Nuclear's big return, British creators revolt & Elon sends an email
44:27 Share

Hi folks, it's Mark from the podcast Food with Mark Bittman. As the weather gets warmer, it's time for lighter meals and Whole Foods Market has just what you're looking for with great everyday prices. Look for the yellow low price signs that help you save money without compromising the quality you expect. Nothing, absolutely nothing in the store has any high fructose corn syrup, for example.

Just shop with confidence. Save on the best of spring with great everyday prices at Whole Foods Market. This episode of the Times Tech Podcast is sponsored by Vanta. Let's talk about something that might be keeping you up at night, cybersecurity. According to Vanta's latest State of Trust report, it's the number one concern for UK businesses, and that's where Vanta comes in.

Whether you're a startup, growing fast or already established, Fanta can help you get ISO 27001 certified and more without the headaches.

And Vanta allows your company to centralize security workflows, complete questionnaires up to five times faster and proactively manage vendor risk to help your team not only get compliant, but stay compliant. So stop stressing over cybersecurity and start focusing on growing your business in 2025. Check out Vanta and let them handle the tough stuff. Head to vanta.com forward slash the times tech to learn more.

Because when it comes to your business, it's not just about keeping the lights on. It's about keeping everything secure. Hello, this is the Times Tech Podcast with you, Danny Fortson. Danny in the Valley, Silicon Valley. And you, Katie Prescott in the City of London. So later in the show, Danny, we're going to be hearing a conversation you had about the future of clean energy. Which is what exactly? Nuclear, or as George Bush used to say, nuclear technology.

nuclear is back and it's all about AI. So, you know, just like everything else, it's all due to AI. So that's what we're going to talk about. And I think we're going to talk about that in the first stories that we're looking at today. Before we get onto that, one from your side of the pond and one from mine. Where do you want to start? Let's start over there. What's happening in jolly old London? Well, there's

There's been a bit of pushback against your part of the world, actually. Oh. And I found out a bit. The man running my local newsagent went over there in the morning and he was wondering what was going on because all of the newspapers had arrived that morning with exactly the same front page.

which is unusual. That is unusual. Given the political splits between British newspapers. Yes, indeed. This was the same. So they all had make it fair, and the AI and fair was highlighted.

Can you guess what it was about? I'm going to just take a wild stab here. Was it about AI as in, you know, the AI unfair? Yeah. And what we keep talking about, which is copyright and what happens between the creative industries, so the media,

and the tech companies when it comes to copyright. And so there was this big campaign pushing back against the tech companies using the creative industries copyright without permission. And in the Times, we also ran an open letter, which was signed by some really big names from the creative industries saying,

who are basically disagreeing with proposals that the UK government's put forward. And the public consultation into these proposals closes this week. So hence the big push from the creative industries over them. And they're saying that going forward, they think that AI companies, that tech companies should be able to use copyright from other people unless...

companies or individuals say that they don't want it to be used. So they have to proactively opt out. And the creative industries say, hang on a minute, that puts all the burden on us and not all tech companies. And we don't think that's fair AI. But the thing that I don't understand, what's in this for the UK and why are they doing this? Because if you think about what AI is doing, the AI developers are doing the platform companies.

Allegedly, they've already taken it all. There's already all of these lawsuits. So why are you saying post everything happening? Yeah, it's okay for you to have stolen this stuff. I just don't understand what's in it for them. And the other thing I would say is that

In the US, nobody's really talking about this. Well, you had the Hollywood writers strike, didn't you? That was partly about AI. Very much about it. But again, these are, there's like flashpoints here. And in the States, they're like, this is all going to be adjudicated in the courts. These companies, a few of them like OpenAI with our parent company, are signing licensing deals with

But mostly this is just going to end up in the courts like we, you know, we had the Getty boss on the other week talking about their case against Stability AI. So they're kind of letting that run its course. Certainly there's a difference of approach between Europe and the UK and what's going on in the US.

The UK government wants to try and get on with something, I think, before these court cases unravel, because that could take ages. What does that mean? Get on with what? With trying to find rules that going forward, because I see what you mean, the horse is bolted, historical stuff is already in the machines. But particularly if you're from a news organization and these models want contemporaneous information and they really do need it to be accurate, then it's like, okay, well, going forward...

we need to make sure we've got rules in place for how copyright's used. And I think there is some thought as well about then looking back into the black box as if this is possible of AI and seeing what is in there already. But the creative industries are very adamant, and they talk about this in a forward-looking way, that for the musicians who are coming down the track, for the articles that you might publish in the Sunday Times this weekend, they want to make sure that that use is fair.

