Arushi Sharma Frank thinks the best way to restore consumer trust in the power sector is to build a better grid.
Arushi Sharma Frank started working in the energy industry in the early 2010s and quickly learned how vital state and federal regulatory bodies are to the energy transition. Years later, as part of Tesla’s U.S. energy policy team, she convinced Texas regulators to allow Tesla to connect a 100 megawatt battery to the grid. More recently, she developed new ERCOT rules that allow data centers to connect to the grid much faster.
Arushi invests in and supports clean energy startups as the founder and principal of Luminary Strategies, a venture studio she launched in 2024. But she also spends a lot of time contributing to efforts to improve grid infrastructure. Without the technology startups she supports, consumers' trust in the grid can’t be restored, she says.
This week on With Great Power, Arushi Sharma Frank describes her vision for “prosperous infrastructure” and her recent work with Texas regulators. She also talks about some of the startups she’s working with, as well as how she sees AI impacting the power sector.
Credits: Hosted by Brad Langley. Produced by Mary Catherine O'Connor. Edited by Anne Bailey. Original music and engineering by Sean Marquand. Stephen Lacey is executive editor. The GridX production team includes Jenni Barber, Samantha McCabe, and Brad Langley.
Brad Langley: Last fall, Texas Governor Greg Abbott was all in on data centers.
Governor Greg Abbott: Texas will be the centerpiece for AI data centers for Google in the entire world.
Brad Langley: That clip is from a November press conference where Google CEO Sundar Pachai said the company was investing $40 billion to build three AI data centers in the state. It's part of a growing nationwide movement.
News clip: Meta, Amazon, Google, Microsoft expected to spend 1.5 trillion building data centers through next year.
Brad Langley: But public backlash over rise in electric bills and concerns about water use and noise pollution have been growing in lockstep with tech companies' ambitions. And as that pushback intensifies, Arushi Sharma Frank has been thinking, "I've seen this movie before." That's because early in her career, she was a lawyer for a natural gas industry group.
Arushi Sharma Frank: The word data center has the energetic negativity behind it that the word fracking did back in 2011.
Brad Langley: In 2015, Arushi joined the power sector, working in regulatory roles for electric power generators and utilities. Ever since, she's been obsessed with how the power grid works and how it could work even better. And while the data center backlash is tied to structural issues on the grid, she says public opposition to data centers and to a large extent AI in general is also rooted in distrust.
Arushi Sharma Frank: It's simple, deep mistrust of technology and a complete disconnect with how we make technology trust happen.
Brad Langley: Data centers are now a big campaign issue in many midterm elections, including Governor Abbott's. In the past year, he's gone from embracing hyperscalers to banning new data center permits while Texas regulators audit the grid impacts and proposals already in the queue. The audit is a step in the right direction, but Arushi says building technology trust is a much bigger job.
Arushi Sharma Frank: With tech energy people putting all of that together, there's translation layers you need. We need that at scale. We need it urgently.
Brad Langley: That's why she's trying to reframe the way that stakeholders both think about and talk about infrastructure.
Arushi Sharma Frank: I'm here to build infrastructure in a way that ensures social prosperity.
Brad Langley: And to make that happen, she's taking on the role of translator between tech and energy sectors and connecting it all to people outside those spheres. Her vehicle for that work is Luminary Strategies, a venture studio she launched in 2024.
Arushi Sharma Frank: My goal is to create full tilt, full market infrastructure solutions and support young companies that are building into the space. I see it as my job to catch the moment and restore the power of language and connect it to actual outcomes and all of my work over the past two years, bringing companies into my platform, working on changing big rules for the big grid. All of it comes down to putting all those things back together in the same box so people see the full picture and they understand that the box means more things for people.
Brad Langley: This is With Great Power, a show about the people building the future grid today. I'm Brad Langley.
Some people say utilities are slow to change, that they don't innovate fast enough. And while it might not always seem like the most cutting edge industry, there are lots of really smart people working really hard to make the grid cleaner, more reliable, and customer-centric.
Today, my guest is Arushi Sharma Frank, founder of Luminary Strategies. This is actually the second time Arushi has been on the show. In episode four of season four, we talked about her time working on Tesla's energy policy team among many other things. We'll link to that episode in the show notes. Today we're talking about her recent work with Texas regulators and some of the startups she supports. We also look at how AI is impacting the power sector, but first I asked Arushi to describe what's behind the rising costs that are feeding the data center backlash.
