Redesigning Nuclear Reactors for Mass Manufacturing with Aalo
Matt Loszak is CEO and co-founder of Aalo Atomics, a startup designing 50-megawatt nuclear power plants made of pods of five 10-megawatt reactors each, targeting the data center market.
Inspired by the MARVEL test reactor that his CTO Yasir Arafat led at Idaho National Lab, Aalo pursues parallel DOE and NRC pathways, betting that factory-built, pad-ready nuclear plants can slash cost, schedule, and grid-interconnection delays. Now 45 teammates strong and raising a Series B, Aalo aims to break ground on its first plant within 12 months and go critical the year after, powering the AI era with compact, clean, always-on fission.
Episode recorded on April 23, 2025 (Published on May 19, 2025)
In this episode, we cover:
[02:31] From SaaS to fission: Matt's founder journey
[07:08] Yasir’s path and the Marvel reactor at INL
[08:56] How national labs help launch commercial nuclear
[12:12] Aalo’s dual regulatory path: DOE + NRC
[13:45] Why Aalo is building reactors for data centers
[17:19] Designing for mass manufacturing, not megaprojects
[21:50] Security differences across nuclear facility types
[25:03] Fuel and enrichment: what Aalo had to consider
[28:02] Switching to LEU+ for supply chain resilience
[31:04] Comparing XMRs, SMRs, and micro‑reactors
[33:25] Inside the Aalo Pod: 5 x 10MW modules with built-in redundancy
[37:15] Regulatory updates on NRC civilian licensing
[38:43] Nuclear + AI: mutually transformative technologies
[41:35] Hiring at Aalo: who they need next
-
Cody Simms (00:00):
Today on Inevitable, our guest is Matt Loszak, CEO and co-founder at Aalo Atomics. Matt and his team are building what they call XMRs or extra-modular reactors. They're designing 50 megawatt nuclear power plants made of pods of five, 10-megawatt reactors each, targeting the data center market. The reactors themselves are inspired by the Marvel Reactor Design from Idaho National Labs, for which Aalo co-founder and CTO Yasir Arafat was chief designer and project lead. Aalo uses a liquid metal sodium design to extract heat from the reactor core, which they believe increases efficiency and energy output compared to a traditional light water reactor. They're pursuing regulatory pathways with the Nuclear Regulatory Commission for commercial data centers, and with the Department of Energy for on-site federal facility usage. Aalo announced a series A in mid-2024 that included Fifty Years, Valor Equity Partners, Harpoon, Crosscut Ventures, and others. From MCJ, I'm Cody Simms, and this is Inevitable. Climate change is inevitable. It's already here, but so are the solutions shaping our future. Join us every week to learn from experts and entrepreneurs about the transition of energy and industry. Matt, welcome to the show.
Matt Loszak (01:38):
Thanks so much for having me, Cody. Excited to be here.
Cody Simms (01:40):
I'm looking forward to really getting an update from you on what you're building at Aalo, and let's start there, because I know everyone I talked to pronounces it differently, and you just clarified for me it's Aalo like all the things.
Matt Loszak (01:54):
That's right. Yeah, it's a word that means the light, so it has a multi-layered meaning for us. On the one hand, we're trying to push towards this bright, optimistic future for humanity, and then it has a few deeper meanings for us as well. For example, we try to hire optimists, people who believe that what we're doing is possible, and want to work really hard at achieving it. There's a few other secret hidden meanings as well that might come to light one day.
Cody Simms (02:18):
Wow. Multi-layered. What language is this from?
Matt Loszak (02:23):
It's Bengali.
Cody Simms (02:24):
Bengali. Wow, cool. I think that is the only Bengali word I know, that I now know.
Matt Loszak (02:29):
Well, it's a good one to know, yeah.
Cody Simms (02:31):
All right. On the notion of optimists, I want to go back to when you and I first met, which is the summer of 2022 I believe, and we got introduced by Ange at Version one Ventures. She had met you I think as a fellow Canadian entrepreneur, and you had just written this post on Medium, which is titled Why I left the 150 Employee SaaS Company I co-founded to get into nuclear power. And I think you obviously wrote that drawing on the fact that you were an entrepreneur, but you also, at that point, I think had very little background in nuclear itself. But, what's amazing is Ange connected us. But then, I feel like at MCJ, we had three or four other triangulation points of like, hey, have you met this Matt guy? He's like doing a nuclear thing.
(03:17):
And so, you very quickly went out and built a network, and just willed your way to becoming a founder in the nuclear space. I want to understand more about that because you now are running, you're in development on a full-scale nuclear reactor. Why don't you describe the journey that you went on, maybe starting with your entrepreneurial path, and how you became a founder, and then what ultimately convinced you that you could go do this thing?