So it's still all up in the air, but the proposal, the proposals close soon. Yeah. And then there'll be some guardrails. And actually, I'd be interested to see how what gets decided here in the UK does end up affecting you in the US. Yeah. And I think if you look at the people who signed that letter, you know, it's Ed Sheeran, it's Dua Lipa, it's lots of very big names. I've referenced it before, but it just reminds me so much of Napster.

Of just like, hey, you're stealing our stuff to create this new thing that's going to completely disrupt our industry and at least in the near and medium term, completely devalue these creative works.

And in the case of AI, creating something that could make it harder for people to continue to kind of make money going forward, continue to kind of do this amazing creative stuff and get paid for it, all that stuff. So it just feels so pliant. It feels so kind of like, okay, like no harm, no foul, but can we agree some rules going forward? It's like, no, play a bit of hardball, you know, extract your pound of flesh from what has been done.

and set up a new regime going forward. I wonder if Hollywood and Silicon Valley is going to blow up again anytime soon. A hundred percent. Because the writer's strike, writer's and actor's strike feels quite a way ago now, a long way ago. Those deals are only three years long.

Three years. And we're already almost a year removed. And AR is accelerating and accelerating and accelerating. And as soon as its use starts being accelerated in Hollywood, I think. Correct. So this is absolutely going to come back. So the other thing that happened this week, apart from all the newspapers having the same front cover,

was a group of British musicians released an album in protest of the proposed changes. And there are tracks on there from all sorts of people. Kate Bush, Damon Albarn of Blur, Billy Ocean, The Clash. So I thought we'd play you a little snippet. Let's hear it. Play a clip. Play a clip. Let's hear it. That was it. That's so funny. So they actually have quote unquote, quote unquote tracks. Oh, here's the new Kate Bush joint.

Press play. So I think we'll just leave that playing in the background for the rest of the episode. Exactly. Yeah, I just think this is all going to get sorted out. It's going to be super messy. Yeah, I just feel like giving away any leverage you have at the beginning of these negotiations feels strange to me. But, you know, I don't know. Maybe I'm missing something. And that's the British media. What's going on in the American media? Yeah, so a few things just today, just this morning.

It came out, you know, Jeff Bezos, rich guy, started Amazon. Amazon guy. Oh, yeah, yeah. You know that guy. He owns the Washington Post and he sent out this kind of directive or announcement to the Washington Post, people that work there, announcing some changes to their opinion pages, saying they're going to focus on personal liberties and economic freedom. Those are going to be like the two, quote unquote, pillars that they now center their work around in those opinion pages saying,

And, you know, that very much kind of echoes, for example, the Wall Street Journal opinion page, which is the most well-respected thought leader-y source on the right in America and has been for decades. They say like free markets, free people, I think is their tagline. So it's kind of marking a real shift at the Washington Post, the head of the opinion section,

stepped down immediately because Bezos said, given this new order, if the answer to do you still want to stay here is not quote unquote, hell yes, then the answer must be no. So the guys left. It's the vibe shift. The vibe shift continues of these very wealthy, often techie type folks kind of throwing their weight around and being like, okay, we're going to kind of get more lockstep with this new administration. Fascinating to watch. Yes. Um,

I'd love to hear your views on one last news thing before we move on to nuclear. And where it's not open letters, strongly worded emails that Elon Musk, President Trump, brologarchy is continuing. Yes. He's been emailing his new, well, it's not fair to call them his new employees, is it? No. Well, so just a little bit of context. On Saturday, Trump posted on Truth Social, his social network that very few people use, but is his personal bullhorn.

He said, Elon is doing this all caps. Of course, we don't we don't use lowercase all caps shouting. Yes, Elon. Elon is doing a great job, but I would like to see him get more aggressive. Remember, we have a country to save, but ultimately to make greater than ever before. MAGA exclamation point. Yeah, exactly. Sound the klaxon. And Elon swiftly responded. Will do, Mr. President. And he did.

by very quickly sending an email to at least many hundreds of thousands, if not millions of federal employees. And it was exactly the same as what he had sent to Twitter early on, which was basically justify your existence. Make a bullet point email of the five things you got done this week. Absolutely. At that moment, just before he ended up cutting half of the employees. Yeah, he ended up cutting 80% of the workers and contractors at Twitter.

And in that email, he sent up a follow-up kind of post on X. He said, failure to respond will be taken as a resignation. And I can just tell you, I know a few people who work for the federal government in different guises. It's so disruptive.