Arushi Sharma Frank: Retail, meaning the costs you and I pay in our homes for power, those retail commodity costs are not necessarily rising, but the cost of the infrastructure that's delivering all of that and the way that we invest in keeping those things bigger and growing, that's going up. And then the contributors to the system, meaning the number of users of electricity, those haven't really gone higher. So the answer there is always going to be simple math. It's rates rise for pro rata per user of the grid if there are less users and the cost of the whole grid are going up.
That's what I'm seeing historically is getting lost in the current narrative on affordability. And then Brad, you can layer on top of that the forward cost of electricity in our regional wholesale markets that is absolutely dependent on the issues of capacity auctions and how the data center load growth scenario, when you put that into the capacity auction and you decide that you're going to procure forward power or forward capacity for all of this new load growth, well gosh, of course the forward cost of power is going to go up and that trickles down into people's retail bills on the commodity side in any way we haven't seen historically. So those are the two sides of the problem.
I actually helped found a coalition earlier this year. It's a independent trade association called Utilize. Its founding members include Tesla, Google, Spark Fund, Bring You Home and more. And the purpose of a coalition is to move us into an era where we can arrest the affordability story by moving towards a common standard across each state where we can determine through simple transparency metrics like which parts of the system are underutilized relative to how much capital is available to deploy across it to fix and improve the grid, and also how many users are actually actively using enough of the grid so we can fix this numerator denominator cost problem. It's a super interesting exercise, but if you think about why this coalition exists in 2026, it's because we've not done a great job in the United States on figuring out how to cost effectively reward efficient use of the distribution and the transmission system. If you don't do that, you just have lower utilization and you have rising costs of the whole infrastructure.
Brad Langley: So late last year you filed a proposal with the Electric Reliability Council of Texas, also known as ERCOT, for new rules that would allow data centers to connect to the grid much faster. It's a framework called provisional controllable load resources, also known as PCLR. First, why did you do this and what was the need you were trying to address?
Arushi Sharma Frank: We were looking at a world where there weren't any great solutions for any kind of large load that's willing to accept the risk of not operating at full power during peak. There was no way for that data center or any other load to accept that risk in exchange for getting some firm power on a timeline of say three to five years. So last year we were really stuck in Texas. Other stuff that was happening simultaneously was the governor of Texas was telling ERCOT, the grid operator, that we need to completely rewrite all of the large load connection rules, which then happened. And meanwhile, I was sitting there at my desk with my advised ventures and with ERCOT staff and thinking, okay, this is the one place where I know we can have a firm plus flex framework. We have to have it. And that's how I started working on it last year.
Brad Langley: So what's in the proposal and how does it relate, if at all, to the PUC of Texas approving ERCOT's batch zero process for connecting large electricity users?
Arushi Sharma Frank: I wrote planning guide revision request 134 and filed it as a draft in September of last year before the batch zero framework existed before it was contemplated. So I wrote this planning guide revision request. We started presenting it in a stakeholder process and it said that if you agree to be operationally flexible, i.e. You accept the same grid ramp and dispatch instructions that all batteries and generators accept in Texas to connect to the grid faster, then you can be in the line for faster power, but you also have to pay for the full cost of your firm grid upgrades. So you're a firm load, you're not different than any other load, you're still in the queue, but once your utility signs off on your interconnection processing, the grid operator can start giving you power sooner if you're willing to be operationally dispatchable. So that's what is called batch zero.
And that process kicked off in December and by the time the survey results from the market showed up asking the stakeholders of ERCOT and all data center developers, "Hey, what do you guys want to see out of our redesign of the entire large load interconnection?" Guess what shows up in the survey responses? The flex to firm proposal I filed, a co-location proposal that simply makes grid loads net neutral to the grid called a private use network. It's actually withdrawal limited private use network. Lots of complicated words here, but basically these two concepts came up as like, oh, we need these. We need dispatchability. We need net neutral for the grid. If we don't have that, all of the new design process you're doing for batch zero is going to be a total washout because no load will be able to connect anything and get allocated any firm power over some five or six year period if we don't have some resources on the system that can connect in advance of this firm upgrades and we can't utilize that system at a higher level.
So with that survey result and the obvious momentum, ERCOT took the 134 proposal and refined it into what they believe would be the most acceptable minimum viable product for firm plus flex speed to power. And they put it inside the final batch zero rules that were just approved by the PUC in June. The speed at which we put a speed to power guideline and a process and then passed it as rules of the grid is just unmatched in the country and in the world. I don't think any grid has finalized rules within 10 months, but we did it in Texas.