Matt Loszak (03:43):
It's funny because for a lot of people, they see this guy who had this HR payroll benefits software suite turning into a nuclear founder and thinking, what the heck is that? Looking at the whole life story, it's interesting how nuclear has woven in and out of my life story all throughout. And the earliest realization for me of how cool nuclear energy is was actually when I was early teenager, and before that, I had pretty bad asthma, breathing attacks. And that was because at the time, Ontario, we had a lot of coal power plants. We actually used to have around 62 smog days per year. But then, very fortunately, the government decided to turn off those coal plants and go all in on nuclear, and it ended up being around 60 or 70% of our grid with the rest being hydro and basically entirely clean. And we went down to zero smog days per year, and my asthma went away.
Cody Simms (04:37):
Oh, wow.
Matt Loszak (04:38):
I thought, my God, this is a very cool technology. And that was part of what fueled my passion for engineering, and physics, and science. I ended up studying physics and engineering in university at Queens, took courses on nuclear, but also other things like relativity, and quantum, and even had a lasers course, which I thought was pretty badass.
Cody Simms (04:57):
Awesome.
Matt Loszak (04:58):
After graduating, I thought it's probably pretty hard to start a business right after graduation in the nuclear field per se. My first business I did was a painting business. I decided to torture myself in my hardest year of my physics degree, and simultaneously run this house painting business. I would study until midnight and for study breaks I'd be doing cold calling to draw a business. I hired 30 employees for that business, and we did $100,000 of painting homes, and I learned that, hey, business is really cool. It's cool to work your butt off and have some success there, but painting isn't the most interesting to me. But, I was like, yeah, maybe one day I can combine all these interests. And that was always what drove me is finding a way to dabble in things and see what I really loved, because I tended to do better at things I really loved.
(05:45):
And so, I liked business, I liked science, and engineering, and physics. But then, yeah, because starting a nuclear company after graduating is hard. I thought, hey, let's start in software. I did one company that you might call a learning experience. The second one I started in 2016 with a co-founder, Kevin, so that was the Hume, so that was HR, payroll and benefits software, and we became a pretty big HRS player in Canada. We processed over $7 billion of payroll. We actually just had an exit for over a hundred million.
(06:15):
It was a great learning experience, and a lot of fun, and learned a lot about structuring companies, and scaling, and recruiting, and all those good things. I wasn't the most passionate about HR software. And so, I think that blog post that you're referencing was what I had written when I was deciding really what is the culmination of all my deepest passions in life? That really was Aalo. And so, that's the long story, and if you want, I can dive into how I met Yasir, but maybe if you want to dig into anything else first, happy to do that too.
Cody Simms (06:46):
That's a great overview, and congrats on the recent exit of Hume that happened a few years after you've since departed, but a testament to the company that you helped build. It ultimately hopefully turned into a good outcome for you and for the folks you brought on board as well. How did you meet Yasir? He is this very accomplished nuclear engineer. Maybe give us a bit about his background and how you two came together.
Matt Loszak (07:08):
Yasir was the chief architect and project lead on the Marvel program at Idaho National Lab. This was like a scaled down test reactor, one of the first advanced reactors that was under construction in the US in decades, and it was the first reactor that was ever approved by the DOE for construction. Usually the NRC is the regulator, but for government labs, and other forms of non-civilian soil, the DOE can be the regulator. Yeah, Yasir, before starting Marvel, he also was the person who formed the Vinci program at Westinghouse, which was their micro reactor entrant and still is. And before that, he did a lot of R&D and worked a little bit on the AP 1000. Westinghouse was where he was the longest after his university degree. For both he and I, we were both interested in nuclear for a long time.
(08:01):
Obviously, he went right into the industry after university. We were both entrepreneurial minded. I think he also always wanted to start a business. But, when we met, it really felt serendipitous because we had a very complimentary skillset. We had very similar values, and a view of what nuclear should become, and even a similar sense of humor and just really got along quite well, which I think is important for a co-founder who you're doing this business with for the long run. We met and decided that maybe at some point it makes sense to start a company together and found a way to make it work.
Cody Simms (08:36):
Help me understand the pathway for building a research reactor at INL like Marvel, and how it then becomes a commercial project. Does the government still play any kind of role in the ongoing commercialization of the technology? Help us understand how that all works.
Matt Loszak (08:56):
Yeah, so Marvel was never really intended to be commercialized directly. What it was really meant to do was help catalyze the industry through providing data that can be really helpful for a lot of different players in the industry. When Yasir was going to pitch this idea to the folks at INL and DOE, the whole intention was saying, "Hey guys, in the US, we haven't built one of these advanced reactors in a long, long time." First of all, maybe it makes sense just to show that we can do it. There's a lot of value in that. It might sound simple, but just the very act of doing that helps to rekindle and revitalize, rehydrate these dried up or dead aspects of the supply chain and the muscle of the industry to some extent. That was a big part of it. There was another program called Pele, which is also ongoing, and that's a high temperature gas reactor. Marvel is a liquid metal cooled reactor.
