Skip to content
Pour Over: AI Energy episode artwork

Jan 21, 2026 11 min

Pour Over: AI Energy

from The Daily AI

In this episode, we pour over the details of AI's energy consumption and the Trump administration's $15 billion power plant proposal. We offer a distilled analysis of the complex energy landscape supporting AI innovation. Chapters 00:00 AI's Power Hunger 01:48 Trump Admin's Plan 10:07 Energy Source Debate 13:30 Accountability for Consumption In this episode, we explore the Trump administration's proposal for tech companies to invest $15 billion in power plants to meet the surging electricity demands of AI and data centers. We also discuss its potential impact on America's power grid, consumer costs, and the national effort to remain competitive in the global AI race. Chapters 00:00 $15 Billion Power Plant Deal 01:48 PJM Grid Operator & Energy Needs 04:44 Alternative Power Solutions 06:45 The Trump Administration's Plan 10:05 Renewables vs. Consistent Power 13:33 Addressing Critiques and Future Links Get the top 40+ AI Models for $20 at AI Box: ⁠⁠https://aibox.ai AI Chat YouTube Channel: https://www.youtube.com/@JaedenSchafer Join my AI Hustle Community: https://www.skool.com/aihustle

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

Pour Over: AI Energy

0:00 11:09
of matches

Transcript Auto-generated, may contain errors

Welcome to the podcast. I'm your host, Jayden Schaefer. Today on the show, we're going to be talking about the latest deal from the Trump administration in regards to having tech companies by $15 billion worth of power plants. Now, this is coming at a time when a lot of these big tech companies are needing more and more power to help fuel the compute for train all of their AI models and power all of their data centers.

So today on the show, I want to break down how this could shape some of America's power consumption and how this deal worth, you know, $15 billion of power generation is going to affect the nation's electrical grid. And essentially, you know, what's going to happen if we don't do this? This is interesting as America is now trying to stay competitive in this global AI race. There's a lot going on here.

Before we get into the story, you probably pay for multiple subscriptions to get access to a lot of these different AI tools. It can definitely add up fast. I had a very similar problem. And so I built AIbox.ai.

So for $20 a month, you get access to over 40 of the top AI models. I put them all into one platform. You get text, image, audio, everything you need all in one place. So you don't have to double tabs.

You don't have to waste money on overlapping subscriptions. If you want to give it a try, there's a link in the description to AIbox.ai. All right, let's get into the story. So the Trump administration is joined by a group of governors in the mid-Atlantic and Midwest.

I think it's kind of bipartisan group that's coming together on this. But they've all asked PJM interconnections. This is the grid operator that serves about 65 million Americans. It's over about 13 different states and is, you know, holding an quote-unquote emergency power auction to procure new generation capacity.

So the idea is that tech companies, especially those that are building these huge data centers everywhere, are going to bid on 15-year power contracts that could help underwrite the construction of a lot of new power plants, essentially trying to keep pace with all the surging electrical demands. And personally, as someone who has seen their electricity costs, you know, when I was living in Arizona, literally double as a lot of new facilities, a lot of subsidies and government grants went out to a lot of these companies, building data centers, building chip fast. And then, of course, the electrical bill of everyone else goes up as those ones are subsidized. So in my opinion, this is a fantastic program.

We need to build more power generation. This is something that's definitely, I think, on the top of everybody's mind in the industry. And how to do that is something that we're all discussing. This deal right here, I think, could unlock up to about $15 billion in new generation capacity.

And of course, electrical demand from data centers is absolutely skyrocketing right now. Utilities are really struggling to keep up because all of the new generation takes years and it is a mass investment if you're going to build them. So it needs actual concrete, physical plants, and it needs these things online. So there's a lot of different strategies that, you know, companies are using to try and make this a reality.

One from Microsoft is they went to the decommissioned nuclear power plant off my island and said, hey, if you guys can recommission this, we give you a billion dollars to do that and then we'll make a contract to buy all of the power that comes out. So we're seeing, you know, decommissioned nuclear power plants come back online or start getting retrofitted to come back online to generate electricity. There's a lot of deals. I mean, we even see like there's a whole scandal of XAI and I believe Memphis where they had a big data center there and they literally just brought on like thousands of these portable diesel generators and turned them all on to power their whole facility because they couldn't get enough power from the grid.

So obviously this is a very acute problem for a lot of these things. And with all this, we need to start, I think doing some, you know, better planning than just firing up a million generators. And I mean, there's some, you know, there's some like good angles like the three mile island generator, but that's not going to be something that, you know, there's like a million of these decommissioned nuclear power plants all over the country. There's a few of these little, you know, sort of side projects or a few of these sort of undiscovered areas that people can capitalize on.

But for sustainable long term, America needs to be building a lot more energy consumption. So with all of that going on, PJM and a lot of power markets in the region have been under a lot of strain. And so these capacity auctions in recent years, they have seen some record high prices. And I think there's real concern on regulators about the reliability as kind of low forecasts are continuing to just increase.

So right now you have the Trump administration's plan, which is looking to help with reliability. It's trying to limit some consumer costs increases and also to try to spread the responsibility fairly. And essentially what they're doing is having the biggest demand drivers, which is tech companies building data centers, help to bankroll the expansion of capacity. Now, I think this is something that the tech giants would do in one way or another on their own.

If I'm being honest, I don't necessarily think you always need the government to step in. But perhaps in this way, they're just getting to the table now and trying to expedite the process of getting a lot of this power consumption built out. I'm sure there's going to be like arguments and, you know, it's good or it's bad from both sides of like a political spectrum in the United States because it involves the Trump administration. So by default, it's going to be controversial.