To people who have mortgages, kids who have dedicated their lives to working in the Department of Justice, Civil Rights Division or Housing and Urban Development, people who have like important jobs doing important things, serving people. All of a sudden being told like maybe you're going to get fired at any moment. You're maybe going to have to fire people on a whim, even though you disagree with it.

And the overarching sense, you know, the kind of the source of these emails is a sense from Trump and Musk that is really anti-government and anti the people who work in government. Like you're a bunch of losers. Go do something else. And it's like when you talk to people in the government, they're like, you know, they want to keep doing what they're doing. But also you have now this sort of Damocles hanging over you and this sense that effectively you're a boss.

doesn't think what you're doing is worthwhile. That's what I wanted to ask you. Is he really the boss? I mean, isn't this actually incredibly confusing if you're one of these employees? Because some were told not to reply. So you're kind of like, who's the master here? Does Elon really have the power to do this? Well, this is the thing. It's like you're already starting to get pushback from other parts of the Republican establishment, including the new head of the FBI, Kash Patel, who's like a total Trump loyalist. He was like...

Y'all don't need to respond to this. Like, don't worry about it. And the Office of Personnel Management sent out an email basically being like, you don't have to respond to this. This is by, you know, you can do so if you want to, but you're not going to get fired. And then you have Republican congressmen who were last week in their districts all over the states before returning to Washington saying,

getting really hammered by their constituents being like, what is happening here? Like, we feel like we're being caught up in this chaos machine. And so you're starting to get a lot of pushback around what Musk is doing. And oh, by the way, a bunch of people they've fired, they've had to immediately hire back like days later. That's like a nuclear safety part of the government. They hired almost all of them back five days after firing them all in mass. Because they're like, hmm,

What they do is actually kind of important. Should we look at something a bit more solid? Yes. Yeah, so obviously everybody, including us, is talking about AI everywhere, all the time. Just last week, Elon Musk's ex-AI released their latest seemingly impressive model, Grok 3. You've had Anthropic.

Release a new model this week that is getting lots of kind of rave reviews. They look set to be close to closing a new $3.5 billion funding round, which would value the company at over $60 billion. So that means I think our friend Dario and a lot of his top team are now, assuming that happens, would be paper billionaires.

Apple committed to investing $500 billion in the US, which includes a new 250,000 square foot AI server manufacturing facility in Houston. It goes on and on and on. There were a lot of billions in that introduction. Billions. AI, AI, dollars, dollars, dollars. Exactly. Data centers, data centers, data centers. You need to kind of...

the infrastructure is data centers. This is the core of all of it. And of course, to run all those data centers, you need energy, energy, energy. And of course, clean energy would be best bet for varying reasons. You know, coal no longer in vogue. So what's the answer? So as we have kind of mentioned here and there before, nuclear has come back in vogue.

because it is seen as if you can make it work, reliable, clean, what they call baseload power. It's not renewables that depend on the sun shining or the wind blowing. It can just be on, always on, and not, you know, fill the environment with toxins. Uptime, as they like to say in data centers. Indeed. Uptime is everything. And fashion is, of course, circular. Nuclear is back. And so I wanted to have someone on who is kind of right at the forefront of this. It's a company called

called Aalo Atomics, A-A-L-O, Atomics. And I had on their founder, CEO, a guy called Matt Lozak. They're based in Austin, Texas, and they're having a go. They want to build, they want to kind of lead this generation of new, smaller, modular nuclear power plants that we're going to so desperately need. ♪

So you're Canadian, you're now in Texas. How and why did you end up going from your home country to Texas to try this really hard thing that hasn't been done in decades? Well, I grew up in Ontario and I used to actually have breathing problems like asthma. And we used to have 62 smog days a year. And that was because we had a lot of coal plants.

And when we turned off all our coal plants in the early 2000s, we went from 62 to zero smog days per year. And my breathing problems went away.

And actually, my sister, the same thing happened. And she moved to England, to London, and her breathing problems came back. So again, this air pollution thing is real. And I'm surprised it's not spoken about more. But anyway, that was my introduction to nuclear. And basically, I grew up living in the grid of the future without really fully appreciating it. Because Ontario replaced coal with nuclear. Exactly. So Ontario is 60% to 70% nuclear powered. Oh.

Oh, wow. And they turned off all the coal plants in those early 2000s. And so I did engineering and physics in university, 10 years of software startups, had a good outcome, built a meaningful HR and payroll business that processed billions of dollars of payroll. And that was great, but always wanted to take those learnings from software and business and bring it back to science and engineering where I came from. Yeah.

