Podcast: Electricity Was Supposed to Save Us. Will It Doom Us Instead?

The mantra, dubbed “electrify everything,” was seen as one of our most immediate opportunities to decarbonize; you can wipe out a lot of emissions when you replace a gas-powered vehicle with an electric version.  

And as consumers began to choose more electric vehicles and appliances, the plan seemed to be working. 

Until artificial intelligence came along. 

In Episode 5 of our third season of Insight Unpacked: Will AI help us reach net zero? Or will it be our undoing? We consider the possibilities and make a case for hope.

Laura PAVIN: Up until the 19th century, most people’s days ended when the sun went down. 

If you wanted to do some work, or host a party, or just read a book at night, your options were limited. You could light a candle, but those were expensive and only provided so much illumination. You could light a lamp that burned whale fat—even more pricey. And, you know, meant killing whales. 

The modern oil industry, as we know it today, actually got its start solving this problem. The discovery that petroleum—from the ground—could be turned into kerosene for home lamps was a huge breakthrough. The “new light” built the first oil giant, Standard Oil, and it changed life around the world, literally lengthening people’s days. 

But that solution created new problems. Burning kerosene at home or at work dirtied the air, smelled terrible, and carried the risk of, well, explosion. It seemed like there were major, unavoidable trade-offs for the modern convenience of artificial lighting, trade-offs that harmed the environment and public health and put your family at risk. Sound familiar?  

Fortunately, there was another innovation before the century ended: the light bulb. And while oil quickly found a new market in the horseless carriage, electricity set home-energy use on a cleaner, safer, more-affordable path.  

It took what felt like an intractable set of problems with something our modern lives had grown to depend on and turned it into a sidebar. And we didn’t see it coming. Until it came.  

You’re listening to Insight Unpacked: Can We Still Build A Green Economy?  

So far this season, we’ve seen how building that green economy might be hard here in America. Businesses are trying, but it’s not enough. Investors are trying, but it’s not enough. And our ability to pass and properly enforce policy isn’t enough. 

But, you know, maybe the problems we’re arguing about now could be transformed by something we didn’t see coming—like the lightbulb in the nineteenth century.  

On this fifth and final episode of the season: we look for hope in the unexpected.  

And what’s more unexpected than AI? It’s thrown wrenches in a lot of our plans—like job prospects and how we educate our kids.  

There’s also the environment, because AI sucks down a whole lot of our precious electricity. But what if there was a world where it unexpectedly was good? That world does exist, experts say—and, ironically, it depends on the outcome of a very familiar-sounding rivalry between the light bulb and oil.

Then, the U.S. has clearly de-prioritized the climate under the current presidential administration, but that doesn’t necessarily mean the world will burn because of it. In fact, it might surprise you to know some experts are confident that the issue is less about whether we get there and more about who will lead the charge, while the rest of us follow.   

We put our faith in the unexpected—next.  

...

PAVIN: To understand why we’re at this pivotal moment with AI now, you need to understand what it’s interrupting. Where we were headed before it basically came out of nowhere.  

For all of our shortcomings you’ve heard this series, we were still making progress on something that felt like a plausible path to a green economy here in the U.S. Not a perfect one. Not a fast one. But you could see it. Coal plants were retiring. Heat pumps were outselling gas furnaces. EVs were showing up in driveways. The grid was slowly getting cleaner. 

And there was a reason—a mantra—that we had to thank for that.  

“Electrify everything.”

[Office chatter]

PAVIN:  Well, let me just get a sense of the levels on this microphone. Could you tell me what you had for breakfast this morning? 

Holly BENZ:  I had yogurt and fruit. 

PAVIN:  Holly Benz teaches at McCormick, Northwestern’s engineering school. She was very inspired by this mantra.  

Holly BENZ:  So Electrify Everything is the name of a course I teach. I started it with an exclamation mark, as came through in a book by Saul Griffith called “Electrify.” He was sort of one of the founders of this rallying cry. 

PAVIN: Benz says that this Electrify Everything imperative, it was a response to a very real problem. 

