Subscribe
Share
Share
Embed
Stripe cofounder John Collison interviews founders, builders, and leaders over a pint.
John (00:00):
The US did have this long running anti-nuclear bent.
Julia (00:04):
I can't tell you how much has changed in the last few years. They changed the mandate of the NRC and they said, “You are no longer focused just on safety.”
Department of Energy pathways have also been unopened and unblocked for test reactors.
The last 10 years or so, it's been like 1% of nuclear fuel used in the US was actually domestically mined. We're still buying almost 50% of our uranium from Russia and kind of ex-Soviet states.
John (00:28):
How capital-intensive is it going to be to get your product to market?
Julia (00:32):
In the hundreds of millions.
The biggest organization that showed up against Diablo Canyon was called Mothers for Peace. This organization was that old that it was related to the nuclear Cold War, right? I’m part of this Mothers for Nuclear organization—
John (00:47):
Mothers against their mothers.
Julia (00:49):
And then I brought some graphite.
John (00:50):
Okay.
Julia (00:51):
Do you want to take that out? I'll show it to you.
John (00:52):
Sure. Yeah.
Julia (00:53):
This is what we are using for our moderator in our reactor. It has been designated a critical mineral. It's an important one—
John (01:00):
But this is just graphite. This is not Uranium 235 or something—
Julia (01:05):
I wouldn't have brought you that, would've been a little bit dangerous to be holding.
John (01:10):
Julia DeWahl is the co-founder of Antares, designing nuclear micro reactors for the military and other off-grid use cases.
Cheers.
Julia (01:23):
Cheers. It's great to be here.
John (01:27):
So you started out at first Opendoor and then SpaceX. What did you learn at each one of those that you took to Antares?
Julia (01:36):
Yeah, sure. I think one of the big things is always about customers. I mean, at the end of the day, this is why you build a business. At Opendoor, a lot of us were pretty young starting the business. I was one of the early employees there and none of us had ever bought, let alone sold a house. So a lot of it was learning what that was all about and I remember in the early days we were working on writing the copy for the website and we're like—
John (01:58):
You have no idea what a homeowner—
Julia (01:59):
We don’t really know much about this whole thing, so what can we do about that? We just went and sat outside a bagel shop for a Saturday and a Sunday morning and just bought people bagels and asked them about their processes, buying and selling a home, and learned a ton. And really thought about what language they were using, what were the pain points, and that was really important to just really get to know what we were trying to deliver on as a business. And then fast forward to joining SpaceX after Opendoor, I joined as Starlink was coming out of R&D and into go-to-market. That's why I was joining, to lead that team and the business operations there. And we launched and we got our first customers and so I invited our first customers to come join us via Zoom for an all-hands. And it was honestly the first time that many of the engineers had actually heard from the words of a customer themselves about the experience, and of course everyone loved the internet. The fast internet was amazing. Rural Idaho, this is a dream. But when we got to talking about the setup process, everyone kind of was like, “Oh, that sounds rough.” They're talking about multiple Home Depot trips. They didn't know if they had to cut down trees on their property, they couldn't get a pole long enough. And so—
John (03:03):
A lot of it was simplifying people's understanding of what it would mean.
Julia (03:06):
People kind of didn't realize how the setup was so painful and it led to basically this cross-functional SWAT team that was dedicated to mounting options to be sold alongside the dish. And if you hadn't—
John (03:20):
First-party or third-party?
Julia (03:21):
We did it ourselves actually. We said it's important that we have a good experience for that because actually it is a big part of the experience and so we spun up a whole team. A lot of attention was given to mounting, which is like who ever would've thought? And I think, customers, again, they are the reason you start a business. Just bringing this back to Antares, I mean the reason we started Antares was to say the government, particularly the military, needs more energy resilience, especially for critical infrastructure. We just do so much more now in the cloud with satellites. Our missile defense sites are sitting there being powered by diesel generators still. Shouldn't we upgrade that?
John (03:58):
Okay, so I'll bite. What is Antares?
Julia (04:01):
Sure, sure. We're building micro-reactors, so small nuclear reactors and we are building in particular for critical infrastructure, largely for the military. Although we do think that as costs come down, there will be an interesting commercial market too. But basically anywhere where you're off grid for sure, or you just want resilience behind the grid as it exists today. Cyber attacks or whatever the grid goes down, you want backup power. It's a really great option.
John (04:27):
And if one megawatt is a small neighborhood, if one gigawatt is a big nuclear power plant like in The Simpsons, what are your SMRs?
Julia (04:38):
Sure, yeah. There's basically three classes size-wise. You have the really small ones, which are the micro-reactors, and we're actually one of the smallest of those. So we're even sub one megawatt. We're kind of in the hundreds of kilowatt scale. You could think of that as hundreds of houses worth of electricity. Then you have the SMRs, which is kind of this middle ground bucket, which is an interesting kind of neither here nor there, but potentially interesting. I'm sure we'll get into it. Kind of in the hundreds of megawatts or even 50 megawatts. And then you have the gigawatts, a thousand megawatts and that's what the current power plants are today.
John (05:07):
So you're a sub a megawatt.
Julia (05:08):
So we're super small.