(09:49):
These are both programs that are really trying to help the US just get back into building things. This program is all the data. A lot of the design aspects are readily available to those who want to leverage certain parts of it, so I know some companies have used some parts of the control drum architecture for Marvel. Others have used different parts of the fuel data. These are all meant to catalyze the industry.
(10:13):
And so, what's been helpful for us is the idea that with Marvel going through that whole process, it's caused the INL and DOE to really think through how to do a lot of these things that are in the minutia of building a reactor that might not have existed had that not been happening. Certain siting processes, certain environmental processes, regulatory processes, supply chain processes, these are all things that once Marvel and Pele were going through these procedures, now all of us other reactor developers have the benefit of going through a more streamlined process, because this was done recently. It's really beneficial for the country for all these different vendors.
Cody Simms (10:54):
And for folks who want to understand more the INL side of how Idaho National Lab is thinking about these programs, we did have Dr. John Wagner on as a guest. I think that was your intro, Matt, originally, maybe 18 months ago or so as a guest on the pod. And it was a great episode, really full of a lot of insights about how the national labs work, and in particular how Idaho National Labs is working in a lot of their focus on advanced nuclear reactors. And it sounds like you all have a continued very deep relationship with INL. I think you're using some of their testing facilities still in a significant way, is that right?
Matt Loszak (11:27):
For our first plant, we are doing DOE authorization as well. This is the Aalo-X, it's our experimental reactor. Like I was saying, leveraging a lot of these pathways that have just been established, and now can be used in a more streamlined way. And so, the lab has been great to work with. There's a number of other vendors building test reactors at the Dome, which is a famous site. It's where EBR-II used to be. There's a few vendors that are doing NRC authorized plants in the periphery of the land of Idaho National Labs. There's a lot of different people trying different things. It's pretty exciting.
Cody Simms (12:01):
And will the first wave of NRC work that you need to do be part of that Aalo-X project that's inside INL, or is that purely a DOE authorized reactor that you're building?
Matt Loszak (12:12):
For us as a company, our strategy on regulatory, we're pursuing the two in parallel, because DOE is for non-civilian soil, NRC is for civilian soil. Our estimation is that DOE might be a little bit more streamlined. And so, we think that plant might go critical before the NRC regulated plant. But, the nice thing is all the engineering work, and the math, and the physics applies equally to both regulators. It's just packaged up a little bit differently. And one really interesting thing that's actually brewing that I think is worth highlighting, this is hot off the press from the past few weeks even, is we always perceived the DOE authorized plant to be like our calling card, a good way to get the first one built, prove things out, and then scale rapidly thereafter. But, we never saw it as a major deployment strategy for scale.
(13:01):
However, in the past few weeks there's been a bit of a sign that Trump and Energy Secretary Wright had been pushing for the deployment of a lot of co-located nuclear and data centers on non-civilian soil. And this is pretty interesting. I think it's coming from a place of really trying to compete with Russia and China on AI and nuclear to make sure we don't fall behind frankly. What's really interesting is if this becomes almost like a wartime effort of keeping up in this really critical technology, then the DOE authorization pathway could become way more relevant for scale than we initially estimated. Really interesting space right now.
Cody Simms (13:40):
Super fascinating development. Why don't we take a step back? Describe the Aalo Reactor.
Matt Loszak (13:45):
For us, we really focus not just on the reactor, but really the whole plant. The way that we always think of this is that the reactor is not the product. The plant is the product. And the reason this is relevant is, I think it makes sense to kind of take a step back and outline how we see the lay of the land in nuclear. Bear with me as I just rant on this for a second. We think that the ideal way to deploy nuclear in the west, and really the world, is to mass manufacture it, make these plants and factories. And if you look at nuclear today, it's a lot of bespoke onesie, twosie plants. There's never really been a factory that's mass-produced reactors and deployed them widely. And you can ask why this is. We think that it's largely because in the past, there was this more government-driven utility model where there wasn't really the right type of demand.
(14:35):
And taking a step back further, you might want to ask, well, what is the right type of demand? What will drive the generation of these factories? And the AI data center movement is so incredibly unique and well-suited to be the guiding force that bootstraps these factories for two main reasons. One is it's the fastest growing, and two, is it's the highest willingness to pay. We're talking about bringing on tens of gigawatts, which is like tens of new cities. Each gigawatt is like a million homes, but in a very short period of time, maybe in the next 10, 15 years. Because these data centers are so lucrative, they have a really high willingness to pay for power. This is a really, really awesome recipe for getting this factory set up, because everyone knows that the first of a kind is going to be more expensive than the end of a kind.
