You’re out of free articles.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:
It took the market about a week to catch up to the fact that the Chinese artificial intelligence firm DeepSeek had released an open-source AI model that rivaled those from prominent U.S. companies such as OpenAI and Anthropic — and that, most importantly, it had managed to do so much more cheaply and efficiently than its domestic competitors. The news cratered not only tech stocks such as Nvidia, but energy stocks, as well, leading to assumptions that investors thought more-energy efficient AI would reduce energy demand in the sector overall.
But will it really? While some in climate world assumed the same and celebrated the seemingly good news, many venture capitalists, AI proponents, and analysts quickly arrived at essentially the opposite conclusion — that cheaper AI will only lead to greater demand for AI. The resulting unfettered proliferation of the technology across a wide array of industries could thus negate the energy efficiency gains, ultimately leading to a substantial net increase in data center power demand overall.
“With cost destruction comes proliferation,” Susan Su, a climate investor at the venture capital firm Toba Capital, told me. “Plus the fact that it’s open source, I think, is a really, really big deal. It puts the power to expand and to deploy and to proliferate into billions of hands.”
If you’ve seen lots of chitchat about Jevons paradox of late, that’s basically what this line of thinking boils down to. After Microsoft’s CEO Satya Nadella responded to DeepSeek mania by posting the Wikipedia page for this 19th century economic theory on X, many (myself included) got a quick crash course on its origins. The idea is that as technical efficiencies of the Victorian era made burning coal cheaper, demand for — and thus consumption of — coal actually increased.
While this is a distinct possibility in the AI space, it’s by no means a guarantee. “This is very much, I think, an open question,“ energy expert Nat Bullard told me, with regards to whether DeepSeek-type models will spur a reduction or increase in energy demand. “I sort of lean in both directions at once.” Formerly the chief content officer at BloombergNEF and current co-founder of the AI startup Halcyon, a search and information platform for energy professionals, Bullard is personally excited for the greater efficiencies and optionality that new AI models can bring to his business.
But he warns that just because DeepSeek was cheap to train — the company claims it cost about $5.5 million, while domestic models cost hundreds of millions or even billions — doesn’t mean that it’s cheap or energy-efficient to operate. “Training more efficiently does not necessarily mean that you can run it that much more efficiently,” Bullard told me. When a large language model answers a question or provides any type of output, it’s said to be making an “inference.” And as Bullard explains, “That may mean, as we move into an era of more and more inference and not just training, then the [energy] impacts could be rather muted.”
DeepSeek-R1, the name for the model that caused the investor freakout, is also a newer type of LLM that uses more energy in general. Up until literally a few days ago, when OpenAI released o3-mini for free, most casual users were probably interacting with so-called “pretrained” AI models. Fed on gobs of internet text, these LLMs spit out answers based primarily on prediction and pattern recognition. DeepSeek released a model like this, called V3, in September. But last year, more advanced “reasoning” models, which can “think,” in some sense, started blowing up. These models — which include o3-mini, the latest version of Anthropic’s Claude, and the now infamous DeepSeek-R1 — have the ability to try out different strategies to arrive at the correct answer, recognize their mistakes, and improve their outputs, allowing for significant advancements in areas such as math and coding.
But all that artificial reasoning eats up a lot of energy. As Sasha Luccioni, the AI and climate lead at Hugging Face, which makes an open-source platform for AI projects, wrote on LinkedIn, “To set things clear about DeepSeek + sustainability: (it seems that) training is much shorter/cheaper/more efficient than traditional LLMs, *but* inference is longer/more expensive/less efficient because of the chain of thought aspect.” Chain of thought refers to the reasoning process these newer models undertake. Luccioni wrote that she’s currently working to evaluate the energy efficiency of both the DeepSeek V3 and R1 models.
Another factor that could influence energy demand is how fast domestic companies respond to the DeepSeek breakthrough with their own new and improved models. Amy Francetic, co-founder at Buoyant Ventures, doesn’t think we’ll have to wait long. “One effect of DeepSeek is that it will highly motivate all of the large LLMs in the U.S. to go faster,” she told me. And because a lot of the big players are fundamentally constrained by energy availability, she’s crossing her fingers that this means they’ll work smarter, not harder. “Hopefully it causes them to find these similar efficiencies rather than just, you know, pouring more gasoline into a less fuel-efficient vehicle.”