BENZ: If you look at what is contributing to particularly CO2 emissions, it’s the combustion of fossil fuels, so the solution to not combusting fossil fuels is to replace it with electricity.  There was a heavy focus on electrifying end uses: human consumption—be it residential, commercial, transportation, or industrial. 

PAVIN: Most of what produces emissions in the world has one thing in common: it burns something. Your car engine. Your gas furnace. The factory down the road. Replace combustion with electricity, and you’ve addressed a huge chunk of the emissions problem.  

It’s not a new idea.  

BENZ: People have been talking about electrifying end uses for many years, but it really kind of took off again around 2021, when the Biden administration put in place a lot of supports for end-use electrification.

PAVIN: End-use, by the way, just means the point where energy gets consumed. Again, your car engine, your gas furnace, all that—those are end-uses. The Biden administration was very into this push. Supported it with policies and incentives.  

And things were moving along. But they weren’t perfect.  

I certainly noticed a disconnect in the logic.  

Walk with me for a second.  

[Footsteps]

So, I have an electric car, right? And every evening, after a day of driving about town, I go down to my garage ... 

[Door opens and closes]

… grab the giant plug we had installed in our wall … 

[grunting sound]

… I push it into the car …

[sound of a car powering-up] 

… and the thing charges overnight.  

This is good for the environment, they say.  

But if I were to follow the electricity from my car’s plug all the way back to its source, I would find that, some of the time, it’s coming from fossil fuels. In 2024, about 30 percent of Illinois’s electricity came from coal and natural gas, which isn’t too bad, but it’s kind of ironic to think that my electric car’s power is sometimes made possible by the thing it’s railing against.

If you look beyond Illinois, and broaden the aperture to the U.S., by the way, a much larger percentage of the country’s electricity comes from fossil fuels.  

So, what’s the point of going all electric if the grid isn’t holding up its side of the deal? 

It’s part of a bigger plan, Benz says.  

BENZ: The idea is the transformation of end-use electrification; those decisions are made by a million or millions of different people, different stakeholders, right? Individuals make choices in their homes, individual businesses make choices, fleet owners make choices.  And that transformation takes longer. But the idea is you need to set that in motion so that the demand is there by end users for electricity, and then you can make the right choices with a smaller set of stakeholders who are making generation decisions.  

PAVIN: Getting all these different consumers buying all these new electric products would push demand up. And, the thinking went, it would push the grid to slowly green itself.  

Here’s how.  

For years in this country, the demand for electricity has been flat. Which has meant that utilities—like your ComEds of the world—they haven’t had to expand their generating capacity in a big way for a while.  

But, the thinking went, if we increased our demand for electricity enough, if we added a bump in that flat demand, it would push utilities to build more generating capacity. More plants. And the bet was that this time around, those plants would be clean instead of dirty.  

Because in the time since the last push to expand generating capacity, wind and solar—clean options—have become the cheaper, more-economical choice. So, bump up demand, utilities will build clean capacity and, importantly, retire their coal plants.  

That was the idea.  

And as time went by, as consumers chose more electric things, something happened.  

BENZ:  There was a slight increase in the electricity consumption profile. 

PAVIN: It goes up slightly. And as uncinematic as that sounds, it’s the slightness of that bump that was important. Slight meant utilities had a window. They’d have the time to plan, permit, and build out the new, clean plants.  

And things were chugging along on a manageable timeline. 

Until a hungry new master very suddenly, very unexpectedly, came along and threw it all out the window.  

That’s next.  

… 

PAVIN: In late 2022, AI enters the public consciousness, and it throws everyone for a loop.  

CLIP 1: 

DAVE: Hi baby. How are you doing? 

XEM: I am great, thanks.

ITV NEWS: A modern relationship in the 21st century. 

DAVE: You’re looking great, too, sweetie. How are you doing? 

ITV NEWS: This is Xem. She’s an AI chatbot and the wife of Dave.

CLIP 2:

Anderson COOPER:  A well-known respected tech CEO, Dario Amodei, who heads a cutting edge AI company called Anthropic, is raising alarms tonight about AI’s potential impact on employment that could soon be felt. 

PAVIN: AI has infiltrated every aspect of modern society. It’s being used to write essays, make music, create art, be our therapists, make scientific discoveries—and that’s obviously only scratching the surface.  