(15:22):
You're going to need a customer willing to pay a bit of a premium for those earlier versions. But then, once you get that set up, and you've got the repeatability of the factory and the predictability on cost and timeline, it's really this incredible situation you can be in, because then you can come down that cost curve, and then compete on all sorts of other different markets beyond that initial wedge of AI data centers, so good old-fashioned utility power. And then, anything from even new markets like industrial process heat, desalination, chemical processing, all sorts of things that never would've been possible without the right type of demand to get the factory set up to begin with. And I think it's also worth highlighting why the factory is so important. Again, if you look at all these bespoke reactors that were built before, the hard thing is a lot of people look at nuclear and ask why we're not building more of it.
(16:13):
On the surface, they assume, well, maybe the regulator is stopping us. The thing is, there's actually eight utilities that already have approved licenses to build large water-based reactors, and they're choosing not to, because of a fear of cost and schedule overruns. Again, the reason why the nuclear plant factory vision is so compelling is because it can solve that main problem around the challenge of predictability on cost and timeline. We're really in a really special moment in history now where nuclear really for the first time in its history, even in the early days, has this chance to set up this infrastructure which could allow it to be just way easier to deploy. And I think this would be a very good thing for the world because it's clean, it's base load, it uses very few resources, very little land. It works anywhere. To me, and I'm biased, it really is the ultimate energy source for humanity. I'm really excited because I think the window is open to do something really special here, and we're working our butts off to achieve this vision.
Cody Simms (17:12):
What about your design approach makes it mass-manufacturable in a way that a light water reactor is not?
Matt Loszak (17:19):
Light water reactors have been attempted to be manufactured in factories before. The large water-based reactor that was built recently was fairly modular. One of the challenges there was that it wasn't fully vertically integrated, so there was a lot of miscommunication and that's just a challenge. If you build large things, you very quickly realize no matter how good separate vendors can be, there's just inherent challenges to integrating lots of different vendors.
Cody Simms (17:44):
That's my perception is a lot of the challenges with the Vogtle plant, for example, were challenges of project management, maybe more so than challenges of engineering per se.
Matt Loszak (17:54):
Yeah, exactly. It's worth highlighting, this is not purely a technology challenge. But then, also, you can ask yourself, well, what technologies would be really well suited for mass manufacturing? And for us, one key thing is we think a lot about supply chain, and availability, and readiness to scale. And one big change we had in the past year was we shifted to a more readily available fuel because just in the name of speed and economics. I'd much rather have an imperfect reactor built sooner than a perfect reactor built later, because the world needs this ASAP. For example, we're using off-the-shelf uranium dioxide fuel that's in a very typical form factor that can be at a very typical enrichment level that can be mass-produced, and not only be available for the first reactor, but also be available to scale really rapidly thereafter, so that's one thing.
(18:41):
And then, another one is if you compare using a liquid metal as a coolant to water, there are a few advantages that we think we can get to. One is, it allows you to make the reactor much more compact, and the plant much more compact, which saves cost. Another one is its atmospheric pressure. The safety case is just a little bit different. Nuclear is already very safe. The whole idea here is you're going to achieve that same level of safety a bit more simply with these types of technologies.
(19:08):
And so, the reason that's not an obvious slam dunk, as it sounds too good to be true, is there actually were a number of groups that did sodium cooled reactors in the past, and oftentimes, they thought it would be this holy grail of economics, because of the compactness, and the inherent safety, and so on. There were some that did have good success. EBR-II operated for decades with very few issues. There are good success stories, but there's also a graveyard of a few different groups who just had struggles with maintenance, and operations, and so on.
Cody Simms (19:39):
Any kind of molten salts or liquid cooled metals are notoriously challenging to work with, right? You've got to iron out a lot of the corrosiveness and all of those bumps when you're actually trying to build these things at scale.
Matt Loszak (19:51):
Yeah, exactly. And the way that we think about this is, this is not fusion where there's multiple Nobel Prize winning discoveries that have to be made before it's really feasible or economical. These are engineering challenges. Maybe a good analogy is like SpaceX landing a rocket. Nothing from first principles physics precludes that. It's just it's a tough engineering challenge. And similarly, our equivalent to landing a rocket might be how fast can we plumb a leak? Maybe we're just glorified plumbers or something.
Cody Simms (20:18):
At the end of the day, it's all a fancy way to create steam, right?
Matt Loszak (20:22):
Yeah, exactly, in a lot of pipes. This is our core engineering challenge is to unlock that holy grail of economics, can we resolve any leaks or clogs in a matter of seconds or minutes instead of months or years? If we can do that, then we'll be in a really good spot, and we're doing a lot of that non-nuclear testing and proving that to ourselves de-risking in the next 12 months, now that we've built the full-scale prototype in the past couple of months.