In her recent Substack post, Su described three possible futures when it comes to AI’s role in the clean energy transition. The ideal is that AI demand scales slowly enough that nuclear and renewables scale with it. The least hopeful is that immediate, exponential growth in AI demand leads to a similar expansion of fossil fuels, locking in new dirty infrastructure for decades. “I think that's already been happening,” Su told me. And then there’s the techno-optimist scenario, linked to figures like Sam Altman, which Su doesn’t put much stock in — that AI “drives the energy revolution” by helping to create new energy technologies and efficiencies that more than offset the attendant increase in energy demand.
Which scenario predominates could also depend upon whether greater efficiencies, combined with the adoption of AI by smaller, more shallow-pocketed companies, leads to a change in the scale of data centers. “There’s going to be a lot more people using AI. So maybe that means we don’t need these huge, gigawatt data centers. Maybe we need a lot more smaller, megawatt-size data centers,” Laura Katzman, a principal at Buoyant Ventures, told me. Katzman has conducted research for the firm on data center decarbonization.
Smaller data centers with a subsequently smaller energy footprint could pair well with renewable-powered microgrids, which are less practical and economically feasible for hyperscalers. That could be a big win for solar and wind plus battery storage, Katzman explained, but a boondoggle for companies such as Microsoft, which has famously committed to re-opening Pennsylvania’s Three Mile Island nuclear plant to power its data centers. “Because of DeepSeek, the expected price of compute probably doesn’t justify now turning back on some of these nuclear plants, or these other high-cost energy sources,” Katzman told me.
Lastly, it remains to be seen what nascent applications cheaper models will open up. “If somebody, say, in the Philippines or Vietnam has an interest in applying this to their own decarbonization challenge, what would they come up with?” Bullard pondered. “I don’t yet know what people would do with greater capability and lower costs and a different set of problems to solve for. And that’s really exciting to me.”
But even if the AI pessimists are right, and these newer models don’t make AI ubiquitously useful for applications from new drug discovery to easier regulatory filing, Su told me that in a certain sense, it doesn't matter much. “If there was a possibility that somebody had this type of power, and you could have it too, would you sit on the couch? Or would you arms race them? I think that is going to drive energy demand, irrespective of end utility.”
As Su told me, “I do not think there’s actually a saturation point for this.”
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Any household savings will barely make a dent in the added costs from Trump’s many tariffs.
Donald Trump’s tariffs — the “fentanyl” levies on Canada, China, and Mexico, the “reciprocal” tariffs on nearly every country (and some uninhabited islands), and the global 10% tariff — will almost certainly cause consumer goods on average to get more expensive. The Yale Budget Lab estimates that in combination, the tariffs Trump has announced so far in his second term will cause prices to rise 2.3%, reducing purchasing power by $3,800 per year per household.
But there’s one very important consumer good that seems due to decline in price.
Trump administration officials — including the president himself — have touted cheaper oil to suggest that the economic response to the tariffs hasn’t been all bad. On Sunday, Secretary of the Treasury Scott Bessent told NBC, “Oil prices went down almost 15% in two days, which impacts working Americans much more than the stock market does.”
Trump picked up this line on Truth Social Monday morning. “Oil prices are down, interest rates are down (the slow moving Fed should cut rates!), food prices are down, there is NO INFLATION,” he wrote. He then spent the day posting quotes from Fox Business commentators echoing that idea, first Maria Bartiromo (“Rates are plummeting, oil prices are plummeting, deregulation is happening. President Trump is not going to bend”) then Charles Payne (“What we’re not talking about is, oil was $76, now it’s $65. Gasoline prices are going to plummet”).
But according to Neil Dutta, head of economic research at Renaissance Macro Research, pointing to falling oil prices as a stimulus is just another example of the “4D chess” theory, under which some market participants attribute motives to Trump’s trade policy beyond his stated goal of reducing trade deficits to as near zero (or surplus!) as possible.