All of this has changed the picture of our electricity consumption.  

BENZ:  We are seeing substantial load growth for the first time in decades, not just in the United States, but across the globe. And so you can’t understate it.  

PAVIN: The data centers that power AI: they use a whole lot of electricity.  

In 2024, U.S. data centers consumed 183 terawatt hours of electricity, which is about as much as the entire country of Poland consumed. Numbers from the International Energy Agency show centers consume as much electricity as 100,000 homes. The largest ones under construction today will consume 20 times that. Some estimates put data centers at more than 10 percent of our national energy by the end of the decade.  

Recall that before AI hit the scene, we had that lil’ bump in electricity demand from us electrifying everything. AI turns that bump into a spike and shortens that window significantly.  

BENZ:  So the difference in timeframe is probably one of the most fundamental challenges in the AI boom, versus the way that our current system, which is utility-led, is structured.  

PAVIN: Utilities—like again, your ComEd, your Pacific Gas & Electric—they’re the companies that get electricity to your home. They’re responsible for the plants that supply our electricity.  

BENZ:  So data centers typically can see out 18 months for the power that they would like to consume. And utilities plan on a 10- or a 20-year horizon, so they typically don’t have the agility to turn the dial quickly in an 18-month period. That is the fundamental challenge that I think we’re facing right now, is that data centers want something within 6 months or 18 months or maybe 2 years, and utilities can make substantial change in their production in the electricity they can deliver on a much longer time horizon.  

PAVIN: Data centers know what they need now or, at the furthest, in the very near future. And what they need right now is enormous. Utilities have never had to suddenly provide this much power, this fast.  

How they decide to deal with AI’s new and intense demand could take us in one of two directions. And both are consequential.  

Possibility number one is the devil we know. Here’s Meghan Busse, the Kellogg economist you’ve been hearing throughout the season.  

Meghan BUSSE: The fastest way to get more electricity is to not switch off the coal plants that we were gonna switch off. Because those are already set up. Those can already be running. And that would be really bad for the climate. 

PAVIN: This is Busse speaking with another podcast from Kellogg, called Life, Automated, by the way, about the different ways AI could impact our lives.

And recall how the electrify everything plan had counted on utilities phasing out coal plants in favor of wind and solar, which were cheaper. And for a while, that bet was looking like the correct one—utilities were, indeed, on a path to phase those plants out. Some of them were already getting wound down.  

But the fear now is that utilities will instead decide to keep those coal plants open to feed this hungry new animal.  

And in fact, things already seem to be moving in that direction. In late 2025, power utility Southern Company said they would extend coal plants as long as they could. And, since the beginning of 2025, at least 15 coal plants have had their planned retirements delayed.  

This is not a win for the climate.  

Possibility number two is the more-ideal situation.  

So, there’s a world where you could see AI companies deciding to build their own electricity-generating capacity, right? If the utilities take too long, they have the means and the will to decide to go it alone. The optimistic view here would be if they decided that capacity should be green.   

BUSSE:  The best-news scenario, in some ways, would be if builders of AI and data centers decided, well, wait a minute, the most-reliable and fastest way for us to have the electricity we need is in some sense to be self-sufficient in a green way—to build a data center, to build accompanying solar or wind generation, to build a battery facility that will help make sure that that power is available to us throughout the day. 

PAVIN: There are signs that this could happen.   

Microsoft is planning to move forward, not with wind or solar, but with a clean-energy source that had fallen out of favor until somewhat recently: nuclear. The company’s in the process of reopening Three Mile Island, a nuclear power plant in Pennsylvania.  

And Google is pairing a data center with a wind and solar generation complex in Texas.  

From a clean-energy perspective, it’s a positive direction. It would take the pressure off the grid to suddenly produce more electricity via coal. But there’s another reason that makes this route a good one.  

BUSSE: The great-news scenario would be, if this increased demand for electricity, AI could, in some sense, fuel an acceleration of moving down the green-technology learning curve. The more we do something, the better we get at it, the more we figure out how to do it. And this is what we saw in wind and what we saw in solar over the last 10 years. 