John (05:09):
I see. Okay. So you're kind of similar nuclear reactor size to what would be on a submarine or something like that.
Julia (05:13):
Smaller than that even. Those are megawatt scales. So we're like, think about we're the size of a car reactor. Think about replacing a diesel generator that's used to power something that would be on the order of 200 to 300 homes of electricity. So it's quite small.
John (05:26):
And so a big part of your focus is providing power in places where you really need that reliability. You're talking about replacing diesel generators, is that kind of the design center for you?
Julia (05:38):
Yeah. And anywhere where you're going to enable a capability, a mission capability, that wouldn't really be possible otherwise. So there's actually some really interesting space applications, potentially underwater applications as you move to… So you move away from what they call exquisite systems to attributable systems. This is where a really small reactor can come in very handy for the submarines are obviously multi-megawatt-powered systems, but these could be kilowatt-powered, smaller systems underwater. But yes, anywhere where you're off grid where today you're relying on what they call the fuel supply chain of bringing in fuel via helicopter or barge repeatedly and that being a vulnerability, this is a great option.
John (06:21):
So you're the Starlink of electricity, bringing it to places where—
Julia (06:23):
We're actually perfectly paired. And I wish we could offer a consumer product, that's not on our roadmap right now, but maybe one day, like backpack-sized nuclear reactor.
John (06:34):
I presume it kind of can't be on your roadmap, but why not? Is selling to the government is much easier than selling to something—
Julia (06:40):
So many reasons. I think that you always want to try to get your beachhead where it's the easiest to achieve. And so we believe, I believe, that premium power is where you need to go with nuclear. And there there's two places for that. One is what we're doing with Antares, which is to go to the military, help them with mission-critical needs where they're not cost-sensitive, where they already pay more for power, and go there and support them with this next gen better version of resilient power. Or you go and you sell, what we're seeing now I think, which is really interesting with the hyperscalers is they have carbon commitments. They want clean firm, they want 24/7 power. They're actually willing to pay above market rates for that and they will pay above market rates. So this premium power thing is a really interesting dynamic of where nuclear is kind of well positioned.
John (07:27):
What are the underwater applications?
Julia (07:29):
Well, it's really just, think mini-submarines.
John (07:31):
Oh, okay.
Julia (07:32):
Just really small systems. What drones are to maybe larger missiles. That's kind of the equivalent. It's not super far along just yet.
John (07:41):
Got it. Okay. What does the path look like between now and producing power for the military? I guess, first in a pilot environment and then kind of at scale outside of a pilot environment.
Julia (07:54):
Yeah. Well, it's always what are the next milestones that you can do to prove your technical ability and raise more capital to keep going with your company. I think the biggest risk often to hardware companies is you don't get far enough along technically to merit additional investment. So we're thinking a lot about our milestones for reactor development. And so, next year a criticality test is kind of the first milestone. By the end of 2027, we want to have our first demonstration unit demonstrated. And what's really exciting is there's just so much good policy happening right now from the DOE and elsewhere that is supporting reactor tests and pilot programs. And so there's just so much happening where they're opening up federal land, they're removing… Starting to remove some of those NEPA type regulations that really just hold everything up. So there is a clearer path now than there really ever has been for reactor builds.
John (08:52):
Oh, we'll get to the regulatory side. I have a lot of questions, but what is your criticality test in 2026? What does that mean?
Julia (08:58):
Yeah, it means that you're testing that you can actually start your reaction. So you've figured out the right amount of moderator to fuel and you start it going critical. You don't have it continue on for a while. You don't typically have a power convergence system that's actually taking the raw energy and turning it into electricity. So it's just one of those initial tests. Can you start to put all your pieces together?
John (09:23):
And fusion guys are all focused on this, producing more power than you put in. I know that's different. That's a fusion thing, but yeah, when do you actually produce power?
Julia (09:37):
Yeah, producing power will be your first reactor test. It's like can you do the whole shebang? You can get this thing up and running. It's producing power, and that's what we're aiming to do by the end of 2027.
John (09:47):
Wow, that's pretty soon.
Julia (09:48):
Yeah.
John (09:49):
What does the market look like for these small reactors? Who are you going to compete with?
Julia (09:54):
Well, the market for micro-reactors is one of these ones where it's like, well, there are no micro-reactors today on the market to just buy and use, so what do you think they're going to be, right? And so I think, first and foremost, it's going to be the most critical assets where you just cannot have them go down, which is why defense is the perfect first market for it. And so it's going to be all of your critical infrastructure on every single base that there is. Obviously you'll start with the most critical of critical missile defense sites and off-grid locations, things like that. And that's kind of defense. Then you think about commercial, basically you have to see where the cost curve gets you to “How cheap can it get?” This is just like space. Those first expensive launches into space that better have been a really important thing you were putting up there. But now that launch cost has come down, you can start to be like, “Oh, what else could we do? We can do some more stuff that's maybe not so mission critical.” I think we'll see the same thing in the commercial industries. Oil and gas has a lot of areas where they use diesel generators to do things like compressed natural gas along their pipelines and they actually have carbon commitments, which are really interesting that they're sort of thinking about. Ironically, can they move towards some more renewable or nuclear type resources? So I think there's going to be some commercial applications there too. Mining, for example, too. Anything that's just off grid, hard to power today, really hard to get fuel into. It's the perfect market.