Brad Langley: So then looking at the technology response to data center demand, there's a lot of interest in building data centers that are flexible power users. We had Varun Sivaram on the podcast as a company that you advise and invest in his company Emerald AI, which uses AI tools to build more responsive and grid-friendly data centers. What does it actually take to get AI infrastructure onto the grid and allow it to scale? And maybe as a follow-up to that, just based on what you were talking about, has AI become almost like a dirty word alongside data centers and how do we maybe shift that perception as well?
Arushi Sharma Frank: Oh yeah. So yes, absolutely. The word artificial intelligence, it's fascinating. I actually work with, in Washington, since you mentioned the midterms, I think I might be one of the only people in DC who has a background working at a company like Tesla and enabling gigawatts of clean power solutions along with anything else that the market can create with those solutions, who speaks to conservative America from a voice of free markets. That's who I am. I'm a free markets champion. I'm a member on the C3 Solutions Advisory Board in addition to sitting at a nonpartisan think tank. I do events with organizations like the Daily Caller News Foundation, and I speak to MAGA America as well in that platform. In fact, I write for them as well because I think the obligation at this point, the fix is not a fix. I don't like the word fix anymore because it puts a Band-Aid on something that needs to be healed from the inside and you can't heal from the inside unless you are putting capital, private capital in the space of being actively, vociferously, unapologetically committed to social prosperity.
Emerald AI is in my prosperous infrastructure platform. I've been advising that company when it just had a few employees and ever since. They are there because Varun's commitment to show the world that real capital will back grid asset data centers, it will back a better prosperous outcome for society is absolutely essential against the tidal wave of billions of dollars going into projects that do not have any ascertainable, discernible social prosperity outcome. Not one that people can see because the people in charge of those investments are not thinking about the social layer that they must bring along and galvanize to build their company and win those private returns. For the utility managers who are listening to this podcast, another person who strongly believes in this is Patti Poppe. I was just on a podcast hosted by one of her former employees, Matt Smith, who mentioned that he learned from Patti this concept of people, planet and profit.
And there's a very intentional piece to that order where if you want to move off of data centers and AI being dirty words to Americans, whether it is MAGA America, Democrat America, I-don't-care America, it doesn't matter. If we need an agnostic platform for this, we need people like Varun, we need people like Zach Dell—Base Power also part of my platform. We need these founders who come with capital and reputation to be new spokespeople for the private sector and also be inside communities and be inside layers of narrative creation that show actuation of community change and community prosperity and respect. And I think respect is also one of the things that you can't stick a Band-Aid on. If you don't respect the amount of power people have over their land, their autonomy, their prosperity in this country, you just can't win. And I see a lot of that.
Almost every day there's some announcement or some article written by, I don't know, a current Texas railroad commissioner or a current community champion in Colorado or in Utah, and it doesn't really matter where it's coming from. It could be red states and blue states, could be an affordability narrative, it could be a resources narrative, but those narratives exist because like I said earlier, we did not do enough as a capital industry to put the right word in front of the word infrastructure when we said we're going to come build this thing in your neighborhood. That to me is deeply wrong and it takes many, many, many, many, many micro rights to slowly work towards correcting such a big wrong.
Brad Langley: So as we're seeing with data centers as a growing backlash against AI, how is the backlash impacting the use of AI in the power sector more generally?
Arushi Sharma Frank: I don't know if there is any impact. I say that because the timeline, think about the power sector's evolution to adopt AI and to build for AI as one wheel moving on a track autonomously, and the other track has the data center tech industry moving on it. The power wheel moves from point A to point B to point C very slowly. And so a lot of the tech sector is shock that you see in capital markets and in deal decks and in investor proposals and shareholder calls. Frankly, it's happening on its own track that has a lot more volatility and a lot more of the stop and starts. On the utility side, it's slower track. With that being said, the one area where I think there is going to be immediate impact with things like the moratorium in Texas is that it's going to be harder for utilities, credit worthy utilities, to underwrite their own forecasts of what they believe they need to raise cheaper capital for to build for these loads.
So even if you have excellent large load tariffs in a specific state that ensure the general public against the potential rising costs of data center connections, that's great, but that also means the utility, which also is funded by investors, institutional investors, and then some retail investors. Those entities need to go into equity and debt markets to raise for the thesis they have built today for how much data center load is coming to their system. And that thesis is risk averse. It is designed to keep headroom in the grid. It is designed to protect against reliability crises. So that forecast is tied to their financial wellbeing and how they represent it to investors and how credit agencies look at that forecast relative to what they believe is a viable, durable investible strategy in that utility stock. So if there's going to be an immediate sort of slowing or wobbling of the wheel that the utility systems are on to try and accommodate AI, that's the biggest challenge.