And I actually left a little bit early from that company, six years in after building it up to 150 employees and left a few million dollars on the table and some equity that I gave up to move on to something that I felt more meaningful, which is Allo. So if you could give a brief description of what Allo is and what you're trying to do. Yeah, what we're doing is we're building factory mass manufactured, smaller scale, but not too small, nuclear reactors.

The product that we're really trying to build here is something that's purpose-built for AI data centers. The product, for example, is not the reactor. The product is something we're calling the Allopod, which is five reactors, one turbine in a footprint that is much smaller than the equivalent energy-consuming portion of a data center. So it could be co-located there with a data center.

It's an advanced reactor, so it has these kind of good levels of inherent safety that allows you to co-locate with a pretty tight emergency planning zone so that it's just very easy to just co-locate. There's a bunch of interesting advantages of this approach. For example, when you have five reactors, one turbine as the product, it achieves a very high level of uptime. So these data centers, they need power for the most part around the clock or at least something that can be deployed remotely.

whenever the power is needed. And so it's a beautiful pairing for that reason, because if a reactor is down for refueling or maintenance, you have a bunch of others to make up the difference. And also with that level of availability, you can essentially, maybe not for the first project, but very soon thereafter, once capacity factor is established, do something that essentially allows you to bypass the grid.

So right now there's major delays with grid interconnects. Also transmission distribution is often half the cost of delivered electricity. So if you have something on site that does not require that grid backup, then it's a pretty unique value prop that we think is just gonna be insanely compelling to these data center companies. And we're already seeing quite a bit of early traction towards this for these first few deployments.

So that's essentially in a nutshell how we're getting started. When was the last nuclear reactor built?

and kind of plugged in in America. Vogel is the last kind of gigawatt scale reactor that got turned on a couple of years ago, kind of infamously went over schedule, over budget, in large part because in the West, we struggle so much with large scale infrastructure projects right now. And that Vogel plant, that's like,

the giganto power plant we all think of when we think of a power plant, right? Yeah, exactly, exactly. And that's part of the problem that we see is, one, the factory mass manufacturing can help bring in a lot more certainty around cost and schedule. Because the funny thing is, a lot of people don't realize that

There are eight utilities that already have approved licenses to build AP1000s, which are these gigawatt scale plants. But they're choosing not to because of a fear of cost and schedule overruns. So the whole idea is that factory mass manufactured brings you that predictability around cost and schedule.

But the second thing I would just highlight there is, you know, we don't think that these gigawatt scale plants are the be all and end all ultimate solution for data centers, because every two years you have to take it down for a month for refueling. To put in new uranium, basically, and swap out the old stuff. Exactly. Exactly.

So if you're able to do this, what we call N plus one configuration with an all-o pod, for example, with five reactors, one turbine, this would allow you to basically achieve that full uptime because during refueling our maintenance, you've got the other reactors to make up the delta. So we had the Vogel one, but it feels like nuclear has effectively been dead for 50 years or since Three Mile Island. Is that a fairly accurate description of what's happening here? Yeah.

I'd say growth has been stalled. I don't know about, you know, calling it dead might be an overstatement. What was it? The reports of my death have been greatly exaggerated or something. So, I mean, you know, nuclear still powers 20% of the US with clean baseload energy.

It's still got a pretty big workforce, like 100,000 people in the US. But there's been pretty incredible growth projections, I think, in the past five years, especially the public, policymakers, lots of customers, frankly, are all waking up to the

You know, the realization of how incredible nuclear really is. I think in the light of that, we're seeing a huge amount of growth. We're going to have to triple the workforce from 100,000 to 300,000 people to pull off the growth projections that are seen here. And those growth projections that surge in demand, is that really about growth?

Data centers?

Like Amazon, when they got started, had to choose a wedge market and they chose books, right? And then eventually they expanded into other markets. And for what we're doing, the data center angle is really compelling because it's a huge amount of demand.

They want clean baseload. And our solution is really, really optimal for them. Like there's not a lot of other great solutions. In all our conversations, it's basically like, you know, yes, if you guys pull this off, this is exactly what we need and we'll buy as much of it as possible. The big question is always just for a deep tech company is bridging that, you know, chasm of death for the first deployment to prove what you think is possible. This episode of the Times Tech Podcast is sponsored by Vanta.

Let's talk about something that might be keeping you up at night. Cybersecurity. According to Vanta's latest State of Trust report, it's the number one concern for UK businesses. And that's where Vanta comes in. Whether you're a startup, growing fast or already established, Vanta can help you get ISO 27001 certified and more without the headaches.

And Vanta allows your company to centralize security workflows, complete questionnaires up to five times faster and proactively manage vendor risk to help your team not only get compliant, but stay compliant. So stop stressing over cybersecurity and start focusing on growing your business in 2025. Check out Vanta and let them handle the tough stuff. Head to vanta.com forward slash the times tech to learn more.

