a, a macro look at what’s going on in the world and one of the things that is.

Clearly an issue for the United States is energy production. And so, um, you know, stay on top of this stuff. This is, you know, this is where the puck is headed, right? Um, ai, all these things that are, are really, uh, driving the next decade of growth. Really depend on it. Anyway, that is it for me. This week on Wealth Formula Podcast.

This is Buck Joffrey signing off. If you wanna learn more, you can now get free access to our in-depth personal finance course featuring industry leaders like Tom Wheel Wright and Ken McElroy. Visit wealthformularoadmap.com.