The Nuclear Comeback: Can It Power America’s AI Future? | Seth Grae

David, it's so great to be on with you. Thank you.

You know, my graduate degrees are in law and in business, not in engineering or science. And I originally came into nuclear through the law. When I was in law school, I wrote a paper in the form of a brief to represent someone who was a nuclear engineer in the then Soviet Union and ended up actually taking on that case and helping him actually get released. And through that, got very interested in talking to people around that case and things about nuclear. And then as a lawyer I had worked on the international transfer of high technology that became only nuclear technology and I was very intrigued in the nuclear technology. The people in the industry, they were just remarkable, fascinating people with remarkable careers trying to do big things. There were some very interesting international aspects about the nuclear power industry to me. So, that's really how I originally got interested into the industry through being a lawyer.

That's very true and it's a very law and rules-based industry and understanding the legal framework, the regulation from the US regulation to the safety culture to the rules around transferring the technology to protecting that intellectual property and contracting is all very relevant.

I try to avoid just using the word nuclear for exactly the reasons you're saying. People might hear the word nuclear and think of the word that follows it as being nuclear war or nuclear weapons or nuclear medicine. And if we're talking about nuclear energy, I say nuclear energy or more specifically, I'll often say nuclear power. And I think of power as energy that is always available at the full amount. And that is one of the key benefits of it in a world with a lot of intermittent energy supplies such as by wind and solar that aren't available at the full amount if the wind isn't blowing or the sun isn't shining and might not be available at all. And for modern industries such as data centers with AI, they need tremendous amounts of energy that is constant, that's reliable, it's available all the time, and that also preferably is clean. And nuclear emits no CO2. It emits nothing harmful into the atmosphere. Emits nothing into the atmosphere at all. So nuclear is something that nuclear energy, nuclear power is something that I think is a very important part of the energy mix. If you're going to have a modern economy with a lot of always on clean power, you really can't get there without nuclear as part of the diversified energy mix.

Well, the US has generated about 20% of our total electricity usage by nuclear for about half a century. And that's not bad within a diversified energy mix that has included hydroelectric and natural gas and coal and sometimes some geothermal and some other generation as well. And I think that one of the reasons why the US did not increase the amount of use of nuclear energy as much as some countries did is an abundance of resources in some other areas such as the ability to have hydroelectric that some countries don't have. There was also, I think, an unfounded public fear of nuclear that in many ways is actually the safest industry in the country and nobody has ever died from radiation in the history of the US nuclear power industry. Just as you hear if you go on a flight, it's the safest part of your day in an industry that so prioritizes safety. Well, if you go on a flight to visit a nuclear reactor, you're really going to have a safe day. And it's everything in between that's statistically more dangerous because this is an industry that focuses on and delivers on such a high standard of safety. And I think that you see that in communities where there are reactors. So the reason for this fear of nuclear power really traces back to 1979 to Three Mile Island in Pennsylvania with fuel that melted down. And this was at the time the China Syndrome movie was in the movie theaters. And this led to abandoning of large numbers of nuclear projects, cancelling of large numbers of nuclear projects. Well, nobody died at Three Mile Island. Nobody got killed at Three Mile Island. And now as there's such a need for surging electricity in the region, the remaining plant at Three Mile Island is in the process of reopening with strong public support locally in the state of Pennsylvania in a bipartisan way led by the Democratic Governor Josh Shapiro and with tremendous national support led by the Republican President Donald Trump. And so of all the reactors to be reopening and with such popular support that it's Three Mile Island, I think really shows how people have come to understand that we need nuclear as a growing part of the energy mix and the fact that 100% of that power will be used for Microsoft data centers, you know, just shows where the new demand is coming from and what they're looking for to how they can find that much power.

