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:
A conversation with Ben Goldfarb about his road ecology book Crossings.
An alternative title for journalist Ben Goldfarb’s fantastic new book, Crossings, could have been Squashings. “Wait a minute,” I thought to myself about 25 pages in. “Have I been duped into reading a book about … roadkill?!”
The answer wasn’t precisely no, although Crossings is also about so much more (its subtitle: How Road Ecology Is Shaping the Future of Our Planet). From cliff swallows that have evolved to have shorter wings to better avoid zooming cars, to Oedipal cougars stranded in the highway-wrapped Santa Monica Mountains, to the trials of one surprisingly charismatic anteater named Evelyn, Crossings observes that “the repercussions of roads are so complex that it’s hard to pinpoint where they end.”
Goldfarb, though, attempts valiantly to untangle them, and the result is as funny, heartbreaking, enraging, and enlightening as anything I’ve read this year. “There may be nothing humans do that causes more misery to more wild animals than driving,” he writes, but planet-warming emissions are only the most prominent part of that story. Ahead of Crossings’ publication next Tuesday, Goldfarb and I discussed the promise (and drawbacks) of the EV transition and autonomous cars on road ecology; the short-sightedness of infrastructure budgets; and how bad people are at driving. Our conversation has been condensed and edited for clarity.
When you told people you were working on a book about road ecology, did they take it as an invitation to share their personal, unsolicited roadkill stories with you?
Absolutely, they did. I wouldn’t say it was unsolicited: I’m always — I don’t want to say I’m happy to hear roadkill stories — but I’m certainly interested in stories and there were lots of them. One of my favorite ones was a guy who told me that he’d recently hit a squirrel and he was so confused and upset and unhappy about it that he actually called 911. He didn’t know what else to do. And the 911 operator basically said, “Uh yeah, the squirrel is dead.” I mean, to me, that sort of gets at how viscerally upsetting and disturbing roadkill can be. It’s something we see constantly and ultimately take for granted in a lot of ways but committing it ourselves is, of course, a miserable feeling. I just hit an owl a few nights ago and I’m still losing sleep.
One of the things I was most astonished by while reading this book is how well-sourced it is — the texts and interviews you bring together are so broad and enriching. Do you have any idea how many books you read? Tell me a little about how you approached the research.
Oh, geez. Let’s see — two shelves of that bookshelf [behind me] are road ecology reference books. So, several dozen. I can’t claim that I read them all cover to cover, but certainly I drew a lot from other books. I think I ultimately had close to 300 sources in the book who were just invaluable founts of help and knowledge and information.
One of the challenges of writing about road ecology is it’s not necessarily a single discipline. It’s really an umbrella that covers many different disciplines. Roadkill science is its own sort of subset. The impact of forest service roads on contributing erosion to streams is a whole science unto itself. The impact of improperly built road culverts as fish passage barriers — I mean, there are 10,000 papers about that alone. So every chapter was sort of learning a new science unto itself.
You write that “among all the road’s ecological disasters … the most vexing may be noise pollution.” We do a lot of coverage of the future of driving here at Heatmap, and I suppose I was hoping to learn that electric vehicles and cutting-edge advances in automotive technology would help solve at least this problem. Can you tell me why you’re less optimistic?
EVs are much quieter; their engines are silent, which is helpful, especially in an urban context. They’ll ultimately reduce noise pollution and that’s profoundly important. We tend to overlook noise pollution because we’re so awash in it but it’s one of the great public health crises of our time. You read the literature about the health impacts of road noise and it’s horrifying — I mean, literally, it’s elevating our stress levels, it’s increasing our risk of heart attack and diabetes and stroke, it’s taking years off of our lives, mostly without our noticing it. So anything we can do to reduce noise is fundamentally positive. And EVs are part of that.
The drawback, the reason that EVs aren’t a panacea, is that engines aren’t the only thing that makes noise on a car. Above 35 mph, most of what you’re hearing is tire noise: the grinding of the tire itself against the pavement and the little air pockets in the tread popping — “pattern noise” is what that’s called. I wrote most of this book while living a half mile or so from I-90 in eastern Washington state and I could just hear, every time I stepped out of my house, that monotonous hiss of the interstate. That’s tire noise, not engine noise. And tires have gotten much quieter over time, which is good, and hopefully they’ll continue to get quieter, but just electrifying vehicles is not going to solve the problem of road noise even if it does help in urban settings.