Because when it comes to your business, it's not just about keeping the lights on. It's about keeping everything secure. Hey, you know what would make your customer service help desk way better? Dumping it and then switching to Intercom. But you're not quite ready to make that change. We get it. That's why Finn, the world's leading AI customer service agent, is now available on every help desk.

Fin can instantly resolve up to 80% of your tickets, which makes your customers happier and gets you off the customer service rep hiring treadmill. Fin by Intercom, the leading customer service AI agent now available on every help desk. It took a lifetime to find the person you want to marry.

Finding the perfect engagement ring is a lot easier. At BlueNile.com, you can find or design the ring you've always dreamed of with help from Blue Nile's jewelry experts who are on hand 24-7 to answer questions and the ease and convenience of shopping online. For a limited time, get $50 off your purchase of $500 or more with code LISTEN at BlueNile.com. That's $50 off with code LISTEN at BlueNile.com.

I've been hearing about small modular reactors for 15 years. In my previous life, I covered energy when I was back living in the UK and whatnot. And there was like a lot of excitement and optimism around this idea of smaller, quicker to build, theoretically easier to get approved nuclear reactors.

Why is now different or why are you guys going to be successful when this is just, it feels like just kind of been stuck in the mud for years and years and years? It's interesting. There's two layers of an answer I would give. One is the short version I can start with. And if you want to go deeper, we can look at what happened in the 1950s to 1980s. Because there were smaller reactors that were built back then. And there's an interesting history there as well.

But I think what you're referring to is the fact that, okay, that was the first atomic age in the 70s and 80s and 90s. It slowed down. And then your question to the point is, in the 2000s, there was the thought that there would be a resurgence, and it did not happen. So there's a few factors there. One is there was the collapse of the economy in 2008. That slowed things down.

Two is there was Fukushima 2011 that made things difficult for nuclear. It stopped it in Japan. Globally, it slowed it down. But really, that wasn't why it stopped. That was more like the final nail in the coffin, which the coffin was already getting set up at the time. So the key thing was natural gas was expanding during that time.

during those years. - Of course, fracking, fracking dropped the price of natural gas. - Yeah, so natural gas was not a major part of the grid in 2000. And by 2016, it surpassed coal. Why will this time be different? I think like in the long term, I think it's inevitable, like nuclear is inevitable.

It's insanely energy dense. It's clean. It's baseload. It's just a question of when. And so right now, it feels like all the pieces are in place for this to happen. You've got

arguably more public awareness, probably driven by social media and things like that, where a lot of people are very pro-nuclear these days. And it's got bipartisan support politically. So Trump is very pro-nuclear, which is great for the next four years, as was Biden, frankly. The demand is there like it wasn't before. So again, with data centers, it does present a very unique opportunity. So one is the pace of development, the scale of development,

But two is the willingness to pay a premium. So you saw some of these big tech companies, for example, Microsoft signing a PPA with turning Three Mile Island back on, which is crazy. That was a moment for a lot of people were like, because Three Mile Island is such this totemic thing in people's minds, especially of a certain generation. And Microsoft of all companies be like, no, no.

We want to pay the money to turn part of this back on because that's how desperate we are for clean baseload power to power our data centers. Yeah, yeah. And not only the poetic side of it, right? It's also the dollars and cents. So they were willing to pay 10 to 15 cents a kilowatt hour, which is a very high willingness to pay.

So if you think about what does it take to pull off the small nuclear reactor factory vision, it wasn't really possible in the first atomic age because they were going larger. They were largely state-driven projects. There wasn't really private industry to drive the demand for a small reactor gigafactory.

But now there is. So this is really a unique window because the demand is there. It's from the right types of people. It's at the right speed level, at the right price point, with the right level of political and public support. Is there an analogy to be drawn here between what you guys are trying to do and generally the kind of this private push for new modular reactors with like what's happened in space?

Like going from the NASA era state driven to a bunch of private companies taking a completely different approach to.

There's a push and then there's a pull of like, well, once you get kind of launch costs down, then you can do a whole bunch of other stuff. And so there's like also a pull from like private industry as well of just like kind of a rethinking of that whole approach. Yeah, there's a lot of similarities there. There are also some differences, you know. So similarities wise is like for space, there was kind of this one big obvious way to reduce costs, which is don't throw away the rocket.