I think the rest of it is just noise because frankly, utilities will build the firm power by a date certain, period. And if you paid for it, they'll build it. Really the backlash comes in figuring out whether or not it was a financially smart decision for the utility system itself, because even if loads pay 6X for a load that they never actually bring to the grid, you've still committed a whole bunch of utility infrastructure, labor, capital, OPEX to a whole system that will be fundamentally underutilized if that load doesn't come.
Brad Langley: We talked earlier about the idea of prosperous infrastructure, a concept I love the naming of. A couple of follow-up questions on that. What do investors and technology companies understand about infrastructure today that they didn't understand five years ago and what are they still missing?
Arushi Sharma Frank: Absolutely. Infrastructure technology for investors, I think the emerging issue is that investors are more cognizant that they have been repeatedly encountering the same missing middle surprise, which is the technology is great, it exists, it's flourishing, but the market architectures to absorb that and create a real opportunity, it requires all of these things that investors traditionally did not look at, which is commercial rules, regulatory reform, operating architecture that fits the basic physics of the grid and acceptance of all of that by community members and what it means for what it does to advance the socioeconomic goals of a specific region. And now that missing middle problem is so big and it's so huge for the reasons we discussed that they have to move into a space where all of those things become a part of the founder thesis, they become a part of the pre-seed thesis.
And frankly, I mean, I don't know what's a bigger signal of this than my existence. I came from a world where I already knew that when you invest in the grid itself and in the people and in the rules of the grid, it becomes compounding infrastructure, right? Every time you build a better market, you get capital. The capital brings the technologies. The technologies work because the operators understand them and trust them. The whole thing makes the whole of society more resilient. I started advising Base Power within two weeks of leaving Tesla. Zach Dell and Justin Lopas, amongst the many things they have to their credit as founders, as co-founders, already understood that they need people who understand that the prosperous infrastructure thesis of building for all of these moats simultaneously and building into a socially prosperous system is the only way to do private returns well in a hard space like batteries and grids.
Before, investors didn't care if these companies had regulatory or market design experts. They didn't care about it. In fact, policy, regulatory, market architecture, whole of systems as skill sets were things that companies put into their plan post-raise, post-seed, post-series A. And now it's like startups, if you're in this space to make a difference, you better come in with a strong leader in regulatory policy and even better if that person has credibility as an investor or a board advisor because all of that is needed and now investors are really starting to wake up to it.
Brad Langley: So you've been on the show before, therefore you know this question is coming and I'm interested to see if your answer has changed or if you've acquired any new ones in the past two years, but we call the show With Great Power, which is a nod to the energy industry. It's also a Spider-Man quote, "With great power comes great responsibility." So Arushi, what new or evolved superpower do you have in your arsenal these days?
Arushi Sharma Frank: What's my evolved—I forgot that you asked me this question the first time. My new power since being on your show last is that when I put language back together, entire ecosystems follow the reparation I have made and they build new things on top of it. Every tech company needs anchors who have the ability to remold and reshape and heal the language that is broken in our systems so that infrastructure can be built on top of it. And frankly, I'm really excited and it also gives me a sense of how much responsibility comes with that, that okay, now I've done step A, which is make PCLR real. Now I need to follow through with step B because now gigawatts of load might come to Texas or not come to Texas because of my work.
Brad Langley: Well, Rushi, you are clearly a market mover and a very busy person, so I thank you for not coming on the show just once but twice and we'll hopefully make it three times the charm here in next couple of years. Thank you so much for your time.
Arushi Sharma Frank: I'm delighted to be here and excited for your listeners to get this one.
Brad Langley: Arushi Sharma Frank is the founder and principal of Luminary Strategies. With Great Power is produced by GridX in partnership with Latitude Studios. Delivering on our clean energy future is complex. GridX exists to simplify the journey. GridX is the enterprise rate platform that modern utilities rely on to usher in our clean energy future. We design and implement emerging rate structures and we increase consumer investment in clean energy, all while managing the complex billing needs of a distributed grid.
Mary Catherine O'Connor produced a show. Anne Bailey is our senior editor. Stephen Lacey is our executive editor. Sean Marquand composed the original theme song and mixed the show. The GridX production team includes Jenni Barber, Samantha McCabe, and me, Brad Langley. If this show is providing value for you and we really hope it is, we'd love it if you could help us spread the word. You can rate or review us at Apple and Spotify, or you can share a link with a friend, colleague, or the energy nerd in your life.
As always, thanks so much for listening. I'm Brad Langley.