Instead, oil markets are primarily “responding to the recession risk that comes from the tariff and the trade war,” Dutta told me. “That is the main story.” In short, oil markets see less global trade and less global production, and therefore falling demand for oil. The effect on household consumption, he said, was a “second order effect.”
It is true that falling oil prices will help “stabilize consumption,” Dutta told me (although they could also devastate America’s own oil industry). “It helps. It’ll provide some lift to real income growth for consumers, because they’re not spending as much on gasoline.” But “to fully offset the trade war effects, you basically need to get oil down to zero.”
That’s confirmed by some simple and extremely back of the envelope math. In 2023, households on average consumed about 700 gallons of gasoline per year, based on Energy Information Administration calculations that the average gasoline price in 2023 was $3.52, while the Bureau of Labor Statistics put average household gasoline expenditures at about $2,450.
Let’s generously assume that due to the tariffs and Trump’s regulatory and diplomatic efforts, gas prices drop from the $3.26 they were at on Monday, according to AAA, to $2.60, the average price in 2019. (GasBuddy petroleum analyst Patrick De Haanwrote Monday that the tariffs combined with OPEC+ production hikes could lead gas prices “to fall below $3 per gallon.”)
Let’s also assume that this drop in gas prices does not cause people to drive more or buy less fuel-efficient vehicles. In that case, those same 700 gallons cost the average American $1,820, which would generate annual savings of $630 on average per household. If we went to the lowest price since the Russian invasion of Ukraine, about $3 per gallon, total consumption of 700 gallons would cost a household about $2,100, saving $350 per household per year.
That being said, $1,820 is a pretty low level for annual gasoline consumption. In 2021, as the economy was recovering from the Covid recession and before gas prices popped, annual gasoline expenditures only got as low as $1,948; in 2020 — when oil prices dropped to literally negative dollars per barrel and gas prices got down to $1.85 a gallon — annual expenditures were just over $1,500.
In any case, if you remember the opening paragraphs of this story, even the most generous estimated savings would go nowhere near surmounting the overall rise in prices forecast by the Yale Budget Lab. $630 is less than $3,800! (JPMorgan has forecast a more mild increase in prices of 1% to 1.5%, but agrees that prices will likely rise and purchasing power will decline.)
But maybe look at it this way: You might be able to drive a little more than you expected to, even as your costs elsewhere are going up. Just please be careful! You don’t want to get into a bad accident and have to replace your car: New car prices are expected to rise by several thousand dollars due to Trump’s tariffs.
With cars about to get more expensive, it might be time to start tinkering.
More than a decade ago, when I was a young editor at Popular Mechanics, we got a Nissan Leaf. It was a big deal. The magazine had always kept long-term test cars to give readers a full report of how they drove over weeks and months. A true test of the first true production electric vehicle from a major car company felt like a watershed moment: The future was finally beginning. They even installed a destination charger in the basement of the Hearst Corporation’s Manhattan skyscraper.
That Leaf was a bit of a lump, aesthetically and mechanically. It looked like a potato, got about 100 miles of range, and delivered only 110 horsepower or so via its electric motors. This made the O.G. Leaf a scapegoat for Top Gear-style car enthusiasts eager to slander EVs as low-testosterone automobiles of the meek, forced upon an unwilling population of drivers. Once the rise of Tesla in the 2010s had smashed that paradigm and led lots of people to see electric vehicles as sexy and powerful, the original Leaf faded from the public imagination, a relic of the earliest days of the new EV revolution.
Yet lots of those cars are still around. I see a few prowling my workplace parking garage or roaming the streets of Los Angeles. With the faded performance of their old batteries, these long-running EVs aren’t good for much but short-distance city driving. Ignore the outdated battery pack for a second, though, and what surrounds that unit is a perfectly serviceable EV.
That’s exactly what a new brand of EV restorers see. Last week, car site The Autopiancovered DIYers who are scooping up cheap old Leafs, some costing as little as $3,000, and swapping in affordable Chinese-made 62 kilowatt-hour battery units in place of the original 24 kilowatt-hour units to instantly boost the car’s range to about 250 miles. One restorer bought a new battery on the Chinese site Alibaba for $6,000 ($4,500, plus $1,500 to ship that beast across the sea).