It's like, if you can make the rocket reusable, right? That's going to make a huge difference. In nuclear, it's not like there's one single innovation like that. I think the factory manufacturing will really help. And I wrote, for example, a Twitter thread on this and talked about, you know, just how much it can help. It could really bring costs down. But, you know, the reality is like,

For example, paid off nuclear plants today are already able to produce electricity at like two to three cents a kilowatt hour price point. And in the 1970s, nuclear was cheaper than coal. It was sub five cents a kilowatt hour in today's dollars. So can you just give like a one minute description of how your reactor works? Because we have a very tech savvy audience, but we might not have

an energy savvy audience and people kind of, I think in the vaguest of terms, understand what nuclear power is, but most people don't.

I understand actually how it works. Yeah, for sure. So it's pretty magical. Basically, you take nuclear fuel, which is the most often uranium, and you put that in the core. And when you have a sufficiently high amount of it, it basically is like just taking two rocks and putting it together and it produces heat.

So it's very different than oil and gas where you're taking a fuel and you're burning it and you're using it up. In this case, you're taking essentially usually it's metal or ceramic. You're putting it all together and it just magically produces heat. It doesn't look like it's changing, but it's just producing heat.

And when it's spent, people often think of nuclear waste as like green goo. It's not. It looks the exact same when it's done as when you put it in with a few tiny changes maybe and swelling or something. So anyway, you're producing heat and then you flow a coolant around this, these heat producing metals and the coolant takes the heat away. And basically with that, you turn a turbine, you create steam, turn a turbine, produce electricity.

What people don't realize is if you see a cooling tower and you see what looks like pollution coming from that, that's not pollution. That is just steam. It's water vapor. So that's the super high level summary. You said you started two years ago.

I feel like most episodes of the podcast these days are about AI in some form or fashion. But two years ago, it was like ChatGPT had just come out. How much has the conversation changed since you started two years ago, or at least the kind of the temperature of the market when you were like starting out? Be like, no, no, we're going to build this first new advanced reactor in America in decades. You know, I'm just curious how that's all changed. Yeah, I mean, I would just say...

It's been constantly heating up. And this is a big thing about what makes all this possible is like, you know, it takes a village to build a nuclear reactor.

You need, yes, the right technology, the right approach to building a reactor, but you also need financial support. You need customer support. You need government support. You need to be able to recruit the top people from other industries who can bring in their insights and get attracted to say, "Hey, how do you take the learnings from the rocketry or other deep tech industries in the past five, 10 years and bring that to nuclear?"

if there wasn't really an opportunity to actually build something, the best people wouldn't be interested. And it's kind of this snowball effect where when the door opens and there's a bit of a crack and you see the light, then it just kind of snowballs. And that's what we've been seeing in the past few years is...

When we started the company, we saw this opportunity and it's just been getting more and more obvious. And now it feels like it should definitely be possible to build this first reactor. And as long as the economics are what they appear to be, and as long as nothing else crazy external happens in the meantime, I think we'll see this atomic age actually go through to fruition. And I think it's going to be here to stay because...

The last one was largely shrouded in a cloud of confusion and public misperception. And I think that will get corrected this time around. How many other, do you know off the top of your head, how many other kind of

SMR, small modular reactor companies, there are, it feels like there's all of a sudden there's a lot of them. There's a good number. So the way that we think about this is you've kind of got three broad categories, right? And this excludes the kind of, you know, dinosaur gigawatt scale water-based plants.

So you've got the micro reactors, which are one to five megawatts. They're really good for remote diesel or remote military applications. And we should say a megawatt for the folks at home.

That would power how many houses? That'd be about a thousand houses. So usually it's a kilowatt per house. So, yeah. So a megawatt is a thousand homes. So five megawatts is 5,000 homes, for example. Or a very energy intensive military site or a data center of a certain size. And so essentially the price point for these micro reactors is like 20 to 40 cents a kilowatt hour.

Quite expensive. Yeah, yeah. And they often use Triso or HALU fuel, which is quite expensive, but that's okay because their target market is these remote diesel locations. But you're never going to see a data center powered by a thousand micro reactors. Yeah. Right. And then you have the SMRs, which were kind of meant to be this thing that could be good for data centers. But the way the designs have evolved is kind of,

not ideal. And for example, some of the designs, you know, the reactor might be modular, but the plant is not. And so, for example, some of these designs, you know, it's more concrete per megawatt than any other reactor design.

And so it ends up being this like disproportionately huge civil structure that is not all made in the factory. So what we're trying to do is say, hey, maybe there's a third category here of what SMRs are supposed to be. Maybe we'll call it an XMR, like an extra modular reactor. But it's like not just the reactor is modular. The whole plant is modular. The whole plant is actually made in the factory and then shipped on site and assembled like Legos.