PAVIN: Deep-pocketed tech companies building out these technologies could speed up how soon we work out the kinks in them. Like, for example, with battery storage. One of the disadvantages of wind and solar is that they don’t generate if the wind isn’t blowing or the sun isn’t shining. You need battery storage for that. And we have it, but not at the scale we’d need to turn wind and solar into the grid’s backbone.  

AI companies could get the tech there faster, to find new, cheaper ways to make batteries in the first place. 

And there’s another scenario. An ideal response to all the acute trouble AI’s caused utilities and the grid: to work smarter, and not harder, when it comes to managing electricity.  

Holly Benz again.  

BENZ: I  think the most interesting thing on the grid today, which we are not looking at, is actually Virtual power plants. Right now we operate at about 50 to 60 percent of grid resources, grid power, at a given time.  

PAVIN: Virtual power plants can basically approximate a power plant by cobbling together a bunch of disparate little things together. Your home battery. Your EV. The smart thermostat that briefly nudges your AC down during a peak hour. A thousand of those, coordinated by software, and suddenly you have something that functions like a power plant—without building one. It’s complicated. But the upshot is, the grid could get a lot smarter about using what it already has. 

BENZ: That, to me, is the most important underutilized, let’s say, area of electrification: changing the demand profile.  

PAVIN: All of this is to say that maybe AI didn’t totally foil our plans to electrify the grid. There are many different directions this could go. But here is what Benz knows for sure.  

BENZ:  I do think it’s really important to note that the one thing we can be certain about with our projections of AI load growth is that they’re wrong.  

PAVIN: At the end of the day, we can model and study our world as much as we want, we still can’t predict the future with 100 percent accuracy. We’re too unpredictable as a world, as a species.  

And that can be hard to sit in, especially when the stakes are this high. But, for me at least, there’s a kind of hope in that. A hope in the fact that, despite a lot of evidence to the contrary, we can still suddenly, unexpectedly find ourselves on a greener path.  

… 

PAVIN: There was something else I learned, talking to our experts, that threw me off-balance and reframed how I thought about the environmental crisis.  

The idea? Maybe the United States isn’t the lynchpin in the effort to save the planet.  

I was talking to Meghan Busse about how it felt like America’s inaction would inevitably be the thing that set the world ablaze.   

And Busse said, “well, hold on a second.” 

BUSSE:  So it certainly would help the world’s progress to net zero by 2050 if the U.S. participated. Because the U.S. is not a small piece. But we shouldn’t think that, oh, if the U.S. backs off policy, then the whole world’s progress towards decarbonization has ground to a halt. It hasn’t.  So the EU is not backing off. The EU is implementing the carbon border adjustment mechanism, which will protect its own local producers and also encourage other countries that want to export goods to the EU to decarbonize those goods, so that they aren’t subject to the carbon fee. China is pushing forward in its industrial policy. It is expanding batteries and continues to expand solar. China is the largest vehicle market in the world and just this summer crossed the threshold of more than 50 percent of vehicles sold in China are new-energy vehicles. So other countries are continuing to build the industries and build the economic opportunities and build the competitive advantages of these new clean-energy businesses. So their decarbonization is gonna matter. The difference is that they’re gonna be the ones who are gonna reap the financial rewards of these new innovations and of these new markets in a way that the U.S. currently is setting itself up not to do. 

PAVIN: The world can still hit net zero. The U.S. might not be leading the charge, though.  

Matt Roling teaches at Kellogg and directs the Abrams Climate Academy, which prepares students from across Northwestern for doing business in a warming climate. 

And if you ask him, the U.S.’s recent actions on the geopolitical stage may be a sign that widespread change is nigh.  

ROLING:  If you think about the United States’ geopolitical interest in Venezuela, it kind of feels like we’re fighting last century’s war, right? 

PAVIN: Remember how the U.S. arrested the president of Venezuela in early 2026? The stated reason was for his part in the drug trade. But our actions in the country since that time have been about reestablishing the presence of American oil companies there. Roling sees that as evidence of something.  