John (11:17):
Yes, you're probably following the US military, there's all these controversies right now with people talking about fuel quality and fuel contamination issues in Asia, I think it is. But yeah, presumably places where you're worried about fuel availability at all times and ensuring that you constantly have access to fuel, that is the place where there'll be the most pull for these solutions. That's where generators work worst, versus—
Julia (11:39):
And there's a famous defense quote, I think it's “Unleash us from the tether of fuel,” which means just like, gosh, the logistics supply chain has been such a challenge. I mean it was 50% of, I think, the Afghanistan casualties were from just logistics, and so having situations where you can reduce that risk helps and it's a great kind of market fit.
John (12:04):
Yes. You were saying this new market for these small reactors today. Why not? We've had nuclear efficient technology for a long time.
Julia (12:14):
Yeah, why hasn't it been done? Gosh, there's so many reasons. We ended up… I think there was just a macro really negative climate around nuclear in general, so it impacted anything small as well, coming out of the seventies. So you had macro forces related to inflation, cost of capital. People stopped building, which coincided with the environmental movement, which was extremely anti-nuclear. They actually sort of conflated nuclear weapons testing even with, it was just like “no nukes” with energy. It's like, oh, that's actually pretty different. So a lot of things converged in the seventies and then by the eighties we stopped building in nuclear.
John (12:50):
Okay, so you just view it as a part of the general anti-nuclear swing and that also led to the cessation of construction of new nuclear plants in the US and everything like this. Just your corner of the market also got caught up in that and that's why we haven't ended up with them. Were there ever products available at this size range?
Julia (13:11):
There were, but they were pretty few and far between. Mostly test reactors, small things built. We actually put one reactor into space in the 1960s to test that out.
John (13:22):
Did it work?
Julia (13:22):
Yeah, for a bit. I mean a few months and was like, “Let's try to do some systems in space with power.” It's important to have power in space and I think that's a whole other frontier.
John (13:32):
Obviously the US did have this long running anti-nuclear bent. Where I grew up in Ireland similarly, there was the exact same thing and Ireland was very paranoid about a particular British nuclear power plant, Sellafield, that was just across the Irish Sea.
Julia (13:48):
Actually, I don’t know. Does Ireland have nuclear?
John (13:51):
Ireland has no nuclear power. The UK does. France does. Ireland does not, and I believe it's banned and so—
Julia (13:57):
Interesting. There's actually many states in the US that have bans as well, but they've just been falling left and right. It's so cool to see.
John (14:04):
Well, I was going to ask this. The vibe has really shifted over the past 10 to 15 years. Why?
Julia (14:13):
Honestly, it reminds me of gay marriage. And if you look at the polling—
John (14:18):
It came from the youth.
Julia (14:19):
It definitely comes from youth for sure. But it just kind of… Enough things add up and it just swings and no one ever looks back, type of thing. That's how it feels. The percentage point change we've seen in the last five to 10 years is almost 20 points. I mean, it's like 17 points or something. So it's 60 plus percent of people support nuclear. It's actually almost surprising it's that low in some ways. But yeah, I mean it's just huge public sentiment shift. And I think the old hippie nuclear weapons testing folks are kind of dying out. What is wild though is I did little activism related to Diablo Canyon Power Plant here in California, which was slated to be shut down this year.
John (14:57):
Sure.
Julia (14:58):
And just a few years ago, Newsom had to kind of begrudgingly say like, “Oh, you know what? We probably should keep this thing on. We can't afford to have the power going out.” But I listened in on some hearings. I actually gave some testimony there, and it was so funny to me. The biggest organization that showed up against Diablo Canyon was called Mothers for Peace, and it's like, do you understand the irony of your name? This organization was that old that it was related to the nuclear Cold War, right?
John (15:23):
Totally.
Julia (15:24):
And it's like, gosh, we've moved on. We were joking. I sort of am part of this Mothers for Nuclear organization that's just like tongue and cheek—
John (15:32):
Our mother against their mothers.
Julia (15:33):
Our mothers against their mothers. And it's just funny how old school a lot of that opposition is.
John (15:39):
Yeah. So I guess there’s a “science advances one death at a time” dynamic going on here, where the memory of the 1960s and peak nuclear confrontation and Three Mile Island and all these things. As all that fades, maybe it creates an environment where people can be more supportive of nuclear.
Julia (15:59):
Yeah, absolutely.
John (16:00):
Yeah. What was your activism around Diablo Canyon?
Julia (16:02):
It was just exactly that. It was like, let's make sure that this actually passes and we don't shut it down. There was a lot of people just calling and saying, “Hey, listen, nuclear is super safe. It's carbon free. Sensibly in California, we care about these things. Like, let's support this. Let's overturn whatever legislation was done 10 years ago that said we should shut this thing down.” And it did pass. So now we do still have it open.