And if you look at the, you know, further, like what is purpose built for data centers, it looks kind of like that pod that I described where there's redundancy, the five reactors, one turbine, et cetera. So that's how we kind of narrow it down from maybe, you know, dozens or even a hundred.

people in the space to, hey, what are we doing differently? What is our target market that is not shared by others? That's kind of how we slice and dice the market. And how big are your reactors? What we're doing, like the reactor fits on the back of a truck. And then there'd be another 20 or so modular modules shipped on trucks to site. That roughly like, you know, per reactor, it's like the size of a two bedroom apartment for the actual plant portion. Gotcha.

And the idea is you're going to, I'll go all being well, you create a gigafactory where you basically have different assembly lines pumping out different parts, different Legos, basically that you can then all ship to the site and kind of,

build as big or as small as you want to a degree yeah exactly and the whole idea is like you know another nice analog to the data center industry is they have a very kind of fractal modular approach too right you've got the gpus and then they're in racks and they're in rows and then they're in pods and bays and so on and that is just repeated in a parametrically scalable fashion

And it's the same thing with this. So you'd have, you know, one pod is the 50 megawatt block because each of our reactors initially are 10 megawatts electric. So 50 megawatt blocks. And this lines up nicely with the average building block of a data center, which is also 50 megawatts.

And they tend to build these out in those types of chunks in terms of 50 megawatts of demand. I imagine when you started, again, going back to the chat GPT moment and everything leading up to that moment, I feel like a lot of – you talk to a nuclear company startup in that world. The argument was climate. It was like we need zero emission energy.

baseload power. And now it's like the data center kind of narrative has overtaken that to a degree. And especially with the new administration, they're kind of like not huge fans of the whole climate change agenda. How does that change the kind of environment for you or does it at all? Does it matter? Because it does feel like that driving, that driving kind of logic behind this whole new kind of atomic renaissance is,

Not that long ago, it was climate change. And now it feels like it's kind of almost shifted quite dramatically. In the short term, it's nice. And we're thankful that nuclear has bipartisan support. So regardless of why, you know, the Trump administration is pro-nuclear. But we also wanted to use a Canadian analogy, skate to where the puck is going. To me, it's

Where this is going is just, we need clean air. And this is kind of maybe a bit of a narrative that doesn't come up often. And I might start to try and push this harder. It's like, you know, climate change is one thing. And I think that that still matters. But to me, like the crazier thing is like,

is air pollution in terms of just breathing air quality. And this is actually a much bigger thing in China. But it still impacts here. It's like when I drive around on the highway and I see a truck putting out this black stuff into the air, I'm like, why are we okay with that? And then people go in a huff and puff about nuclear waste, which has never harmed anyone. It's like there's a real disconnect here between things like air quality, which actually does lead to deaths. But it's just so indirect.

And it's kind of like maybe shortens people's lives by, you know, some amount of time that doesn't feel as visceral as maybe some kind of weird mental link people have between nuclear energy and nuclear weapons, which are totally different things anyway. So just like, I think, you know, the bottom line is all

all these big tech companies that are going to have all this demand for data centers, they do still really care, obviously, about ESG and clean energy. Whether it's altruistic or for marketing purposes, it's irrelevant. They want this. And so that's why we have this window of opportunity right now where ESG

A lot of the developers for data centers are looking for ways to kind of feather off of gas. Yes, build gas now, but feather off of it onto nuclear as soon as it's ready. Will that be before 2030? Will it be in the early 2030s? There's different thoughts and schools of thought on that. And we're obviously working to make it work as fast as possible. But the desire is there and that's what matters.

That's really, really interesting. I find this idea that the rise in AI has created a window of opportunity for

for new energy sources is absolutely fascinating. Like you, I've been covering small modular reactors, mini nukes or whatever we call them. Such a long time. Rolls-Royce developed them here, ones that look a little bit like, you've probably seen them, like woodlice, silver woodlice, whatever. But you always wondered what they would be used for. And actually data centers feels like a perfect marriage. It is interesting because also it's like,

And I think this is kind of a broader point, but it's like, you know, when everybody's freaking out about AI and what it's going to do, which is, I think, totally valid in terms of jobs and society and everything. But it's also going to create all these kind of second order effects, like all of a sudden, if, oh, if we could crack economical factory made technology.

nuclear reactors that replace coal and gas, that's probably a good thing societally. And the same thing was like, you know, everybody was wondering, I felt like crypto for 15 years or since 2008, since the white paper has been a solution looking for a problem.