Not to keep raining on the parade, but you also write that autonomous vehicles could be “the gravest challenge to road ecology since, well, roads.” How do driverless cars change the road ecology calculus?
I think the answer is, we don’t know yet. From a large animal avoidance perspective, I think they’re ultimately going to be really helpful. Yes, it’s fun right now to dunk on Tesla and Waymo and all of these autonomous vehicle companies whose products are still very buggy, but, you know — probably there are people who will read this and take exception with this idea, but I’m ultimately pretty optimistic that the AVs will solve most of those problems and become better drivers than human beings.
And that’s the thing that always gets lost when somebody posts a video of an AV doing something stupid — human drivers do stupid things constantly, right? We’re horrifically bad drivers. Tens of thousands of people die in the U.S. every year because of it. And one of the things that we’re really bad at is avoiding large animals. We don’t see that well at night, they jump out unexpectedly, and our reflexes are too slow to slam on the brakes. I think that AVs will be much, much better at avoiding those deer and elk and moose than we are because those are large animals and all of [the AV] sensors that are designed to avoid pedestrians will be triggered by those large animals.
But, of course, that doesn’t really help a rattlesnake or a prairie dog or any smaller creature. I, for one, go out of my way to avoid hitting those animals, and when my car is piloted by a robot, that’s not going to help; that robot will have no reason to avoid those small animals if engineers don’t design it to do so.
And the broader problem is that autonomy is likely to lead to a whole lot more vehicles on the road. When you can get in your car and it drives itself and you can spend that time watching movies or doing work or what have you, commuting becomes a lot less onerous. Every autonomous vehicle could have a kid in it who’s not able to drive currently. Most of the modeling suggests that there’s going to be a dramatic increase in vehicle miles traveled as a result of autonomous cars. And that’s going to be bad for wildlife, that’s going to make the barrier effect of roads even more severe and make it even harder for animals to migrate across highways.
And commuting traffic, human traffic, is really just the tip of the iceberg when it comes to autonomy. The autonomous delivery fleet, in some ways, is the bigger concern. A lot of the early AVs are going to be delivery vehicles; it’s going to be so easy to summon products to us. So it’s hard to imagine a scenario where AVs lead to less driving rather than more of it, unfortunately.
How did you navigate striking the right balance between the ideals of conservation and the realities of politics and economics in this book? I found myself getting so frustrated reading about the frogs trying to cross Highway 30 in Portland, Oregon, only to then learn that SP-139 in Brazil actually closes a section between 8 p.m. and 6 a.m., when animals are most active. I was like, “Why can’t we do that!”
We do have this very constrained idea of what is possible and that’s why I like drawing upon other countries. You mentioned that road in Brazil that is closed at night through a park; another great anecdote is that in India, they built a new highway through a tiger sanctuary and they just elevated the entire highway on pilings so that animals can come and go underneath the lifted freeway. Of course, that made the project vastly more expensive, but it’s ecologically the right thing to do and is much more radical than anything we’ve done in this country.
I was just talking about this the other day with somebody in the bird ecology world: how our sense of what we can afford is so skewed. I think that people hear the price tag of a wildlife crossing structure and they think, “Oh my gosh, $10 million just to help elk cross the highway, what an extravagant expenditure.” But that’s beyond nothing in the context of national, state, and federal transportation budgets. I mean, $10 million for a wildlife crossing, that’s not even a drop in the bucket. That’s like a molecule of H2O in the bucket. It costs a million dollars to pave a mile of highway, let alone add a bunch of lanes to it. So to me, the notion that we can’t make our infrastructure better for nature because it costs money is incredibly short-sighted and fails to consider how much money we’re spending on our roads already.
A great example of that was the Infrastructure Act, which contains $350 million for wildlife crossings — which is great and wonderful and a step in the right direction. But it also contains billions of dollars for highway expansions and repaving and bridge repairs. And one bird ecologist described that $350 million as “decimal dust,” you know, just nothing in the context of federal transportation. The politics of the possible can definitely be frustrating.