Yin Lu (20:48):
Hey everyone, I'm Yin a partner at MCJ, here to take a quick minute to tell you about the MCJ collective membership. Globally, startups are rewriting industries to be cleaner, more profitable, and more secure. And at MCJ, we recognize that a rapidly changing business landscape requires a workforce that can adapt. MCJ Collective is a vetted member network for tech and industry leaders who are building, working for, or advising on solutions that can address the transition of energy and industry. MCJ Collective connects members with one another, with MCJ's portfolio, and our broader network. We do this through a powerful member hub, timely introductions, curated events, and a unique talent matchmaking system and opportunities to learn from peers and podcast guests. We started in 2019, and have grown to thousands of members globally. If you want to learn more, head over to mcj.vc and click the membership tab at the top. Thanks, and enjoy the rest of the show.
Cody Simms (21:50):
One understanding that I have, correct me if I'm wrong, is one of the reasons that in the US, at least, reactors have largely been these large facilities is also one of security. Nuclear reactors have armed guards and incredibly tight security parameters around them, because again, you're dealing with nuclear fuels. Is there something about your reactors and others of your ilk that can allow those requirements to be less severe?
Matt Loszak (22:20):
That's a great question, and that's one that also people don't often talk to, but can really impact economics as well. For us as a company, we actually did a couple site visits. We did a site visit to a gigawatt-scale nuclear plant, and we noticed that there were 200 armed guards, seven sniper rifle towers, all these security folks. And then, we also did a site visit to a one gigawatt natural gas plant, and that had the same size power output, but instead of 200, it had 20 staff, and maybe six on site, and it had a badge card entry.
(22:55):
There's this question of can we get nuclear to a point where it looks more like the latter than the former? That has some big economic implications. People often think of the cost of electricity as being a function of the equipment cost. But, if you have something generating power over several decades, the O&M quickly becomes a big part of that LCOE, that cost of power. There's actually some good proof points, or existence proofs, and research reactors. Some research reactors use certain-
Cody Simms (23:25):
And by the way, sorry to interrupt you, but if you take that model, if you have one gigawatt plant that require 200 staff, and you say, "Hey, we're going to build a bunch of 50 megawatt reactors," and if each one of those requires 200 staff, you're dead in the water. Your model will not work, right?
Matt Loszak (23:40):
That is fair to say, yeah. Yeah, we're certainly not planning that. And conversely, I think, and we can come back to this, but the idea of having a shared set of staff, and security, and refueling infrastructure across a number of smaller reactors is actually a compelling product for data centers because the redundancy of reactors gives them availability, and that is valuable to them, and they're willing to pay premium for that, especially if it can be delivered quickly. Let's come back to that, because I want to finish answering your other question too.
(24:08):
The existence proof I was mentioning is research reactors, there are some that are actually 10 megawatts. They don't have 200 armed guards. They actually have a key card system badging in. There are proofs of certain reactors when they have different levels of inherent safety and so on that have been able to negotiate with the NRC around having a simpler safety case, because it's warranted. Reactor technology has been developing for 70 years, and it's not all created equal, and it's all safe today, statistically as safe as solar or wind, but that can be achieved in different ways. That's the whole idea here is, in this advanced reactor space, can you make it look more like a natural gas plant operationally? And if you can, then boy, oh boy, are the economics going to look good, because it's got virtually no fuel costs.
Cody Simms (24:57):
How much of that comes down to how enriched the uranium fuel source is that you're using?
Matt Loszak (25:03):
And you're asking in terms of safety?
Cody Simms (25:05):
Yeah.
Matt Loszak (25:05):
Enrichment, generally-
Cody Simms (25:07):
Security too, just like security requirements as well.
Matt Loszak (25:09):
Generally, the thing with enrichment is that's largely impacting your refueling frequency. That's the primary operational thing is if you have high enrichment, you can let the fuel operate for longer, and then there's this whole financial analysis you can do there on, well, does the fuel cost more upfront? It's a CapEx versus opex trade-off with more refueling costs. There are some implications with higher enrichment. If you take this reductum ad absurdum where you think about the extreme case, if you had 95% enriched fuel, that would require very intense security, because that's weapons grade. But, that's the whole point of HALEU is below 20%. You can't really make a weapon out of that per se.
Cody Simms (25:50):
Correct me if I'm wrong, but most large scale light water reactors are not heavily enriched. Is it UO2, it's like 5% enriched roughly?
Matt Loszak (25:58):
Yeah.
Cody Simms (25:59):
Whereas HALEU approaches 10 to 20% enriched. As I understood it, the advanced reactors are actually using a more enriched, or getting closer to weapons grade, enriched level of fuel, and yet, it sounds like they may also have fewer or less severe security requirements than an existing utility scale reactor might.
Matt Loszak (26:22):
This would be a topic we could dig into for a while. Maybe the simple short answer is anything below 20%, you're safe. The other things to think about, it's not really going to be a concern about this kind of safety concern or security concern. It would be if it was 95%, that's a different story.
Cody Simms (26:38):
Okay, got it.