And now you have AI come out and you're like, everybody's talking about, well, pretty soon we're going to have more agents than humans on the internet. And they're all going to be doing stuff with each other. And maybe that's going to be the thing that crypto finally finds its use of like these micro payments between AI agents or whatever. But you just start to see like the little kind of green shoots of like...

How one thing over here is going to affect all these other kind of industries in different ways that we can't, wouldn't have been conceived of. It's very relevant for the UK and Europe where there's massive concern about the number of data centres which are needed to power AI and whether the energy infrastructure can currently cope.

and what the impact on the national grid will be with all of this new pressure. And so people are casting around for other ideas about how to keep the data centres up, keep the uptime. Because when they collapse, it's such a disaster. I think we spoke about this recently. The UK's finally made them part of critical national infrastructure on a par with...

Power stations. Water. You know, water. Exactly, all the utilities. Because if you're running a hospital and suddenly your digital infrastructure goes down, it's an enormous issue. Well, especially if you're, you know, there's all these like little AI...

that more and more doctors are using. There's a whole array of things where you're like, oh, this is actually just going to be kind of fundamental to a whole bunch of new services if the government gets its way. How much money have they raised so far? How are they doing financially? So they've raised $30 million so far in a Series A. I think they had a $6 million seed around then $30 million Series A. And I think their goal is to raise another $100 million this year. Yeah.

And they're trying to get this, you know, build this first prototype in the US to show that, you know, basically to put their money where their mouth is and be like, okay, we're actually, we can do this. Here it is. You know, from your time covering small modular reactors, I'm sure you know, like,

Just getting this through the regulatory gauntlet is a thing. But their team, they have some quite experienced people who have got prototypes through the process and in a smaller scale with like within government. So I think they have a pretty good shot. You never quite know. But as he said, you know, there's dozens of these small modular companies now. Right.

Danny, last week you were scoffing at the idea of leaving your kids in the hands of a humanoid robot. I completely maintain that view. I completely...

I have not wavered. I sometimes think I trust robots more than humans. But there's more about tech to worry people who have kids than robots. And social media, well, certainly that's the one that worries me. And I know we've talked about this a lot before. If you are worried about your kids online, The Times and The Sunday Times have launched a new parenting hub, which subscribers can access by going to thetimes.com forward slash parenting.

There's all sorts on there, as well as helping your kids navigate the online world. There are school and university guides, advice on mental health and finances, which got me thinking, Danny, when it comes to parenting tech, what would you like to see invented? Well, depending on who you talk to, they say it's already here, but I don't think it's quite ready because I've tried a couple of them. But just like a properly good AI tutor.

for like math and reading and like whatever it may be for for the kids I think that would be just really amazing what about you yes no I agree because I think the worry is when you leave them with a tablet or with a computer a I worry about what they're accessing frankly I don't completely trust the parental controls but b it all feels very passive

And you never really know how much they're engaging from it, even if they are doing a learning program. Anyway, until that properly exists, do go and visit thetimes.com forward slash parenting. And if you have any pioneering parenting tech ideas, you know where to send them to techpodatthetimes.co.uk. That is techpodatthetimes.co.uk. That's right. Or questions and thoughts related to

Any other things we speak about on the pod, data centers, whatever it is. Humanoid robots, parenting tech, whatever it may be. The best place to record a podcast when you're on holiday. You know, there's a car, there's a boot, there's a closet. A duvet over your head. I know, I know. I didn't do that. Maybe I should have. This episode of the Times Tech Podcast is sponsored by Vanta.

Let's talk about something that might be keeping you up at night. Cybersecurity. According to Vanta's latest State of Trust report, it's the number one concern for UK businesses. And that's where Vanta comes in. Whether you're a startup, growing fast or already established, Vanta can help you get ISO 27001 certified and more without the headaches.

And Vanta allows your company to centralize security workflows, complete questionnaires up to five times faster and proactively manage vendor risk to help your team not only get compliant, but stay compliant. So stop stressing over cybersecurity and start focusing on growing your business in 2025. Check out Vanta and let them handle the tough stuff. Head to vanta.com forward slash the times tech to learn more.

Because when it comes to your business, it's not just about keeping the lights on. It's about keeping everything secure. Millions of people have lost weight with personalized plans from Noom. Like Evan, who can't stand salads and still lost 50 pounds. Salads generally for most people are the easy button, right? For me, that wasn't an option. I never really was a salad guy. That's just not who I am. But Noom worked for me. Get your personalized plan today at Noom.com.

Real Noom user compensated to provide their story. In four weeks, the typical Noom user can expect to lose one to two pounds per week. Individual results may vary.