Not to mention, you have a statistic in Crossings that animal crashes cost America something like $8 billion per year.
And that was $8 billion in 2009. So for inflation and accounting for increased collisions over time — yeah, it’s an enormous number that we’re not doing a whole lot about.
Your book is full of so much humor and cautious optimism but when I was reading it, I would sometimes get overwhelmed just thinking about how many roads exist and how many more roads are going to exist and the awful ends so many living things meet because of them. How did you stay hopeful while immersed in these stories?
I think that the book comes off as humorous and optimistic because that’s just my natural register as a writer, but I’m not sure I actually always feel that way. There are times that I feel totally desperate about the future of conservation. One of the challenges of writing about this topic is that there’s no perfect solution, there’s no panacea. We could say “we need more mass transit,” and certainly we need to get people out of cars, but I live in rural Colorado: It’s hard to imagine a public transportation system that is going to meaningfully change driving rates in this kind of very rural, dispersed area that was built around the automobile.
Wildlife crossings are the same thing. They help a specific set of problems, which is roadkill and the curtailment of animal migration. But they don’t reduce road noise, they don’t prevent tire particles from spewing into the environment and killing salmon, they don’t do anything about road salts being applied in ridiculous quantities and destroying freshwater ecosystems. So, again, there is no panacea here and it can be really challenging to confront the scale and the number of different solutions needed to make our roads lie more lightly on the planet.
Is there anything else you would want readers to know about Crossings?
You mentioned EVs in the context of road noise and one of the things that I almost wish I had emphasized more in the book is that when people tend to think about the environmental impacts of transportation, they think about the carbon emissions, right? And the solutions tend to be things like the electrification of vehicle fleets and fuel standards. And certainly, those are good things. But the electrification of the fleet is going to do absolutely nothing for wild animals. In fact, just as AVs could lead to more driving, EVs can do the same thing when it becomes much cheaper to drive your car because you just have to plug it in — the whole Jevons paradox idea that a million EV scholars have written about.
I feel like part of the purpose of the book is to say, look, the carbon emissions from transportation are an enormous problem. But they’re only one of the many, many ecological problems that our car-centered transportation network causes. You can strip the carbon out of our transportation and still not make it benign for the environment.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
The multi-faceted investment is defense-oriented, but could also support domestic clean energy.
MP Materials is the national champion of American rare earths, and now the federal government is taking a stake.
The complex deal, announced Thursday, involves the federal government acting as a guaranteed purchaser of MP Materials’ output, a lender, and also an investor in the company. In addition, the Department of Defense agreed to a price floor for neodymium-praseodymium products of $110 per kilogram, about $50 above its current spot price.
MP Materials owns a rare earths mine and processing facility near the California-Nevada border on the edges of the Mojave National Preserve. It claims to be “the largest producer of rare earth materials in the Western Hemisphere,” with “the only rare earth mining and processing site of scale in North America.”
As part of the deal, the company will build a “10X Facility” to produce magnets, which the DOD has guaranteed will be able to sell 100% of its output to some combination of the Pentagon and commercial customers. The DOD is also kicking in $150 million worth of financing for MP Materials’ existing processing efforts in California, alongside $1 billion from Wall Street — specifically JPMorgan Chase and Goldman Sachs — for the new magnet facility. The company described the deal in total as “a multi-billion-dollar commitment to accelerate American rare earth supply chain independence.”
Finally, the DOD will buy $400 million worth of newly issued stock in MP Materials, giving it a stake in the future production that it’s also underwriting.
Between the equity investment, the lending, and the guaranteed purchasing, the Pentagon, and by extension the federal government, has taken on considerable financial risk in casting its lot with a company whose primary asset’s previous owner went bankrupt a decade ago. But at least so far, Wall Street is happy with the deal: MP Materials’ market capitalization soared to over $7 billion on Thursday after its share price jumped over 40%, from a market capitalization of around $5 billion on Wednesday and the company is valued at around $7.5 billion as of Friday afternoon.
Despite the risk, former Biden administration officials told me they would have loved to make a deal like this.