John (16:23):
Everyone on Progress Twitter likes to complain about the NRC. But for the NRC, we would be living in that Jetsons vision of the future and everyone would have their nuclear-powered car that they used to drive to their nuclear-powered office and everything. Is that true? How good or bad is the current regulatory environment? How much is that changing?
Julia (16:44):
Yeah, yeah. I think it's an overestimated headwind, certainly today. I think if you go back five years ago, 10 years ago for sure, I mean they hadn't licensed anything in decades. There had been nothing new, but I can't even tell you how much has changed in the last few years. It's kind of remarkable. And it's been across both the Biden and the Trump administration, which is amazing to see. I mean the legislation around nuclear passes 95 to 5 type of thing. It's really, really strong bipartisan. What was interesting to see is it started with Biden, a bunch of legislation related to modernizing and reforming the NRC. The biggest thing they did was they changed the mandate of the NRC and they said, “You are no longer focused just on safety.” Because what's the safest number of nuclear reactors to license? Zero. So they expanded that, “You must consider the full environment, the full civilization…” that's great. And then just change things around like the advanced reactor pathway, removing some of the barriers. And then it's continued on with Trump and in fact, he's sort of added even some more teeth and specificity. So right now they've changed it to, you have 18 months to get your answer from the NRC. License or no license. Do you pass or not?
John (17:59):
And what happens if they exceed the 18-month window?
Julia (18:02):
Probably it's in the details somewhere. But the point is that these have been declared that timeliness is important. There's also been some leadership change, which is just more reform-oriented, which is great, and there's just a huge microscope on it. Everyone's watching. Everyone knows now who's submitting their paperwork and what's going to come out the other side. Another thing that happened with some of these Trump executive orders is the Department of Energy pathways have been also unopened and unblocked for test reactors. And so actually Antares is part of this new test reactor program, which allows the DOE, alongside usually Idaho National Lab, to be sort of your regulator as you’re testing your first reactors so that you don't have to spend all this time upfront with the NRC for something that you're just still innovating on and iterating on. That's a really nice, I think, first step before you go through the NRC.
John (18:55):
Wait, so let me make sure I'm getting this right. There is now a Trump-era change where we essentially, for the first time ever, have a second nuclear regulator where you can go the NRC path, but also there's a regulatory sandbox where you can start by working with the DOE and get up and running.
Julia (19:12):
Yeah, that's essentially right.
John (19:13):
How big is the sandbox?
Julia (19:15):
I mean, the DOE has always been a path. There actually technically has always been a path within the military. They license their own reactors with submarines, so it's always kind of been there, but hard to navigate, not well staffed, not well known about. And now it's kind of like everyone's watching, let's go DOE, let's go INL, Idaho National Lab, and let's make these things happen. So there's just been a lot of stated like, “You must enable this thing” and now here's some… It's actually a $0 program, but it gives you unlimited access to staff and all these other things. And the point is everyone's watching to see how fast you can get one of these reactors through safely.
John (19:51):
So are you guys going the DOE route first?
Julia (19:54):
Yep, we are. And the DOD route. I mean to say the Army is now also able to license its own reactors alongside the Navy, which already does. So that will absolutely be our first step. The NRC really only regulates civilian electricity. So civilian power, so we actually don't need to even go there yet.
John (20:15):
I see. So you could be selling a lot of reactors and never go through the NRC.
Julia (20:18):
We'll basically do our test reactors with the DOE, alongside the DOD who has the jurisdiction to regulate, which is great.
John (20:28):
I didn't know that. So this is the rare bipartisan issue.
Julia (20:31):
It's a true… Kind of a gem.
John (20:34):
I can paint three clean energy futures. One is the solar and regular renewables view. Casey Handmer, who was just on this podcast. And solar will get very cheap. Batteries will get cheap enough such that you can go all in on solar and batteries, maybe some other renewables in the mix, but there you go. That's it. Nothing will be able to cost-compete with that. That's one vision. Another is that we will rediscover nuclear fission and we will figure out the spent fuel and everything like that, and we'll just do what we used to do, which is build nuclear power plants at larger scale and that delivers us enough clean energy. A third view is that fusion will work and we'll actually have nuclear fusion at a useful meaningful scale of the grid. Between those scenarios, how would you handicap them?
Julia (21:25):
I think it's… I'm going to leave fusion out because I’m not enough of an expert to know when and how that will happen. I would love for that to happen. How exciting, right? Okay. Between the first two, we do a lot of market manipulation of our energy markets and there's a little bit of flavor of the administration that does happen, especially between renewables and oil and gas. Nuclear has been a little bit actually immune to that.
So there hasn't really been a true market for this stuff. Renewables have been heavily subsidized for a while and nuclear has not been. That did change in 2022 with the IRA. There's now production tax credits for nuclear, which has a lot of ripple effects, but I think we're going to see a lot more progress with solar and batteries in particular. There is a reason, however, that the hyperscalers who are committed to their climate targets are not just going for solar farms. Why are they even thinking about nuclear if solar is so cheap? And it's because actually to get true a hundred percent 24/7 coverage, the battery or the excess build-out is massive. It is really expensive. It's basically on par with nuclear costs. So we're not there yet. It's one thing to say. The other is that the cost of solar at the panel level is dirt cheap and installing it dirt cheap, but to actually integrate it, build the transmission, and then the added grid complexity where you need these peak or gas plants to run when the clouds come and that kind of thing. That actually has a ton of costs and it's sort of hidden costs. I think it's underappreciated certainly by the solar-maxis out there. It's real. And so I think we're always going to want to have multiple sources of energy, right? It's like diversity is good. It's like if one thing… If natural gas prices are through the roof, like, hey, we have these other things. We have this base load nuclear, which has been just so stable for us, and I think we're going to want to have everything. I actually don't think there's going to be a world that's going to be a hundred percent nuclear.