Yeah, it's interesting because I saw today that the word hyperscaler isn't in the Microsoft online dictionary, but they're what a lot of companies think of when they think of a hyperscaler, right? And I think the term mostly is being used now for these enormous companies, be they in the tech industry from the Microsoft angle needing data centers and software to the chip companies like Nvidia. These $4 trillion market capitalization companies that are on the leading edge of many technology developments and usage in data centers for artificial intelligence. And you know what I've heard about Nvidia and some of these other enormous companies like Amazon, Microsoft, Google's parent, Alphabet and Meta Systems, which is Facebook's parent, is that perhaps the biggest limiting factor on their stock price going forward is their ability to get enough power for the data centers with AI. And in that context, it makes perfect sense to go and reopen Three Mile Island, to find the power where you can, you know, to build new reactors. They need the power to execute on their multi-trillion dollar business plans. So these data centers to put in context, New York City all five boroughs use about 8 gigawatts electric all of the time. Think of it every day New York City is using about 8 gigawatts. There's one data center that Microsoft is planning to deploy with OpenAI that will use over five gigawatts for one data center. So one supercomputer complex they call it a campus of data centers that's larger than the power needs of most major cities and it's, you know, 5/8 of New York City's power needs, the largest city in the country. So these data centers just need power at a level that has been unimagined before in our nation's history on how much incremental power you have to add to the grid for these data centers. And there's a notion that we're in almost an existential competition with China that if the American and Western companies don't in a sense win the AI race, you know, it would be very bad for the nation with AI eventually being in almost everything from weapons systems to controlling the electric grid to you ultimately in people's pacemakers and toasters and everything. But really in big picture AI is becoming very important for everything and the most highly valued companies in the history of the world are the ones leading this charge and that need the most new power.

Sure. So there's terms base load power and some people call it firm power. There's peaking power and people talk about peaker plants and you could think of say in a service area say a city the amount of power that is always needed no matter what at any time be it at night or a weekend or in the spring or fall when there's not as much electricity or air conditioning being used. Think of that as the base load that no matter what that amount of power is always needed and then everything above that you're getting to more midlevel generation and peaker plants that kick in as you go up. So for example in New York City there's even some oil burning plants that can kick in if the electric need surges on a hot day when there's a lot of air conditioning needed. Now, that's also called firm power because you need it to always be there. If it dips below that amount, the lights start going off, hospitals lose power. It's pretty bad. And so the base load in a sense is increasing, you know, that line is going up of how much power you need for base load when you're adding data centers when you have other industries like steel being produced with electric arc furnaces they're switching from fossil fuels to nuclear power. When you look at transportation like cars and buses and trucks that have been powered by fossil fuels in the form of gasoline or diesel that are switching to electric cars and electric vehicles plugging into the walls. And there's crypto mining where people are using, you know, electricity for that in large amounts with computers and this is adding to this amount of always on electricity that's needed. So for example in Loudoun County, Virginia, which has more data centers than anywhere in the world near Washington DC, there's just this immense amount of always on electricity needed and without it, the credit card payment systems wouldn't be working. People's photos from their phones would back up and all kinds of critical, you know, national security missions wouldn't be going on. So having power that can produce base load energy is very important. Nuclear is pretty much the best at doing that. And what you find when you build a reactor is it could last almost forever. You could keep refueling it, keep replacing the critical parts. There are reactors that were originally licensed for 40 years that have been relicensed to go to 60 years. They're being relicensed to go to 80 years. Some of them are already looking into going to 100 years. There's a, you know, a joke in some of these plants that the only thing that's actually 40 years old at the end of the original 40-year license is the license on the wall. They've replaced so much of the equipment along the way, but they really become these almost forever generators of reliable power at a very known cost. And when you have fossil fuel costs being very volatile and could go up and down in price so much versus it's expensive to build the nuclear plant, but then it's pretty cheap to fuel it for up to a hundred years. That's just usually the best investment you can make in power generation. So nuclear is just increasingly becoming looked to for base load power, for firm power as we need more of it.

Yeah, I agree and what I'm seeing on Capitol Hill is that this is probably the most bipartisan issue there is, literally support for nuclear power. It's just amazing how strong the support is from both political parties and as we've transitioned from administration to administration how strong the support is in the administrations. So, we've seen, you know, fantastic legislation supporting nuclear power. We saw in May of this year the president's executive orders and there is a bit of a great power competition there that this is the US and its allies versus China versus Russia and industry simply needs the power and without it people's electric bills will keep going up. There won't be as many jobs. There won't be as many good jobs. The modern economy needs jobs that need a lot of power. From, you know, data centers to steel production to all these new factories want to onshore in America, including auto production. It's just almost staggering how we talk to people at utilities and say the amount of power they need to produce five years from now is so much more than they were projecting 5 years ago for 10 years in the future. And now when they look at 10 years in the future, they really have no idea how they're going to get there from here and keep businesses, keep jobs, keep the air conditioning and the heat on. And, you know, that's why they're investing in new energy deployment, you know, including, you know, in particular so much interest in nuclear power.