When I asked Alex Jacquez, who worked on industrial policy for the National Economic Council in the Biden White House, whether he wished he could’ve overseen something like the DOD deal with MP Materials, he replied, “100%.” I put the same question to Ashley Zumwalt-Forbes, a former Department of Energy official who is now an investor; she said, “Absolutely.”
Rare earths and critical minerals were of intense interest to the Biden administration because of their use in renewable energy and energy storage. Magnets made with neodymium-praseodymium oxide are used in the electric motors found in EVs and wind turbines, as well as for various applications in the defense industry.
MP Materials will likely have to continue to rely on both sets of customers. Building up a real domestic market for the China-dominated industry will likely require both sets of buyers. According to a Commerce Department report issued in 2022, “despite their importance to national security, defense demand for … magnets is only a small portion of overall demand and insufficient to support an economically viable domestic industry.”
The Biden administration previously awarded MP Materials $58.5 million in 2024 through the Inflation Reduction Act’s 48C Advanced Energy Project tax credit to support the construction of a magnet facility in Fort Worth. While the deal did not come with the price guarantees and advanced commitment to purchase the facility’s output of the new agreement, GM agreed to come on as an initial buyer.
Matt Sloustcher, an MP Materials spokesperson, confirmed to me that the Texas magnet facility is on track to be fully up and running by the end of this year, and that other electric vehicle manufacturers could be customers of the new facility announced on Thursday.
At the time MP Materials received that tax credit award, the federal government was putting immense resources behind electric vehicles, which bolstered the overall supply supply chain and specifically demand for components like magnets. That support is now being slashed, however, thanks to the One Big Beautiful Bill Act, which will cancel consumer-side subsidies for electric vehicle purchases.
While the Biden tax credit deal and the DOD investment have different emphases, they both follow on years of bipartisan support for MP Materials. In 2020, the DOD used its authority under the Defense Production Act to award almost $10 million to MP Materials to support its investments in mineral refining. At the time, the company had been ailing in part due to retaliatory tariffs from China, cutting off the main market for its rare earths. The company was shipping its mined product to China to be refined, processed, and then used as a component in manufacturing.
“Currently, the Company sells the vast majority of its rare earth concentrate to Shenghe Resources,” MP Materials the company said in its 2024 annual report, referring to a Chinese rare earths company.
The Biden administration continued and deepened the federal government’s relationship with MP Materials, this time complementing the defense investments with climate-related projects. In 2022, the DOD awarded a contract worth $35 million to MP Materials for its processing project in order to “enable integration of [heavy rare earth elements] products into DoD and civilian applications, ensuring downstream [heavy rare earth elements] industries have access to a reliable feedstock supplier.”
While the DOD deal does not mean MP Materials is abandoning its energy customers or focus, the company does appear to be to the new political environment. In its February earnings release, the company mentioned “automaker” or “automotive-grade magnets” four times; in its May earnings release, that fell to zero times.
Former Biden administration officials who worked on critical minerals and energy policy are still impressed.
The deal is “a big win for the U.S. rare earths supply chain and an extremely sophisticated public-private structure giving not just capital, but strategic certainty. All the right levers are here: equity, debt, price floor, and offtake. A full-stack solution to scale a startup facility against a monopoly,” Zumwalt-Forbes, the former Department of Energy official, wrote on LinkedIn.
While the U.S. has plentiful access to rare earths in the ground, Zumwalt-Forbes told me, it has “a very underdeveloped ability to take that concentrate away from mine sites and make useful materials out of them. What this deal does is it effectively bridges that gap.”
The issue with developing that “midstream” industry, Jacquez told me, is that China’s world-leading mining, processing, and refining capacity allows it to essentially crash the price of rare earths to see off foreign competitors and make future investment in non-Chinese mining or processing unprofitable. While rare earths are valuable strategically, China’s whip hand over the market makes them less financially valuable and deters investment.
“When they see a threat — and MP is a good example — they start ramping up production,” he said. Jacquez pointed to neodymium prices spiking in early 2022, right around when the Pentagon threw itself behind MP Materials’ processing efforts. At almost exactly the same time, several state-owned Chinese rare earth companies merged. Neodymium-praseodymium oxide prices fell throughout 2022 thanks to higher Chinese production quotas — and continued to fall for several years.