John (23:24):
But isn’t diversity of energy sources valuable because the solar maximalist would say, “No, it's fine. Solar will just take everything.” And wouldn't the nuclear maximalist say “We figure out how to at scale produce nuclear power plants and why do you need that gas? Why do you need anything else?”
Julia (23:40):
It's a good push if you are able to own through and through your whole supply chain. You make the panels, you mine and process the minerals, which we have zero interest in doing as Americans. All that dirty work is done in China and elsewhere. If you really did all that yourself, then I could say, “Okay, you own it. You can do it right now.” At least you have, hey, if you're natural gas pipelines get shut down. Actually, it's not a problem for us. We're not even an importer anymore. We're actually good. This is actually the reason that… Do you know what percentage of France's electricity is from nuclear?
John (24:15):
It's declining, right? But historically it's been around 70%.
Julia (24:19):
Yeah. Do you know the US?
John (24:22):
US from nuclear? Oh, it's small. It's five, 10%.
Julia (24:25):
It's actually more.
John (24:26):
Really?
Julia (24:26):
Eighteen. And everyone thinks it's below five, 10%. When I got into this, I thought it was like 2%. But anyway, my point is that France made a very, very deliberate decision to actually go to 70%, if not more at the time in the seventies because they had no carbon resources. So they didn't have the ability to say, “Oh, we'll just dig in our backyard and we'll be fine.” And actually uranium, while it does require digging out of the ground or whatever, is so much easier to store. It's a million times as dense as coal. So it's a lot easier to work with than… The reason people get uptight about oil and gas supply is that it's not easy to store, and if it's not easier for you to come by, you better well have good geopolitical relationships.
John (25:07):
Yes. Yes.
Julia (25:07):
So anyway, this world and where it's headed, I think it's still going to involve multiple sources.
John (25:19):
Julia's work involves managing a lot of risk: unsolved physics, complex engineering, and a lot of regulatory scrutiny. Managing risk for an internet retailer shouldn't be that complex, but sometimes it can feel like it. Internet fraudsters now operate on industrial scales. They attack strategically when you're sleeping and they use AI to sharpen their attacks now. Stripe’s defense against fraud comes from its scale. When a payment happens on your site, there's a 92% chance we've seen that exact card before, so it's easier for us to spot misuse. And we recently expanded Radar to block other kinds of fraud, like free trial abuse, promotion abuse and refund abuse. Stripe now processes trillions of dollars of payments each year and every transaction in the Stripe network makes the next one safer for everyone. So you can focus on your product and not on developing a world-class counter-fraud division to let Radar handle the fraud. So you can get back to building your product. Go to stripe.com/radar.
Tell me about the supply chain for nuclear fuels.
Julia (26:15):
Yeah, I mean, so you dig uranium out of the ground and then you need to convert it to a gaseous state so that you can enrich it. You put in a centrifuge, it goes like this, and the heavier isotopes go out and the lighter ones stay in and you can pick those out and make those into your fuel. The mining part has been something that we've hardly kept alive. I mean there was at one point a couple of mining operations going in the US where the last 10 years or so, it's been like 1% of nuclear used in the US was actually domestically mined and now—
John (26:46):
Where are the last uranium mines in the US?
Julia (26:48):
They're places like where you would think. Texas, Wyoming, those sorts of places, in the Western states. But now again, like everything else have said, “Okay, wake up. Let's actually reshore some of this stuff.” So we're starting to see more uranium mining. Conversion and enrichment is still largely done overseas. And yeah, we're still buying almost 50% of our uranium from Russia and ex-Soviet states. So I think it is important that we're thinking about how do we have far more of a domestic supply chain? We have the ability to do it. We have the uranium and actually there's a little bit of Silicon Valley DNA entering the space with General Matter, which is Scott Nolan's company out of Founders Fund. So he's getting into the enrichment space, and so I think—
John (27:30):
And they're a uranium enrichment company?
Julia (27:32):
They're an uranium enrichment company, which is real picks and shovels for a real hard atoms business to do that. But I think it's really meaningful work. I think it's really important.
John (27:44):
Do you care more about the mining, the enrichment happening in the US or just there has to be an independent supply chain for the whole thing?
Julia (27:51):
I think you just want diversity. You don't want everything from one place, and I think we do want to have capability. So I think all of it should be at least partially done here and with our allies.
John (28:01):
How did you end up starting Antares?
Julia (28:02):
Gosh.
John (28:03):
Most people do not happen upon this kind of idea.