The possibility of the (relatively) simple battery swap is a longtime EV owner’s daydream. In the earlier days of the electrification race, many manufacturers and drivers saw simple and quick battery exchange as the solution for EV road-tripping. Instead of waiting half an hour for a battery to recharge, you’d swap your depleted unit for a fully charged one and be on your way. Even Tesla tested this approach last decade before settling for good on the Supercharger network of fast-charging stations.
There are still companies experimenting with battery swaps, but this technology lost. Other EV startups and legacy car companies that followed Nissan and Tesla into making production EVs embraced the rechargeable lithium-ion battery that is meant to be refilled at a fast-charging station and is not designed to be easily removed from the vehicle. Buy an electric vehicle and you’re buying a big battery with a long warranty but no clear plan for replacement. The companies imagine their EVs as something like a smartphone: It’s far from impossible to replace the battery and give the car a new life, but most people won’t bother and will simply move on to a new car when they can’t take the limitations of their old one anymore.
I think about this impasse a lot. My 2019 Tesla Model 3 began its life with a nominal 240 miles of range. Now that the vehicle has nearly six years and 70,000 miles on it, its maximum range is down to just 200, while its functional range at highway speed is much less than that. I don’t want to sink money into another vehicle, which means living with an EV’s range that diminishes as the years go by.
But what if, one day, I replaced its battery? Even if it costs thousands of dollars to achieve, a big range boost via a new battery would make an older EV feel new again, and at a cost that’s still far less than financing a whole new car. The thought is even more compelling in the age of Trump-imposed tariffs that will raise already-expensive new vehicles to a place that’s simply out of reach for many people (though new battery units will be heavily tariffed, too).
This is no simple weekend task. Car enthusiasts have been swapping parts and modifying gas-burning vehicles since the dawn of the automotive age, but modern EVs aren’t exactly made with the garage mechanic in mind. Because so few EVs are on the road, there is a dearth of qualified mechanics and not a huge population of people with the savvy to conduct major surgery on an electric car without electrocuting themselves. A battery-replacing owner would need to acquire not only the correct pack but also potentially adapters and other equipment necessary to make the new battery play nice with the older car. Some Nissan Leaf modifiers are finding their replacement packs aren’t exactly the same size, shape or weight, The Autopian says, meaning they need things like spacers to make the battery sit in just the right place.
A new battery isn’t a fix-all either. The motors and other electrical components wear down and will need to be replaced eventually, too. A man in Norway who drove his Tesla more than a million miles has replaced at least four battery packs and 14 motors, turning his EV into a sort of car of Theseus.
Crucially, though, EVs are much simpler, mechanically, than combustion-powered cars, what with the latter’s belts and spark plugs and thousands of moving parts. The car that surrounds a depleted battery pack might be in perfectly good shape to keep on running for thousands of miles to come if the owner were to install a new unit, one that could potentially give the EV more driving range than it had when it was new.
The battery swap is still the domain of serious top-tier DIYers, and not for the mildly interested or faint of heart. But it is a sign of things to come. A market for very affordable used Teslas is booming as owners ditch their cars at any cost to distance themselves from Elon Musk. Old Leafs, Chevy Bolts and other EVs from the 2010s can be had for cheap. The generation of early vehicles that came with an unacceptably low 100 to 150 miles of range would look a lot more enticing if you imagine today’s battery packs swapped into them. The possibility of a like-new old EV will look more and more promising, especially as millions of Americans realize they can no longer afford a new car.
On the shifting energy mix, tariff impacts, and carbon capture
Current conditions: Europe just experienced its warmest March since record-keeping began 47 years ago • It’s 105 degrees Fahrenheit in India’s capital Delhi where heat warnings are in effect • The risk of severe flooding remains high across much of the Mississippi and Ohio Valleys.