While the U.S. has plentiful access to rare earths in the ground, Zumwalt-Forbes told me, it has “a very underdeveloped ability to take that concentrate out away from mine sites and make useful materials out of them. What this deal does is it effectively bridges that gap.”
The combination of whipsawing prices and monopolistic Chinese capacity to process and refine rare earths makes the U.S.’s existing large rare earth reserves less commercially viable.
“In order to compete against that monopoly, the government needed to be fairly heavy handed in structuring a deal that would both get a magnet facility up and running and ensure that that magnet facility stays in operation and weathers the storm of Chinese price manipulation,” Zumwalt-Forbes said.
Beyond simply throwing money around, the federal government can also make long-term commitments that private companies and investors may not be willing or able to make.
“What this Department of Defense deal did is, yes, it provided much-needed cash. But it also gave them strategic certainty around getting that facility off the ground, which is almost more important,” Zumwalt-Forbes said.
“I think this won’t be the last creative critical mineral deal that we see coming out of the Department of Defense,” Zumwalt-Forbes added. They certainly are in pole position here, as opposed to the other agencies and prior administrations.”
On a new plan for an old site, tariffs on Canada, and the Grain Belt Express
Current conditions: Phoenix will “cool” to 108 degrees Fahrenheit today after hitting 118 degrees on Thursday, its hottest day of the year so far • An extreme wildfire warning is in place through the weekend in Scotland • University of Colorado forecasters decreased their outlook for the 2025 hurricane season to 16 named storms, eight hurricanes, and three major hurricanes after a quiet June and July.
President Trump threatened a 35% tariff on Canadian imports on Thursday, giving Prime Minister Mark Carney a deadline of August 1 before the levies would go into effect. The move follows months of on-again, off-again threats against Canada, with former Canadian Prime Minister Justin Trudeau having successfully staved off the tariffs during talks in February. Despite those earlier negotiations, Trump held firm on his 50% tariff on steel and aluminum, which will have significant implications for green manufacturing.
As my colleagues Matthew Zeitlin and Robinson Meyer have written, tariffs on Canadian imports will affect the flow of oil, minerals, and lumber, as well as possibly break automobile supply chains in the United States. It was unclear as of Thursday, however, whether Trump’s tariffs “would affect all Canadian goods, or if he would follow through,” The New York Times reports. The move follows Trump’s announcement this week of tariffs on several other significant trade partners like Japan and South Korea, as well as a 50% tariff on copper.
The long beleaguered Lava Ridge Wind Project, formally halted earlier this year by an executive order from President Trump, might have a second life as the site for small modular reactors, Idaho News 6 reports. Sawtooth Energy Development Corporation has proposed installing six small nuclear power generators on the former Lava Ridge grounds in Jerome County, Idaho, drawn to the site by the power transmission infrastructure that could connect the region to the Midpoint Substation and onto the rest of the Western U.S. The proposed SMR project would be significantly smaller in scale than Lava Ridge, which would have produced 1,000 megawatts of electricity on a 200,000-acre footprint, sitting instead on 40 acres and generating 462 megawatts, enough to power 400,000 homes.
Sawtooth Energy plans to hold four public meetings on the proposal beginning July 21. The Lava Ridge Wind Project had faced strong local opposition — we named it the No. 1 most at-risk project of the energy transition last fall — due in part to concerns about the visibility of the turbines from the Minidoka National Historic Site, the site of a Japanese internment camp.
Get Heatmap AM directly in your inbox every morning:
Republican Senator Josh Hawley of Missouri said on social media Thursday that Energy Secretary Chris Wright had assured him that he will be “putting a stop to the Grain Belt Express green scam.” The Grain Belt Express is an 804-mile-long, $11 billion planned transmission line that would connect wind farms in Kansas to energy consumers in Missouri, Illinois, and Indiana, which has been nearing construction after “more than a decade of delays,” The New York Times reports. But earlier this month, Missouri Attorney General Andrew Bailey, a Republican, put in a request for the local public service commission to reconsider its approval, claiming that the project had overstated the number of jobs it would create and the cost savings for customers. Hawley has also been a vocal critic of the project and had asked the Energy Department to cancel its conditional loan guarantee for the transmission project.
