VideoToBe
Lang: en

Fusion Energy Will Power the AI Boom

Fusion may still sound like science fiction—but it might not be for much longer. With AI pushing demand for clean power to new highs, a breakthrough may finally be close. In this episode of YC Decoded, General Partner Gustaf Alströmer explains the history of fusion, the physics behind it, and the engineering challenges that stalled it for nearly a century. He also looks at how Helion Energy is approaching the problem differently, as they develop a new fusion system expected to deliver power to Microsoft by 2028. Chapters (Powered by https://ChapterMe.co) - 00:00 - Intro 01:30 - Fusion 03:50 - Helion 06:50 - Hardware focused

Video thumbnail

Convert your Audio Video Files to Text

Speaker Separated Transcription98% Accurate
Free for files under 30 minutes
0:00|Gustaf Alströmer:
These massive data centers are the beating heart of the AI boom that's reshaping our world, and they're hungry for more electricity. That's one of the reasons that for the first time in a decade, demand for energy in America is spiking. To keep up, we're burning more and more fossil fuel. And fast. But there's another path. A clean, safe, virtually limitless source of power.
0:20|David kirtley:
Fusion energy. Fusion is the way the sun works, only here on Earth we've actually struggled to be able to do it in a commercializable, cost-effective way. We've built now a number of machines showing that we can do that. So what would it take for fusion to power the future? Let's find out.
0:39|Gustaf Alströmer:
For nearly a century, scientists have chased the promise of nuclear fusion. In the 1930s, physicists realized that fusing light atoms could release immense energy. The first experimental reactors came a few decades later in the 1950s. So how have we not still conquered fusion almost a hundred years later? To understand why, we first have to understand the science. Nuclear fusion is the process of smashing together two light atomic nuclei.
1:04|David Kirtley:
You actually can take those two atoms, the core of those, the nuclei of those atoms and push them together, overcome their internal atomic forces. And what happens when you do that? They actually fuse and form a heavier element.
1:17|Gustaf Alströmer:
Unlike nuclear fission, which splits heavy atoms, fusion joins small ones and can yield far more energy per gram of fuel.
1:25|David Kirtley:
That new element has a tremendous amount of energy. That helium is born with on the order of 10,000 times the energy that it started with.
1:34|Gustaf Alströmer:
That energy comes from the tiny difference in mass between the reactants and the products, converted to kinetic energy. It's only about 0.1% of the fuel's mass being converted into pure energy, but that's millions of times more efficient than fossil fuels. To make fusion happen, you must first turn fuel into plasma, where electrons roam free and nuclei move fast enough to occasionally collide and fuse. You then heat the plasma to roughly 100 million degrees Celsius so the charged nuclei can overcome their electrostatic repulsion and fuse together. But how do scientists confine the plasma long enough for fusion reactions to occur?
2:12|Gustaf Alströmer:
There's a few ways to do so. Magnetic confinement uses powerful magnetic fields in reactors known as tokamaks to trap the charged plasma. There's also what's known as inertial confinement, which compresses fuel with lasers, or pulsed power, fusing them before it all flies apart. And then there are hybrid concepts that try to combine these methods. So that's Fusion 101. Heat fuels into plasma at extreme temperatures, find a way to keep it contained and stable, and finally capture the energy. If it sounds tricky, That's because it is. And that balancing act is what made fusion such a big challenge to bring from idea to reality.
2:49|Gustaf Alströmer:
Uniting fusion is only half the battle. You also have to get more energy out than you put in. In the world of fusion, reaching a breakeven point has been an elusive goal for decades. And an economically valuable reactor, one capable of truly changing the energy landscape, would need to do far more than simply breakeven. Only one facility had claimed to reach this breakeven point. In December 2022, the National Ignition Facility at Lawrence Livermore Lab produced fusion output equal to laser energy delivered to its target. While it's great news, there's still a long way to go.
3:22|Gustaf Alströmer:
Historically, fusion products have shown that scaling up means massive machines, multi-billion dollar budgets, decades-long delays, and the old refrain that fusion is always 20 years away. But thanks to companies like Helion, the future is closer than ever before. Helion is taking a unique approach to fusion, combining proven ideas in a new way to build a smaller, faster generator and aiming to succeed with others that fail time and time again. Helion's fusion reactor is built around a unique kind of confinement system. Unlike the large, donut-shaped Tokamaks that dominates fusion research, Helion's device is linear, compact, and pulse-driven.