Julia (28:04):
No, you don't really happen upon it. I was working at SpaceX at the time when I first started reading about nuclear, and it was when I was actually working on getting Starlinks into Ukraine. So after Russia invaded Ukraine. And I just started reading more about the conflict there, probably more than I normally would have because I was actually speaking with Ukrainians and thinking about what was going on there. And I was a history major, so I think there was just a little bit of like, oh, geopolitics, resources are important. Russia has a lot of resources. Randomly, my cousin is the CEO of a glass factory making literally glasses like this, and he is like, “We're going bankrupt because of energy prices.” And you're like, “Golly, what is going on?” And then I read about Germany and the nuclear power plants. They were shutting off to turn on coal in a place that is like… Green party. Crazy, right?
So I was like, what could be so compelling that someone would do something so counter to their own good as a country, as a nation? And at the time, again, I knew nothing about nuclear. I didn't know the difference between fission and fusion, honestly, at the time. I certainly didn't know that it was a pretty big part of our grid and other things, but it just was one of those completely under-celebrated, under-championed by anyone. The climate tech… Climate people really weren't into it either. And so it just kind of got me curious about it and then just started reading more, wrote a piece that I posted on Twitter, and it's actually how I met Jordan, who's my co-founder and he's the one who actually had the vision for micro-reactors for the military. And I just said, what a great beachhead. Again, premium power, great thesis to go to market with and just really got into the topic.
John (29:42):
How did you arrive at that as a beachhead?
Julia (29:44):
I think you just say, where can you go that you don't have to compete on that LCOE level against the cheap solar panels? Where do you go where there's a real need and mission assurance for something critical to the military? They will pay for that. This is also an area where they were obviously behind the Manhattan Project. They're here to help develop new technologies, and so what a great partner to have when trying to do something like this.
John (30:11):
Has DARPA played with this stuff at all? Or has it come from… It sounds like you're working with the main branches of the military.
Julia (30:16):
Yeah, we are not working with DARPA, mostly because it's actually not as technically risky as you might even think.
John (30:24):
They prefer things further at the risk curve.
Julia (30:26):
Exactly. We actually, in a funny way, it's not even really a tech problem. Totally. I mean it's hard. It's hard to build these things. It's not—
John (30:34):
But someone just needs to go out to—
Julia (30:35):
It’s not novel in that way.
John (30:37):
That's interesting. What lessons have you brought from your prior experience at SpaceX, at Opendoor to building Antares?
Julia (30:44):
One of the things I really took away from SpaceX was just really getting into it. It's a very confrontational culture and it's not afraid to say, “Screw your next meeting. We need to keep talking about this.” And I love that. I want to bring that energy everywhere I go. It's like if you are unsure about something, if you're questioning something, you should probably stop and dig into it. We had this thing called the “five whys”, I don’t know if—it's like one of these Silicon Valley little quips, but it's like that on steroids, and I think that the pausing to probe super deeply, question things, really hash things out was a big takeaway. I learned like, wow, it's really effective at getting to the right answer, at making a decisions, at changing course when you need to. And so I've tried to bring some of that to Antares.
John (31:42):
Investors are hot and cold on hard tech depending on when you're raising money and everything like this. Presumably there’s a J curve in investing in Antares where there's definitely some R&D before you eventually start paying it back with revenues. Just how has the experience of raising money for this?
Julia (32:02):
It's going to get harder. I think it hasn't been hard to date only because there is interest and excitement around nuclear. It's also just fewer dollars in the beginning. You're kind of in the honeymoon phase of all of that. I think it will get harder. I think now the time’s a ticking. It's like, okay, let's see your next technical milestones. This criticality test. Obviously the full reactor at the end of 2027. I think it will be challenging. And there are several reactor companies out there, so it's a bit of a race. I think it's going to be fun. I mean, I wonder what Polymarket would have to say.
John (32:37):
Several reactor companies out there, even at the end of the market that you guys are going after.
Julia (32:42):
Everyone has their different niche. I think our niche is, again, the smallest, the most DOD-oriented. There's a lot of micro-reactor companies or SMR that are going after data centers and the data center buildout, which is an amazing market. I mean amazing. I think people will pay a lot more. Again, going back to the premium power point, they will pay more for power for this. Not exactly our target market.
John (33:07):
So your niche in the market, it seems like you guys have selected a part of the market where maybe there's less competition.
Julia (33:17):
There's still some people going after the military applications. There always are. I mean, if you have a good idea, there's going to be a few people working on it.
John (33:23):
How capital-intensive is it going to be to get your product to market?
Julia (33:27):
In the hundreds of millions. It depends on what you mean by to market. So I think our big milestone is like, can you build a reactor, prove it, turn it on, it's producing power. Then I think you can think about interesting layers of capital in terms of pre-orders for your reactor.
John (33:45):
This was going to be my next question. When does the money arrive?
Julia (33:47):
Yeah, exactly. Bigger government grants. So we're kind of still in the hundreds of thousands of… Single-digit millions of dollars of grants that we are getting right now. But you're moving into the tens of millions of dollars R&D grants, and then you want to eventually move into that kind of program-of-record real dollars. There's a really interesting program right now that the Army has put out called the Janus Program. Are you familiar with the COTS program that SpaceX used to bring to kind of develop their first reactor?