First of all, on the cost, it's more expensive to build a nuclear power plant than to build a gas plant or a coal plant or most other kinds of power generating plants. But then the cost of operating the plant every day is much lower. The cost of the nuclear fuel over the life of the plant is much lower than the continuous cost of buying the natural gas or the coal and it is not nearly as volatile in price and overall over the life of the plant, nuclear turns out always to have been a good investment and, you know, in fact often the best investment and for states for countries that are looking at the long-term power needs think of it like building the interstate highway. It might seem like an expensive thing to build if you don't already have an interstate highway system but you'll be very glad when you have it when you look back a couple decades later. The benefits of it will tremendously outweigh what the upfront costs were and economically will have been a fantastic decision to make and nuclear waste is one of those areas where again, like safety, I think there are some misconceptions. So for all of the nuclear power used in one lifetime in the United States for everything attributable to one lifetime uses nuclear fuel rods put into reactors if it was all generated by nuclear that add up to the size of a can of soda. And when you take the fuel rods out of the reactor, they look exactly like when you put them in, there's no liquid waste. And so all of the waste in a sense from the lifetime if all electricity were, you know, from all the factories that made your clothes to everything attributable to nuclear be the size of a can of soda. And about 96% of that would be uranium, pretty much like what you could dig up out of the ground. So if you separate out the uranium, the other wastes in there are about the size of a pinball for all of the waste for one lifetime. And it's all contained in that solid form. So there's nothing in the ground, there's nothing in the atmosphere. It's very easy to safely handle. And in France, for example, they reprocess to reuse uranium and use plutonium that is produced too. That's about 1% of the waste. So it's only about 3% that they have to handle. It's called fission products and transuranics. And they mix that with other substances that they safely bind with into a glass log through a process called vitrification and bury it. And there's one room in France where under the floor is all of the nuclear waste in the history of the country. And it can't hurt anyone and it's not dangerous and it's not a weapons threat or anything like that. So, it's something that can be handled very well. And I think that here in the United States, we're handling it well where we take the fuel out of reactors and we put it in these pools on the reactor sites called spent fuel pools. And then after a few years, we put it into what are called dry casks, these steel and concrete casks outside on a concrete pad near the reactor where they're very safe. You could go up and have a picnic next to those casks and, you know, you won't get any radiation. In fact, you could go and stay there for a year and get less radiation than if you eat one banana once because of the potassium in the banana. And then eventually these might go to what's called interim storage where we'll have a centralized location where we put a lot of these casks, you know, in one or more places, you know, above ground in these concrete and steel casks. And then we'll either put them below ground in what's called a high level waste repository or reprocess and reuse most of that material in reactors to generate power again. With only these very teeny tiny amounts that we need to handle and, you know, and bury underground and even those materials for some of the newer kind of reactor technologies can be reused or what's called, you know, by the engineers transmuted down to lower level materials. So it's actually technically not a challenge. It's more of a political issue. And again, when you look at these local communities around the plants, they've realized that this is not a safety issue to the community. It's not a health issue. There's no problem at all. So again, of all places, Three Mile Island, Harrisburg, Pennsylvania isn't against reopening that reactor and over time putting more of these dry casks there on the site.

I think that the nuclear industry's practices and the regulations from the government, you know, have been honed over the decades but this is an area I think the industry has always handled well. You know, again, nobody has ever been hurt by this in the history of the United States, and it would be very hard to figure out how someone could be hurt by this given how it's handled and how it's handled so well. You could find online some videos of people taking these casks that they put the casks into and, you know, have the equivalent of fly a fighter plane into it and nothing happens to the cask or have two locomotives crash head on into each other that have these casks on them. Nothing happens to the casks. But even if they cracked open, it was very unlikely. But in an instance like that, again, this is solid material that you could put back into a cask. So it's really not this big danger that some people think that it is.