Matt Loszak (26:38):
But, yeah, you're right that a lot of startups are using 20%, we're using 5%. But, the reason we're using five is not because of security or safety, it's because of supply chain. We want to be able to scale really, really fast, and 5% is what's available, so we're starting with five, and we're going to push that up a little bit higher based on the equipment that's available at these supply chain vendors like GE, Vernova, [inaudible 00:27:00], and Urenco to just get this done.
Cody Simms (27:04):
And as I also understand it, there is also with the type of fuel that big utilities are using today in the US, there's a fairly consistent domestic supply of fuel. But, as you get into some of these more advanced fuels that advanced reactor projects are using, you start to become dependent on, for example, Russian supply chains, which I know is certainly a challenge given what's going on in the world today.
Matt Loszak (27:25):
HALEU was a cool idea from 2005 until 2022, but then Russia invaded Ukraine, and the supply chain became a real challenge, and yeah, some companies are switching off, some are trying to find a way to make it work. DOE has announced a few vendors that are going to give a morsel of HALEU for their first reactors. But, I think yes, scaling could be a challenge with that fuel decision. We'll see how things evolve on the HALEU side. Hopefully, we want there to be a big supply chain there, because we want to look at that option down the line for ourselves as well. Yeah, in the short term, we're using LEU, LEU+ for ourselves.
Cody Simms (27:58):
And you said that was a relatively recent new choice to use that fuel source?
Matt Loszak (28:02):
Yeah. In the past year, we've been analyzing this, and talking to the supply chain, because we've been ... We're going to be placing the order for our fuel in the next few months, so this has been in the works for the past year. We put out a video on U-Zr hydride that showed the inherent safety of it, and all this cool stuff. But, that video was actually ... We're like, well, we already decided to move off, but we're like, we might as well publish this video, because it's just so cool anyway. It's a really cool fuel. It's just one really-
Cody Simms (28:28):
That's used heavily in France, I believe, if I'm not mistaken, right, URZH?
Matt Loszak (28:32):
Well, it's made in France at TRIG International, but it's not really specifically used heavily there. It's used in all the research reactors around the world.
Cody Simms (28:40):
Oh, okay.
Matt Loszak (28:40):
Really quick sidebar. This fuel is so inherently safe. You can basically try to make it melt down. You can pull the control rods out, which are usually the thing that slow the reaction. It's like disabling the brakes in your car or something, and it actually just shuts itself off from the physics of the fuel. And what's really interesting is Edward Teller, the person who's behind the most devastating weapon humanity ever created, the fusion bomb, not fission, but the fusion bomb, he created that, and he also then went on to create the world's most inherently safe nuclear fuel, U-Zr hydride. I feel like one day someone should make a movie about this stuff.
(29:14):
Imagine being a physicist. Physicists don't want to kill people, right? That's not what they want to do. But then, the government and the military was like, "Hey, you guys have this technology, why don't you guys go make a weapon?" And they're like, "No, but we want to make energy. This can actually solve so many of humanity's challenges." It's just cool to see the individuals behind this, the same person who made that weapon also made the world's most safe fuel. I think it's a really beautiful story.
Cody Simms (29:36):
And what caused you then to move to LEU+?
Matt Loszak (29:39):
Supply chain.
Cody Simms (29:40):
Got it.
Matt Loszak (29:40):
With U-Zr hydride, there were a few challenges. One was supply chain, that was the big one, because we really wanted to be able to place the fuel order in the next few months, and start building this plant next year, and go critical the year after. There were some other interesting challenges that I think would've been solvable. For example, hydrogen dissociation and the U-Zr hydride would cause the fuel to be a worse moderator over time, which would cause the reactor to become less effective over time, or the fuel to be less effective over time, which would've been solvable. But, really the main thing is the supply chain. We're trying to set up this factory that can pump out a lot of these Aalo Pod products for these data centers in just a few years. And so, you need a fuel that can follow suit with that level of demand.
Cody Simms (30:27):
Awesome. And I know that, that was getting into the weeds for folks, but I think it's fascinating to hear all of the incredibly complex decisions that go into building a new nuclear reactor design, everything down from how can I source the fuel, what are the properties of the fuel going to be? What are the safety constraints of it? I'd love to pull us back up a layer maybe, and I've heard you describe Aalo as an XMR, which is maybe different than an SMR, which is the phrase many of us have heard, or small modular reactor. Explain how you think about the difference there, and is there more to this than just an unbranding exercise?
Matt Loszak (31:04):
It's definitely a little bit of a cheeky term. The last thing the nuclear industry needs is another acronym. I think there is some meaning behind it. The way that we think about the market is that the large-scale reactors, the PWRs, and those are great, they're safe, and so on, produce a lot of clean energy for humanity. But then, you've got all these startups doing these smaller reactors, and there's a whole swath of sizes and form factors. We tried to dissect that and understand where we fit in. You've got micro reactors, which are usually one to five megawatts. The LCOE is sometimes like 20 to 40 cents per kilowatt-hour, which is definitely expensive, but that's okay, because their target market is remote diesel. This comes from when you go that small, you have some challenges on energy density and the efficiency with neutron loss and things like that.