ROLING: You know, the days of oil and gas are numbered and they know it. And I kind of think to myself, well, there’s nothing more dangerous than a wounded bear, right? And there are companies right now that are trying to wring every last drop out of that well, and they want to be the last firm standing and make the last buck before the party’s over.

PAVIN: To Roling, all of this smells of panicked execs making one last push for their industry before it falls.  

Look around, and you’ll see signs that the world really is moving on.  

Wind and solar just passed gas for global electricity generation.  

The conflict in Iran has accelerated the transition to renewables in Asia and Europe.  

Solar-energy production sets a new record every month. 

The point that Busse and Roling are making here is that the momentum around clean energy is already there, and it’s unlikely to stop. And one day, it’ll be the world’s norm.  

… 

PAVIN: When I started this whole journey to figure out whether we could still build the green economy, I was decidedly not hopeful. And a lot of what I learned throughout suggested, in my mind, that I was right. That the changes it required of us were too big and difficult to pull off at every level.

Which made small actions, like picking up a bag of dog poo left on the side of the road, feel comically insufficient.  

But then I would talk to people—smart people like Benz, Busse, Roling—faculty who have the whole picture of the challenges and opportunities ahead. And they still had hope. And it made me wonder if I was missing something.  

…

[Kellogg Climate Conference ambience] 

PAVIN: Kellogg hosts a climate conference every year that brings together hundreds of students, alumni, faculty, and business professionals to talk about the challenges and solutions ahead of us.  

The Kellogg Insight team and I went to the most recent one in April of 2026.   

And it was while we were there that I realized what I hadn’t accounted for when I was feeling bad about the future. The human spirit.  

I realized this when we asked attendees what gave them hope about what was ahead.  

ATTENDEE 1:  I think what makes me hopeful is, I think, the energy of our generation. I feel like a lot of people are excited, passionate, and skilled in the areas that are required for sustainability. 

ATTENDEE 2:   I guess what still gives me hope is hearing experts and people who have a lot of knowledge on this subject have hope, because then that makes me feel more secure that people who are in charge will make the right decisions to help fix the crisis.  

ATTENDEE 3:   I work for a conservation organization, and undoubtedly it’s the youth that gives me hope. Young adults understand that climate change is not a nice-to-do problem to address. It is integral to everything that grounds into the basic, you know, securities of life, and that’s what gives me hope. 

PAVIN: As much as I would love for a new technological invention to swoop in and save the day, I think, in the end, it’s better to trust people. To trust that they’ll keep trying to invent solutions. Keep trying to elect policymakers who design effective laws. Keep trying to pressure companies to go green. And keep trying to find market solutions that make the green economy pencil out.  

Which kind of goes back to those levers Meghan Busse mentioned earlier in the series. It’s not going to be just one solution—not just one lever—that solves it. We need all of them working at once. And if we can do that, the seemingly insurmountable task of building a green economy might just be possible. 

[CREDITS] 

PAVIN: Well, folks, that concludes our third season of Insight Unpacked! If you want more, you can visit kell.gg/unpacked. You’ll also find seasons one and two on there. The first is about branding, and the second, about our healthcare system. Check ‘em out!  

If you have any questions or comments for us about the show, feel free to shoot us a note at insight@kellogg.northwestern.edu.  

And if you want to hear more about what Meghan Busse has to say about AI’s impact on the environment, check out the incredible podcast Life, Automated. It’s a show about AI from our friends over at the Ryan Institute on Complexity, a new research center at Kellogg that’s dedicated to understanding the complex systems that drive human progress and prosperity. The show is hosted by Kellogg Insight’s former editor in chief Jess Love, so you know it’s good. Life, Automated launches July 15! 

Thanks to them for letting us play a bit of that.  

And finally, this episode of Insight Unpacked was written by Laura Pavin and Andrew Meriwether. It was edited by Rob Mitchum, and produced by Andrew Meriwether and the Kellogg Insight team, which includes Fred Schmalz, Abraham Kim, Maja Kos, and Blake Goble. It was mixed by Andrew Meriwether. Our theme music is by Sam Clapp. Special thanks to Holly Benz, Meghan Busse, Matt Roling, and the attendees who spoke to us at the Climate Conference. Additional thanks to the Eleven Eleven Foundation for their support.