The severe weather outbreak that has brought tornadoes, extreme rainfall, hail, and flash flooding to states across the central U.S. over the past week has already caused between $80 billion and $90 billion in damages and economic losses, according to a preliminary estimate from AccuWeather. The true toll is likely to be costlier because some areas have yet to report their damages, and the flooding is ongoing. “A rare atmospheric river continually resupplying a firehose of deep tropical moisture into the central U.S., combined with a series of storms traversing the same area in rapid succession, created a ‘perfect storm’ for catastrophic flooding and devastating tornadoes,” said AccuWeather’s chief meteorologist Jonathan Porter. The estimate takes into account damages to buildings and infrastructure, as well as secondary effects like supply chain and shipping disruptions, extended power outages, and travel delays. So far 23 people are known to have died in the storms. “This is the third preliminary estimate for total damage and economic loss that AccuWeather experts have issued so far this year,” the outlet noted in a release, “outpacing the frequency of major, costly weather disasters since AccuWeather began issuing estimates in 2017.”
AccuWeather
Low-emission energy sources accounted for 41% of global electricity generation in 2024, up from 39.4% in 2023, according to energy think tank Ember’s annual Global Electricity Review. That includes renewables as well as nuclear. If nuclear is left out of the equation, renewables alone made up 32% of power generation last year. Overall, renewables added a record 858 terawatt hours, nearly 50% more than the previous record set in 2022. Hydro was the largest source of low-carbon power, followed by nuclear. But wind and solar combined overtook hydro last year, while nuclear’s share of the energy mix reached a 45-year low. More solar capacity was installed in 2024 than in any other single year.
Ember
The report notes that demand for electricity rose thanks to heat waves and air conditioning use. This resulted in a slight, 1.4% annual increase in fossil-fuel power generation and pushed power-sector emissions to a new all-time high of 14.5 billion metric tons. “Clean electricity generation met 96% of the demand growth not caused by hotter temperatures,” the report said.
President Trump’s new tariffs will have a “limited” effect on the amount of solar components the U.S. imports from Asia because the U.S. already imposes tariffs on these products, according to a report from research firm BMI. That said, the U.S. still relies heavily on imported solar cells, and the new fees are likely to raise costs for domestic manufacturers and developers, which will ultimately be passed on to buyers and could slow solar growth. “Since the U.S.’s manufacturing capacity is insufficient to meet demand for solar, wind, and grid components, we do expect that costs will increase for developers due to the tariffs which will now be imposed upon these components,” BMI wrote.
In other tariff news, the British government is adjusting its 2030 target of ending the sale of new internal combustion engine cars to ease some of the pain from President Trump’s new 25% auto tariffs. Under the U.K.’s new EV mandate, carmakers will be able to sell new hybrids through 2035 (whereas the previous version of the rules banned them by 2030), and gas and diesel vans can also be sold through 2035. The changes also carve out exemptions for luxury supercar brands like McLaren and Aston Martin, which will be allowed to keep selling new ICE vehicles beyond 2030 because, the government says, they produce so few. The goal is to “help ease the transition and give industry more time to prepare.” British Transport Secretary Heidi Alexander insisted the changes have been “carefully calibrated” and their impact on carbon emissions is “negligible.” As The New York Timesnoted, the U.S. is the largest single-country export market for British cars.
The Environmental Protection Agency has approved Occidental Petroleum’s application to capture and sequester carbon dioxide at its direct air capture facility in Texas, and issued permits that will allow the company to drill and inject the gas more than one mile underground. The Stratos DAC plant is being developed by Occidental subsidiary 1PointFive. As Heatmap’s Katie Brigham has reported, Stratos is designed to remove up to 500,000 metric tons of CO2 annually and set to come online later this year. Its success (or failure) could shape the future of DAC investment at a time when the Trump administration is hollowing out the Department of Energy’s nascent Carbon Dioxide Removal team and casting doubt over the future of the DOE’s $3.5 billion Regional Direct Air Capture Hubs program. While Stratos is not a part of the hubs program, it will use the same technology as Occidental’s South Texas DAC hub.
The Bezos Earth Fund and the Global Methane Hub are launching a $27 million effort to fund research into selectively breeding cattle that emit less methane.