John (34:16):
Not by name, but I'm familiar with—
Julia (34:17):
Commercial Orbital Transport System.
John (34:18):
Yeah, yeah. I'm familiar.
Julia (34:19):
It was like the innovation basically to say, “Let's try something that's not cost plus and let's actually do something that's milestone-based and basically we'll cut you if you don't hit the milestone. And it’s fixed contract, we're not going to pay you for whatever.” So anyway, the Army has taken that model a bit and I think used that as inspiration for this Janus Program where they're saying we're going to fund up to three to five companies to build their first prototype alongside us, kind of out of base. And then we will choose a winner and we want to build five to 10 reactors, if not more, on bases. So they're like, the demand signal now is hot. This is great for our future fundraising for the industry in general. It's very, very good demand.
John (35:02):
A lot of ink gets spilled on military US defense contracting. What is your experience as someone who's attempting to sell to the military been?
Julia (35:12):
I mean, people are actually kind of well aware that this cost-plus model is not great. The question is like, well, what are the alternatives? Who's going to go do it? Anduril has been an amazing example of that. So I think a lot of us in hard tech really look to them as what we try to emulate. Every branch of the military has these innovative arms that tell you, “Hey, we want to hear about your ideas.” And they do, actually. You can go talk to them. They're giving out grants for things. I do think everyone wants the innovation. In this day and age of information everywhere, you can learn a lot quickly. They do want to work with startups, so I think we'll only see more of it. But you have things like the government shutdown and the pace of government is slow. So you got to just kind of be able to keep pushing those timelines, again while not running out of money, to be able to make it through that valley of death.
John (36:25):
What's been harder than you expected building Antares? Because presumably you expected it broadly to be hard. You expected there to be a lot of technical risks—
Julia (36:33):
Harder than expected. There's a couple things. I mean, I think, gosh, recruiting is always hard. You're up against the SpaceXs of the world that have just a lot more credibility when you're the no name person at the very beginning. It can be hard.
I think there's a lot of complexity, and even though I said we're making so much progress in regulation, there's still red tape everywhere. Things cost more than they probably should. We spend a lot of money paying the national labs to do things like clear out a pit so that we can test our reactor. And it's like we're paying for that and they're like, here's hundreds of thousand dollars. Here's the bill. So there's stuff that you're like, “Gosh, there's really a lot of expense everywhere and it just slows things down.” And so I guess we were somewhat ready for that, but some of this stuff just seems silly and when you're building in software, you don't have to worry about it. You just sit there code and do your thing and go right to market. When you have to work with the real world, you have to kind of play their games a little bit and you have to follow the process in some ways. And there's not always a way around it. Sometimes it's just through.
John (37:47):
You said there's still a lot of red tape despite the fact that there's been the first Biden-era and then Trump-era improvements. If you were in charge and no Energy Czar, or we'll give you a role with enough powers across all these departments, what would you change? What do you think the specific improvements we could make are?
Julia (38:06):
If I were fully in charge of energy, I would love to see the US government be a buyer. Just like the NASA was a buyer for SpaceX's eventual launch product. I think having the US government be a customer for nuclear would be great. And so we have this a little bit with the DOD. It was a very interesting announcement that just came out a week or two ago that doesn't have a ton of detail, but it's starting in this direction. I think it's very positive. I'm a small government person, so I don't say this lightly, but I do think there are certain things when it's national security related and things like energy resilience, there should be public-private partnership. But anyway, the announcement is that the US government now wants to spend $80 billion on nuclear development alongside Westinghouse. So they’ve basically chosen the technology, and I think that's actually a good thing, because you want to be able to build one thing and then repeatedly build it so that you are not reinventing the wheel every time, you actually have economies of scale, the learning curve that you go down from doing that. So I'm excited about that news. I think that's the type of policy I'd want to see more of. And we will see how it goes. I mean, this is the biggest ever commercial nuclear public-private partnership we've seen.
John (39:24):
And sorry, specifically they're buying from Westinghouse, just a large amount of nuclear power?
Julia (39:28):
Well, it's a combination. So this is where, I mean the details are a little fuzzy, but they've announced $80 billion. They've announced Westinghouse as the partner of choice. Not all those dollars are going right to Westinghouse. Granted, and that's a company that makes just the reactor itself. So there's money that you would spend on a plant or licensing fees or whatever else that surrounds that, but I think it's really positive. I think what it does is it's a catalyst for the whole industry, like supply chain, workforce. People are now paying attention like, oh, there's real dollars to be made here.
There's a huge trillion dollar asset under management company called Brookfield Asset Management, and they've actually just gotten in the game as well in the last couple of weeks. I'm sure there was talk, right? To take the V.C. Summer plant that never got completed in South Carolina and actually finish it. And so again, they now have some nice tailwinds with those new production subsidies. They are partnering with the local utility to, and again, a local utility doesn't have the risk appetite to do this themselves or the balance sheet to do this themselves. And so there's some really nice, I think, market signal you get when the government's saying, “We're going to spend 80 billion on this,” but I think it gets the rest of the industry moving.