New electric vehicles sold in Europe are significantly more environmentally friendly than gas cars, even when battery production is taken into consideration, according to a new study by the International Council on Clean Transportation. Per the report, EVs produce 73% less life-cycle greenhouse gas emissions than combustion engine cars, even considering production — a 24% improvement over 2021 estimates. The gains are also owed to the large share of renewable energy sources in Europe, and factor in that “cars sold today typically remain on the road for about 20 years, [and] continued improvement of the electricity mix will only widen the climate benefits of battery electric cars.” The gains are exclusive to battery electric cars, however; “other powertrains, including hybrids and plug-in hybrids, show only marginal or no progress in reducing their climate impacts,” the report found.
Aryna Sabalenka attempts to cool down during her Ladies' Singles semi-final at Wimbledon on Thursday.Julian Finney/Getty Images
With the United Kingdom staring down its third heatwave in a month this week, a new study warns of dire consequences if homes and cities do not adapt to the new climate reality. According to researchers at the University College London and the London School of Hygiene and Tropical Medicine, heat-related deaths in England and Wales could rise 50-fold by the 2070s, jumping from a baseline of 634 deaths to 34,027 in a worst-case scenario of 4.3 degrees Celsius warming, a high-emissions pathway.
The report specifically cited the aging populations of England and Wales, as older people become more vulnerable to the impacts of extreme heat. Low adoption of air conditioning is also a factor: only 2% to 5% of English households use air conditioning, although that number may grow to 32% by 2050. “We can mitigate [the] severity” of the health impacts of heat “by reducing greenhouse gas emissions and with carefully planned adaptations, but we have to start now,” UCL researcher Clare Heaviside told Sky News.
This week, Centerville, Ohio, rolled out high-tech recycling trucks that will use AI to scan the contents of residents’ bins and flag when items have been improperly sorted. “Reducing contamination in our recycling system lowers processing costs and improves the overall efficiency of our collection,” City Manager Wayne Davis said in a statement about the AI pilot program, per the Dayton Daily News.
Or at least the team at Emerald AI is going to try.
Everyone’s worried about the ravenous energy needs of AI data centers, which the International Energy Agency projects will help catalyze nearly 4% growth in global electricity demand this year and next, hitting the U.S. power sector particularly hard. On Monday, the Department of Energy released a report adding fuel to that fire, warning that blackouts in the U.S. could become 100 times more common by 2030 in large part due to data centers for AI.
The report stirred controversy among clean energy advocates, who cast doubt on that topline number and thus the paper’s justification for a significant fossil fuel buildout. But no matter how the AI revolution is powered, there’s widespread agreement that it’s going to require major infrastructure development of some form or another.
Not so fast, says Emerald AI, which emerged from stealth last week with $24.5 million in seed funding led by Radical Ventures along with a slew of other big name backers, including Nvidia’s venture arm as well as former Secretary of State John Kerry, Google’s chief scientist Jeff Dean, and Kleiner Perkins chair John Doerr. The startup, founded and led by Orsted’s former chief strategy and innovation officer Varun Sivaram, was built to turn data centers from “grid liabilities into flexible assets” by slowing, pausing, or redirecting AI workloads during times of peak energy demand.
Research shows this type of data center load flexibility could unleash nearly 100 gigawatts of grid capacity — the equivalent of four or five Project Stargates and enough to power about 83 million U.S. homes for a year. Such adjustments, Sivaram told me, would be necessary for only about 0.5% of a data center’s total operating time, a fragment so tiny that he says it renders any resulting training or operating performance dips for AI models essentially negligible.
As impressive as that hypothetical potential is, whether a software product can actually reduce the pressures facing the grid is a high stakes question. The U.S. urgently needs enough energy to serve that data center growth, both to ensure its economic competitiveness and to keep electricity bills affordable for Americans. If an algorithm could help alleviate even some of the urgency of an unprecedented buildout of power plants and transmission infrastructure, well, that’d be a big deal.
While Emerald AI will by no means negate the need to expand and upgrade our energy system, Sivaram told me, the software alone “materially changes the build out needs to meet massive demand expansion,” he said. “It unleashes energy abundance using our existing system.”
Grand as that sounds, the fundamental idea is nothing new. It’s the same concept as a virtual power plant, which coordinates distributed energy resources such as rooftop solar panels, smart thermostats, and electric vehicles to ramp energy supply either up or down in accordance with the grid’s needs.