And the thing that I think is not controversial at the end of the day is we need more power consumption. So anything that's going to expedite that process or pay for it or get it done sooner is something that we desperately need. This isn't just for tech giants, but this is for average consumers. If you want your electrical costs to go up, we need more electrical capacity.

And this is something that we're getting absolutely, you know, run over by in regards to China and the expansion of their electrical capacity. They're building way more than we are at a way faster rate. And so they're going to be able to build things cheaper and faster than us because of that if we don't try to keep up. So I think one positive way to think about this instead of just kind of leaving the expansion of the grid to guesswork, the government is signaling a really clear direction so the market can respond.

Long-term contracts bring a lot of certainty that private developers need to invest in big power projects that also guarantee revenue, which can lower financial costs and can help accelerate the construction timeline for new generation. So I think this is overall a good thing. Something else that I know a lot of people like to talk about is a lot of these renewables, you know, electrical sources, solar wind, as kind of being the silver bullet. I think as far as what we're doing today, I think something that a lot of people like to talk about, a lot of the mainstream news does seem to gloss over is the fact that solar and wind is intermittent.

So they only generate when the sun is shining or when the wind is blowing. Data centers on a lot of these really heavy industrial loads. They don't get to take, you know, any sort of breaks or clock out when the grid goes quiet at night. So they need a firm, dependable 24-7 electricity.

Now, the solution in the past for that has been battery storage and today is battery storage. This definitely helps, but it definitely is not a full replacement because batteries paired with solar can cover, you know, definitely a few hours of demand, but they are still limited for long duration loads. And something that I had an experience with when I was growing up, I sailed from California to the South Pacific on a sailboat, which we largely were powering by solar. And, you know, we would have times where we'd have a storm blow over for like four or five days of really cloudy and our batteries were completely depleted.

It was really scary because all of our like autopilot navigation systems, all of our charts, like, you know, everything we're running on our ship was depending on these batteries in the solar and it was all starting to glitch out and stop working. And, you know, lucky for us, the storms passed, we've recharged our batteries and moved on. But this has always kind of been something that's reminded me that while these are amazing, you know, options, we use solar panel for like a whole year on this ship, it's not always the same. Well, you do need, I think, multiple facets, multiple types of energy to power this stuff.

I think for, I think for a lot of these kind of data centers, it's going to be the same thing. When a weather event happens or there's a really prolonged, you know, cloudy period, you still need very consistent solid energy. And so solar and wind can't keep up with that forever. So I think this is why right now the current administration is putting a push to include some traditional base load capacity.

And I think at the end of the day, the best option is multiple sources of energy production. Okay, so I think something else I've heard a lot of people critical on this is, you know, like this whole TechCrunch article that's like, you know, asking why you'd make tech giants pay for the power that they might not use and like maybe this isn't what we need to do. I think if you kind of look at this, there's a couple ways to look at this, but really a lot of these companies are building massive facilities, right? All these data centers, they require continuous powerful electricity.

They aren't just occasional users. If you set up a data center, it's not something that can just like, you know, use electricity every once in a while. Like it's going to be sucking electricity and a lot of it for forever, basically. So whether you're trying to power an AI train cluster or cloud service, critical internet infrastructure they're depending on the power basically like a factory does.

It's very constant and consistent. So when data centers connect to the grid and they haven't really ensured, you know, that there's enough power in future generation, it transfers the risk onto rate payers and families. So electrical costs are even rising in the JPM region. I think that's why we're kind of looking at it and addressing it there specifically right now.

I think the part of this is because the capacity markets are being stressed by demand growth. There's a lot of demand. I personally have seen this. The electrical costs from when I first purchased a house in Arizona has doubled.

I was just looking over my finances and looking at like the last five years of finances and today versus like four years ago, my electrical costs have doubled. I know maybe people could say like, maybe I'm using double as much electricity. I'm not. It's the exact same.

It's the same AC unit. It's the same house. Everything's the exact same. The electrical costs have just literally doubled.

So I think we're definitely seeing this. I think by having these companies, you know, contract for new capacity, kind of better aligned responsibility with the demand. Like they are using this. And I think that this is why it's important to get them to start paying for this in my personal opinion.

I also say it's worth knowing that JPM hasn't been fully on board with this whole idea yet. So it seems more like Trump administration and a lot of these bipartisan governors are kind of pushing this on them. They are reviewing the proposal and they're working through some competing stakeholder interests. I think that's exactly how this probably should work.

Good operators and private companies negotiating in kind of a market framework rather than having this all imposed on them. But I do think that the administration's move has already kind of shifted the conversation in a really meaningful way. I think it sparked a really serious debate about how we plan energy infrastructure. Right now there is so much compute demand and there's a lot of companies that just keep pulling out these data centers and it feels like there's not always capacity and a plan built out to power them once that happens.

Thank you so much for tuning into the podcast today. If you enjoyed the episode, make sure to leave us a rating and review wherever you get your podcast. It helps to show out a ton to be found by more incredible people like yourself. And as always, if you want to go check out AIbox.ai, you can get access to over 40 of the top AI models all in one place for 20 bucks a month.

I'll leave a link in the description. Have a great rest of your day.

Frequently asked questions

How long is this episode of The Daily AI?

This episode is 11 minutes long.

When was this The Daily AI episode published?

This episode was published on January 21, 2026.

Is there a transcript available for this episode?

Yes, a full transcript is available for this episode. You can read the complete transcript on the episode page.

Can I download this The Daily AI episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.