Well, before there was Lightbridge, there was a company called Thorium Power that was looking at using an element called thorium in making fuel for reactors. And there was interest in thorium around the world. It was driven for many reasons including that India was not able to access nuclear technologies and uranium on world markets but has a lot of thorium in the ground and was very driven to try to use thorium in reactors. And there was a notion that uranium was very limited. We might be running out of it eventually and we could use thorium in reactors. And the way thorium works is that you need to what's called seed it. It's what's called fertile. It's not what's called fissile that will produce power. And using what's called an accelerator or using neutrons from uranium in a reactor hitting the thorium, you can get it to become a form of uranium that produces power and that's something that we had been looking into and a lot of people ended up moving away from thorium including us partly because India entered into this nuclear deal with the US during the George W. Bush administration where India could access nuclear technologies and uranium on world markets and that's what India's been focusing on since. And because we found that the benefits we thought we could get from thorium relating to a more energetic fuel, one that could last longer in reactors, one that could provide tremendous safety benefits. There were ways we actually thought were better without using thorium but using uranium in a different form. And we also found that uranium kept being discovered more around the world. And there's just so much of it from Australia to Canada to Kazakhstan and, you know, now much more uranium mining in the United States and other countries that there's really not a sense that we're running out of uranium. And if we do go to reprocessing of this used fuel, that's a tremendous source of uranium there too. So the move away from thorium and toward a very different fuel to accomplish the same missions better was really an evolution that led to Lightbridge. And it was based on talking with some large nuclear utilities that were looking to increase the power output of new reactors and in particular increase the power output of their existing nuclear power plants while, you know, enhancing safety further. They're always focusing on safety. And we started working on ways to re-imagine, to redevelop the nuclear fuel for reactors from scratch in a very different way. Even the thorium was kind of similar to the old uranium fuel which is these ceramic pellets of uranium where a couple of hundred of those are stacked up inside a metal tube. And even thorium fuel was very similar to the uranium fuel that just hasn't really changed much in 60 years. Just very slight tweaks. But Lightbridge went in a very different direction of let's start from scratch. Let's not use pellets. Let's not use tubes. What could we do that can meet what utilities are looking for to get a lot more power out of their reactors with better economics and even enhanced safety? And that's what led us to move from ceramic or oxide uranium to uranium in a metallic form. From pellets stacked in a tube to an extruded rod with several zones in it that are

metallurgically bonded to each other in an extremely high-tech product where we invented the design of the fuel rod as well as the production method for it. And we've started basing this project at Idaho National Laboratory which is the designated United States lab for nuclear energy and nuclear fuels, you know, and benefit from tremendous knowledge and facilities there and support from the US government. And it's remarkable to think that it's only been four months since President Trump signed the nuclear executive orders in May of 2025. Lightbridge has been moving at a speed since then that's really unheard of in nuclear. So we're working at Idaho National Laboratory where since May all of the following has happened. Fasten your seat belts. We received physical highly enriched uranium from the US Department of Energy. Very few companies in the history of the world have received highly enriched uranium from a government. We working with Idaho National Laboratory produced nuclear fuel samples using that highly enriched uranium. Those fuel samples passed being qualified with Idaho National Laboratory.

We were approved for what's called fast testing, which is something Idaho National Laboratory invented. It's an acronym, fission accelerated steady-state testing. That means literally faster testing of nuclear fuel in a reactor. And having that highly enriched uranium means there's more neutrons. The fuel can be tested faster in the reactor. We've been approved for the experiment for going into the reactor. We're supposed to go into the spring outage in the first part of 2026 in the reactor, but it looks like that's being accelerated and we could possibly be in the reactor in November of 2025 instead. And these are all things that have happened in the last four months. And this is really industry and government working together quite quickly, quite safely, using advanced technologies, advanced methodologies to really do things faster than they've ever been done before in advancing new nuclear technology.

Yeah, it's also benefiting from the legislation that's been passed by Congress and signed into law by the president, benefiting from the executive orders which are binding on the executive branch of government. And I think in the end really pressure from these giant companies, these hyperscalers that need the power and are interested in nuclear power that's not, you know, specifically making them look at Lightbridge and the specific test we're doing at Idaho National Laboratory, but that is getting the Department of Energy and the National Labs and the companies like Lightbridge very motivated to work very well together. And in nuclear there's a saying that you should take a long time to plan and then execute quickly. So what you don't want to do is start building a reactor that you haven't finished designing. You want to spend a lot of time planning it and then build it quickly. And it's like that with what we're doing now. We've been preparing for this moment for a long time. We've been, you know, very detailed in getting to this point with our fuel development with brilliant people at the company and at Idaho National Lab who are very excited about this work working on this project and when this moment arose and the government was looking for ways to move more quickly, you know, we're ready and other companies are too. It's a remarkable moment in nuclear. So it's a convergence.