John (40:45):
As you guys were exploring and developing the early stages of your business plan, did you consider for a product that is as tightly regulated as this, did you consider doing any jurisdiction shopping where you develop a product that you sell in Singapore or the UAE or Prospera or some more favorable jurisdiction?
Julia (41:07):
We really didn't. It felt like there was enough signal that we would be able to get through regulatory and see some positive change. There was enough basically demand signal within the US and I don't know, we take pride in also building for our country and it feels good to be developing something for our military and our allies. So yeah, we didn't really shop it around too much.
John (41:31):
And I guess the US military is a large buyer, in this case.
Julia (41:33):
Very large buyer, the largest buyer of non-commodity energy, if not all energy.
John (41:38):
How soon afterwards will you go to… I presume you want to do civilian premium power use cases, like you're saying an oil rig seems like the obvious one that ironically, that it would be hard to get gas. It would be hard to get refined diesel or gasoline to an oil rig. Does that come right afterwards or is that such a different bottlewax from a regulatory point of view that maybe it's like 10 years later?
Julia (42:06):
You're developing concurrently. We are talking with oil and gas already, just understanding them. What do you need? What do you think you'd pay? That sort of thing. So I would say it fast follows, right, a few years later. I think it's going to be such a nice proving ground. If we can make it through and test on a military base, for example, and get a first reactor live and get some revenue from that, then I think it's opened the aperture pretty broadly for what else is out there.
John (42:33):
Does this movement towards reinsuring manufacturing in the US and various blockers people talk about to that between supply chain is difficult in… You've talked about in the fuel context, but also supply chain for manufacturing, talent and the roles that people go into and finding talent in specific areas. I'm just curious, what has your experience been of building a hard tech company in the US?
Julia (43:00):
Supply chain is… Yeah, the bottlenecks. Supply chain is challenging. When you need to work with nuclear grade material and no nuclear plants have really been built. It makes for a very limited number of suppliers, which means costs are higher. Options are just fewer. So if they're not responding to your email like, well, what are you going to do? There's only one other option. And I don't know, they're busy or whatever. So those dynamics are really challenging. The graphite that I brought here is actually from a Japanese company. No one's even processing that stuff in the US. And so you're literally going all over the world for these supplies. So that's challenging. Talent is challenging in the sense that there are so many options for hard tech in the LA region, so you need to have a compelling mission, but nuclear is compelling. So it's been challenging, but not in some ways. If you want to work at SpaceX and you love space stuff and you want to be at that name brand place, we've lost you. But if you're interested in nuclear and being able to build from the very beginning, that's a compelling piece.
John (44:02):
Yes, it's the Shackleton pitch of—
Julia (44:05):
Totally.
John (44:06):
High risk of failure. And that's the pitch that you're making. I like that. What trends in nuclear should we be excited about right now?
Julia (44:14):
So the trend I'm excited about is bringing back online these recently shuttered plants because they don't take too much money to bring back online. And we've seen a couple big power purchase agreements where you have Google partnered with I think NextEra or the Iowa-based utility to bring on Dwayne Arnold, which was shut down about five years ago. And then you have actually Three Mile Island. The site of the infamous accident is also—
John (44:40):
Almost accident.
Julia (44:41):
Well, it was an accident but it was not really a big deal. This much radiation came out. It was the amount of a chest X-ray. Completely overblown.
Julia (44:50):
That's being brought back online, partnership between Microsoft and Constellation, I believe. So that's kind of fun. I mean, you're seeing, let's go take these assets that are just sitting there that shouldn't have been turned off. Bring them back online, bring them back to life. Tons of power coming out of them, and we need it yesterday, right?
John (45:09):
Yeah. I mean, it's so striking. The big trend in the US right now, being that grid demand was flat for the longest time. We were completely not adding power generation nor adding power transmission. And now because of data centers, essentially we have this enormous boom that is just leading to things that maybe previously would've been competitive. People can't get enough of anything. People can't get enough of nuclear, they can't get enough of solar, they can't get enough of natural gas. Just everything is backlogged. It's really interesting.
Julia (45:43):
Even the natural gas turbines are backlogged.
John (45:45):
Totally.
Julia (45:45):
Multiple years.
John (45:47):
Across all the manufacturers.
Julia (45:48):
It is hard when something like this goes into one of these inflection points. And to your point, it was less than 1% year over year growth for 20 plus years. And now it's like people are estimating 5% annual growth next year. Oh my gosh!
John (46:01):
Last question. What kind of culture do you want to build at Antares?
Julia (46:07):
Well, I think it's certainly one of excellence, and I think that means delivering quality product. How you get there, I mean, there's a lot of ways to get there. You can have the combative SpaceX culture. You can have a less combative, maybe more, I don't know, writing culture. They never write that much stuff down in SpaceX. To me, that matters less. Although I tend to be one for a working environment that is energized, positive, preferably. But I think ultimately I would be proud to be delivering on an excellent product that can be delivered to market soon, and we can all go use.
John (46:44):
But okay, so you want the results-oriented culture—
Julia (46:48):
Results speak for themselves in a lot of ways.
John (46:50):
Yeah. Well, Julia, thank you.
Julia (46:52):
Thanks for having me.