Adoption of VPPs has lagged far behind their technical potential, however. That’s due to a whole host of policy, regulatory, and market barriers such as a lack of state and utility-level rules around payment structures, insufficient participation incentives for customers and utilities, and limited access to wholesale electricity markets. These programs also depend on widespread customer opt-in to make a real impact on the grid.
“It’s really hard to aggregate enough Nest thermostats to make any kind of dent,”” Sivaram told me. Data centers are different, he said, simply because “they’re enormous, they’re a small city.” They’re also, by nature, virtually controllable and often already interconnected if they’re owned by the same company. Sivaram thinks the potential of flexible data center loads is so promising and the assets themselves so valuable that governments and utilities will opt to organize “bespoke arrangements for data centers to provide their services.”
Sivaram told me he’s also optimistic that utilities will offer data center operators with flexible loads the option to skip the ever-growing interconnection queue, helping hyperscalers get online and turn a profit more quickly.
The potential to jump the queue is not something that utilities have formally advertised as an option, however, although there appears to be growing interest in the idea. An incentive like this will be core to making Emerald AI’s business case work, transmission advocate and president of Grid Strategies Rob Gramlich told me.
Data center developers are spending billions every year on the semiconductor chips powering their AI models, so the typical demand response value proposition — earn a small sum by turning off appliances when the grid is strained — doesn’t apply here. “There’s just not anywhere near enough money in that for a hyperscaler to say, Oh yeah, I’m gonna not run my Nvidia chips for a while to make $200 a megawatt hour. That’s peanuts compared to the bazillions [they] just spent,” Gramlich explained.
For Emerald AI to make a real dent in energy supply and blunt the need for an immediate and enormous grid buildout, a significant number of data center operators will have to adopt the platform. That’s where the partnership with Nvidia comes in handy, Sivaram told me, as the startup is “working with them on the reference architecture” for future AI data centers. “The goal is for all [data centers] to be potentially flexible in the future because there will be a standard reference design,” Sivaram said.
Whether or not data centers will go all in on Nvidia’s design remains to be seen, of course. Hyperscalers have not typically thought of data centers as a flexible asset. Right now, Gramlich said, most are still in the mindset that they need to be operating all 8,760 hours of the year to reach their performance targets.
“Two or three years ago, when we first noticed the surge in AI-driven demand, I talked to every hyperscaler about how flexible they thought they could be, because it seemed intuitive that machine learning might be more flexible than search and streaming,” Gramlich told me. By and large, the response was that while these companies might be interested in exploring flexibility “potentially, maybe, someday,” they were mostly focused on their mandate to get huge amounts of gigawatts online, with little time to explore new data center models.
“Even the ones that are talking about flexibility now, in terms of what they’re actually doing in the market today, they all are demanding 8,760 [hours of operation per year],” Gramlich told me.
Emerald AI is well aware that its business depends on proving to hyperscalers that a degree of flexibility won’t materially impact their operations. Last week, the startup released the results of a pilot demonstration that it ran at an Oracle data center in Phoenix, which proved it was able to reduce power consumption by 25% for three hours during a period of grid stress while still “assuring acceptable customer performance for AI workloads.”
It achieved this by categorizing specific AI tasks — think everything from model training and fine tuning to conversations with chatbots — from high to low priority, indicating the degree to which operations could be slowed while still meeting Oracle’s performance targets. Now, Emerald AI is planning additional, larger-scale demonstrations to showcase its capacity to handle more complex scenarios, such as responding to unexpected grid emergencies.
As transmission planners and hyperscalers alike wait to see more proof validating Emerald AI’s vision of the future, Sivaram is careful to note that his company is not advocating for a halt to energy system expansion. In an increasingly electrified economy, expanding and upgrading the grid will be essential — even if every data center in the world has a flexible load profile.
’We should be building a nationwide transmission system. We should be building out generation. We should be doing grid modernization with grid enhancing technologies,” Sivaram told me. “We just don’t need to overdo it. We don’t need the particularly massive projections that you’re seeing that are going to cause your grandmother’s electricity rates to spike. We can avoid that.”