4:00|David Kirtley:
Helions generators are what is called a linear topology. So this literally means that it's a very long system. On either end, we inject our fusion fuel, this mixture of deuterium and helium-3, very quickly, less than a thousandth of a second.
4:14|Gustaf Alströmer:
Altogether, Helion uses what's called a magneto-inertial approach. The plasma collides over a million miles per hour. Magnetic fields then confine the plasma, rapidly compressing it to over 100 million degrees, igniting fusion. The whole process happens in under thousands of a second, making the system smaller, simpler, and faster to build than traditional designs. At the heart of Helion's approach is a bold fuel choice. deuterium, and helium-3. Most fusion products use deuterium and tritium, but tritium is radioactive, scarce, and produces high-energy neutrons, which are harder to capture for electricity. Helions' mix allow the reactor to generate mostly charged particles, meaning the energy is already in an electrical form.
4:59|Gustaf Alströmer:
Even better, helium has developed a way to generate helium-3 in its reactors by also fusing deuterium atoms together, solving the long-standing challenge of helium-3 scarcity without needing lunar mining or exotic supply chains.
5:13|David Kirtley:
One thing that we've pioneered here is the formation of helium-3. If you have a fusion system working, you can do that fusion of deuterium, make helium-3, add another deuterium, and now you're making electricity.
5:26|Gustaf Alströmer:
One of helium's bigger breakthroughs is their system for direct energy recovery. Instead of boiling water to spin a turbine like a traditional power plant, helium's machines captures the plasma energy directly through the same magnetic fields used for compression.
5:38|David Kirtley:
The good analogy to this is regenerative braking in electric cars.
5:42|Gustaf Alströmer:
As fusion reactions push back against the magnetic field, electricity is pulled directly out and stored in capacitors.
5:48|David Kirtley:
We take all that electricity, we turn it into electrical current in our electromagnets. And those magnets are the primary compressor, the primary thing we do fusion with. But also, as fusion happens, we can extract that electricity right back with those same magnets.
6:04|Gustaf Alströmer:
This method can reach any conversion efficiencies over 90%, making the entire system smaller, cheaper, and more efficient. Getting to this point wasn't easy. He then broke away from the slow traditional R&D model by running, building, testing, and designing in parallel. While operating one machine, they were already building the next and prototyping the factories needed to mass produce future reactors.
6:26|David Kirtley:
You don't just design a system and then build it and then test it and then learn from it and then design another system. You actually want to do all of that in parallel. A good example is as we were still running Trenta, our sixth generation machine, we were actually building Polaris, our seventh generation machine, based off of the lessons we had already learned. As builders, we want to be building generators, building prototypes. So we're always building hardware as fast as possible.
6:52|Gustaf Alströmer:
This hardware-first mindset focused less on theoretical perfection and more on real-world engineering, helping Helion move from early prototypes to the seventh-generation Polaris system in under one decade. Now, Helion's goal is clear. prove that fusion electricity, not just fusion energy, can be delivered reliably. With a partnership to power and Microsoft data center already lined up, the company is racing to build its first commercial-scale fusion generator, and if they succeed, could redefine the global energy landscape. For decades, fusion was the square in the realm of scientific research. Helion is trying to make it into a product.
7:29|David Kirtley:
Our goal is to have the cheapest source of electricity out there that doesn't generate carbon dioxide and can't make nuclear weapons. And if we can go do that in the world and deploy that quickly and safely, then I think we can radically transform standard of living throughout the world.
7:45|Gustaf Alströmer:
If they succeed, we can soon have a power source that's abundant, carbon free, and built for the needs of a rapidly evolving world. From the fringes of physics to power and technologies of tomorrow, fusion energy can be the breakthrough that reshapes the world.