Cody Simms (31:55):
This would be like Radiant. We had Doug from Radiant on the show fairly recently would be an example. I think of that type of reactor.
Matt Loszak (32:00):
Yeah, yeah, so they're targeting military and remote diesel, which is awesome. The world needs that product. But then, that product might be more challenging to power a gigawatt data center, because having a thousand one-megawatt reactors would just be not the optimal solution, let's say for that scale. And then, you have SMRs, so small modular reactors. And this term came about a while ago when the whole intention was scaling down larger-scale plants for more of a utility model-
Cody Simms (32:29):
And hitting that mass manufacturability that you were hinting at, that's been the goal of this size reactor, I think, right?
Matt Loszak (32:35):
It's funny, the term is almost a misnomer because the goal really shouldn't be to make just a modular reactor. The goal should be to make a modular plant. And whether or not that's what the SMR vision was intending, that is what we think needs the distinction with the whole XMR term. And really, it's almost like it needs a whole new name that says it's just a modular plant. But, that's what we're trying to highlight with this XMR vision is that the whole plant is made in the factory, not just the reactor. This includes the IC modules, and a lot of the secondary system, the primary reactor, and just lots more, like the whole plant really.
Cody Simms (33:11):
Basically, the containment units, the pad that it sits on, all of the concrete that would need to get poured to create an SMR with an Aalo XMR module, you can assemble all of this, or the vast majority of it offsite.
Matt Loszak (33:25):
Yeah, which again, sounds obvious, but the strange thing is sometimes when you dig into some of the existing designs, they're not optimized in this dimension. Some end up being more concrete per megawatt than any other reactor design. That seems suboptimal. That's what we're doing, and what's more relevant or more important than this acronym is really exploring what the product is that we're designing for these data centers. We've had the benefit of having the past two years to really talk to a lot of these customers, and ask them what do they want? And if that was all you did, you'd end up with probably a pretty bad product, because you also have to do the first principles physics and see what's economical. You can't just wave magic wands and do a Enron egg or something. What we're doing is it's a five reactor, one turbine configuration called an Aalo Pod, and each reactor is 10 megawatts, so the total plant is 50.
(34:21):
And this was very intentional, so a lot of data centers they get built out in these 50 megawatt chunks, so 50, 100, 150, et cetera. The whole advantage of having five reactors is that when any one is down for refueling or maintenance, the others can make up the delta. This is also called maybe like an n plus one configuration where you pay a bit of a premium for the extra hardware to be acting as a backup. But, that's valuable to the customer, because then they have a high availability. And I have to tell you that the first plant is going to have some capacity factor challenges. That's a given.
(34:55):
But, the whole idea is, at some point, you could actually have these pods allow these data centers to not require a grid backup. And this is powerful being on site. What it would do is basically save time and money for these developers. Time because the interconnection cues can take five years sometimes these days, and money because people don't often realize T and D can be half the cost of delivered electricity. In fact, in the past 20 years, the overall average power price in the US has been somewhat flat, even though generators have been getting cheaper because T and D has been getting more expensive.
Cody Simms (35:31):
What does T and D stand for?
Matt Loszak (35:32):
Oh, transmission distribution.
Cody Simms (35:34):
Okay, there we go.
Matt Loszak (35:35):
Power wise, yeah. This world we're about to enter where there's going to be gigawatts and gigawatts of new industrial load, whether it's data centers, or even EVs, or other forms of industrial load, it makes so much sense to have a product that's well-sized for on-site power production. The on-site part is really valuable. And if you can pay premium for that high availability with a redundancy, I think you've got a really special product there.
Cody Simms (36:02):
Behind-the-meter power for data centers in a modular way, and as I understand it too, with many reactors, one of the ... Obviously, nuclear is a clean base base load power source. But, one of the challenges is sometimes they do have to go offline when you're doing refueling, for example. And as you mentioned, the lower the enrichment level, the more frequently you are refueling the reactors. And so, these pods can allow for refueling and maintenance to happen in a way that doesn't impact overall power delivery in a major way. Am I understanding that correctly too?
Matt Loszak (36:37):
That's right. For example, a traditional gigawatt scale water-based reactor might have to go down for a month every two years for refueling. And the way that I think about this is, it's like your Tesla or your EV has 100 small battery cells, not one big one. It makes sense. That's on-site power, and it's redundant, so same thing for this.
Cody Simms (36:58):
You talked a little bit about pathways for getting through all of the regulatory requirements on the DOE side. To do this with the data center in Texas or wherever, where the company is headquartered, you need to go through a full NRC process. What does that look like for you?
Matt Loszak (37:15):
We've already engaged the NRC. We've submitted our regulatory engagement plan, and then, we've also submitted a couple of topical reports. We're well underway there. The whole idea is that the DOE progress will help feed into the NRC in our case. They're parallel, but then slightly staggered. You're right to deploy these by the thousands around the country, especially on civilian soil. It will require NRC approval. There's been a lot of great progress at the NRC. The Danube Division is fantastic, and they're very efficient. Cairos has been doing a lot of good work on taking a research reactor approach with NRC 104C.
(37:51):
There's been a lot of promise in slowly moving through these steps methodically. I should really say, step-by-step ferociously. Not slowly, but it's doing the steps required, but doing so with vigor. Like I said earlier, the window is open now where sure, the NRC could be more efficient in different ways, yeah, but never before has there been so much support from the public, from investors, from government, from customers. Now is the time to do this. I think it's going to happen.
Cody Simms (38:18):
I have a philosophical question for you. Some would say, "Hey, more AI and more nuclear. Does the world need these things?" Is this an inevitability or an eventuality that we as a planet need? We're putting a lot of resource into two things that have unknown implications for, not just humanity, but for earth, for life. I'm curious how you think about that.
Matt Loszak (38:43):
Oh, my gosh, yeah. I think AI and nuclear are incredible. I have a bunch of thoughts on this. First of all, if you think about for humanity, what brings us more wealth and wellbeing, it's the ability to do more with less. And there's nothing that I can think of in all of technology, all of humanity that allows for that more than AI and nuclear. Nuclear allows us to do so much more with so much less. And to be more specific, it's like one barrel of uranium is equivalent to two million barrels of oil and gas. I'd much rather just mine one barrel than two million. That sounds a lot easier. And then, it's also just because of that energy density, you use a lot more space, a lot less raw resources, and it's clean, so we can do all these things and not poison ourselves in the air, or have breathing problems, or cause global warming.
(39:29):
And then, on the AI side, I've been interested in AI for quite a long time. Actually, one thing I've never really mentioned is I did a small AI startup that I sold for, I don't know, maybe less than six figures back in the day. One of the co-founders there was Nick Frost from Cohere, so one of the large language model companies. His partner, Aiden Gomez, was one of the original co-authors on the Transformers Are All You Need paper back in the day.
(39:54):
In Toronto, it was a really cool scene of AI with Jeffrey Hinton and a lot of people in there at the forefront. I think AI is just so incredible, the potential for humanity to do more with less. I've had deep conversations with friends about how if we lived a couple of hundred years ago, it would've been easier to contribute to the tree trunk of physics knowledge for humanity, for example. And now, to contribute in science, you have to be so far down some small branch that it feels like it takes so much longer to require the prerequisite knowledge to contribute.
Cody Simms (40:24):
Hard to be Galileo of 2025, I suppose.
Matt Loszak (40:27):
Yeah, yeah. Although maybe AI is this whole new tree trunk itself. Demis Hassabis always speaks about this too, when it's used to help us discover new science and accelerate progress through humanity on whether it's drug discovery, a lot of health related things, especially I think is fascinating because it's very hard to model biology from first principles physics. We don't have the compute, but already with alpha fold, we're seeing that actually AI is really well suited for going toe to toe with the complexity of biology, so to speak.
(40:57):
I think a lot of good for humanity will come from that, and then, name your field, like physics, there's going to be incredible developments in technologies for materials, and maybe it'll help make fusion possible and all sorts of things. I think there's massive opportunity here, and I think nuclear is inevitable, and I think it's just a question of how fast can humanity wrap its mind around it and get comfortable enough to allow it to help us enter the future.
Cody Simms (41:23):
Matt, that was an awesome way to share some of your final thoughts here on the interview. Anything else we should have talked about with respect to Aalo, with respect to the company you're building? Love to make sure to give you a chance to share anything more.
Matt Loszak (41:35):
Thanks again for having me on. Longtime Listener. It's really an honor to be here, and we're hiring aggressively. A year ago, we were like five or 10 people, now we're 45. In a year, we'll be 80 or 90. We're aiming to start construction on the first real plant in 12 months. That's our internal goal. Don't tell anyone. We're hiring aggressively to achieve that, and aiming to go critical the year after. And so, yeah, hiring, fundraising, if people are interested in investing, we are doing a series B right now. Otherwise, really appreciate the time in having us on.
Cody Simms (42:04):
Well, thanks for taking the time to share with us what you've been up to, and for being one of my go-to people I call when I'm looking to learn more about nuclear. I've always enjoyed our conversations, and it's awesome to see what you've been building.
Matt Loszak (42:16):
Yeah, you bet. Thanks again.
Cody Simms (42:18):
Inevitable is an MCJ podcast. At MCJ, we back founders driving the transition of energy, and industry, and solving the inevitable impacts of climate change. If you'd like to learn more about MCJ, visit us at mcj.vc, and subscribe to our weekly newsletter at newsletter.mcj.vc Thanks and see you next episode.
