Fusion Was a 70-Year Joke. This Machine Just Became a $15.5 Billion Bet

[00:00:00]There's a building in Everett, Washington with no windows. Inside it sits a concrete vault, walls 5 ft thick, wrapped in layers of radiation absorbing plastic. And inside that vault is a 16 ft machine that when it fires glows a bright fuchsia pink. It runs at 150 million° C. The core of the sun runs at 15 million. This machine is literally 10 times hotter than the center of the sun, and almost no one on Earth has been allowed inside the room with it.

[00:00:30]Recently, one journalist got in. What he described sounds less like a science experiment and more like a scene from a sci-fi movie. Miles and miles of black cables, stacks of capacitors seven high, two plasma rings slamming into each other at a million miles an hour. and Microsoft has already signed a contract to buy electricity from the commercial version of this machine starting in 2028. Today, we figure out if this is the real thing or the most expensive failure in the history of energy. Before we jump into the good stuff, I have a quick favor to ask. I've posted a poll on the community tab. Go let me know what you think. With that said, let's get into today's video. The company is called Helion Energy. The machine is called Polaris. It's their seventh prototype. The earlier ones were named after Starbucks sizes, Grande, Venti, Trenta, until they ran out of coffee metaphors and they had to start using stars. That tells you something about the culture because for 70 years, fusion energy have been treated like sacred physics done by governments in stadiumsiz machines on 20-year timelines that never actually arrive. Helium is trying to do it like a startup. Build it, break it, build a better one seven times in 13 years. The Polaris system is surreal in person. It lives in a building called Ursa. The majority of that building isn't even the machine.

[00:01:58]It's power supplies. 2500 capacitor units stacked seven high across a four* 4 pallets storing over 60 megajoules in a single pulse. the largest magnetic pulse power bank on the planet. The cables connecting those capacitors to the reactor, if you stretch them out end to end, they would reach 720 mi across Washington state and back. When Polaris fires, here's what happens in 1,000th of a second. [music] Hydrogen gas is injected at both ends of a 60 ft tube.

[00:02:30]It's ionized into plasma. Two rings of plasma are then accelerated toward each other at a million miles hour.

[00:02:37]>> [music] >> They collide in the middle, compressed by the magnetic field, and for a single instant, the interior hits 150 million°.

[00:02:46]High-speed [music] cameras catch a flash of fuchsia pink.

[00:02:50]Then the whole thing recharges and fires again. Now, here's what makes Helion fundamentally different from every other fusion company on Earth. When most [music] fusion startups imagine a commercial reactor, they imagine a giant kettle. Fusion heats a fluid. The fluid boils water. The steam spins a turbine.

[00:03:10]And yeah, that's how almost all electricity has been generated since the 1800s. Helium's founder, David Curley, looked at the plan and asked a simple question, why? His answer skipped the steam. When fusion reactions happen in Polaris, the resulting [music] charged particles push outward against the magnetic field. That push, just like breaking in an electric car, induces [music] an electric current directly. No heat cycle, no turbine, no moving parts.

[00:03:42]You pull electricity straight out of the reaction back into the capacitors and fire. Again, Helion claims this approach is fundamentally more efficient [music] than steam-based fusion. The question has always been whether it actually works at Kale. Now, let's be honest about the counterargument, cuz it's real. Helium doesn't publish its research. Independent physicists cannot audit the numbers. They've also missed deadlines before. The original target was net electricity from Polaris by 2024. Didn't happen. And the sharpest criticism comes from someone who should know. John Slow, the physicist whose original research helped create Helen's core technology. He co-founded the company, then left in 2018. In a scientific American investigation published in May, Slov said the fundamental problem hasn't been solved.

[00:04:33]At the speeds helium fires his plasma, severe instabilities cause what he called a catastrophic loss of magnetic flux before fusion can do enough useful work. In his view, they've hit a wall that [music] speed and engineering alone can fix. And there are signs the timeline is slipping. Scientific American now describes Orion as delivering power by 2029, a year later than the original Microsoft contract. If Helion misses that deadline, they face financial penalties from both Microsoft and partner Constellation Energy. Now those critiques deserve a fair hearing.

[00:05:10]Then February 2026 happened. [music] Helion announced that Polaris had achieved 150 million degrees C and became the first privately developed fusion machine in the world to demonstrate measurable dutyium treatium fusion in Malaga, Washington about 3 hours from Everett. [music] Helium broke ground firstly in July 2025 on Orion. The first commercial fusion power plant in human history. The one that's supposed to deliver electricity to [music] Microsoft in 2028. And since February, the momentum has only accelerated. Last month in June, Helion raised $465 million in a series G round, nearly tripling its valuation to 15.5 billion.

[00:05:56]Total capital raised over $1.5 [music] billion. Ford Motor Company's executive chairman, Bill Ford, joined as an investor. Time named Helian one of the 100 most influential companies of 2026.

[00:06:10]Sam Alman, who owns roughly a third of Helion, a stake now worth over 1.6 billion, stepped down from the board in March. The reason? Open AAI is now exploring its own power purchase deal with Helion. That's right, the company behind Chad Gupt might become Helion's second major customer. But perhaps the most telling move, Helion quietly revealed it's building a new machine called Tiny Merge, 1/8 the size of Polaris. Because key technical questions still need answers before Orion's final design can be locked in. They also secured Washington state's radioactive materials license and air emissions license for Orion. The regulatory first for any fusion company. So where does this leave us? The physics is real. The temperature records are real. The contract with Microsoft is real.

[00:07:02]Actually, the concrete is being poured right now. But the energy gain, the one thing that would definitely prove fusion power is no longer 20 years away, has not been confirmed. Not by Helion, not by independent reviewers, not by anyone.

[00:07:16]And the co-founder, whose physics made this machine possible, says the design has a fundamental flaw. What we do have $1.5 billion in total funding, a $15.5 billion valuation, regulatory licenses secured, a construction site in Malaga, and a deadline that's possibly shifted from 2028 to 2029. [music] For the first time in my life covering clean energy, the sentence fusion in two years is [music] attached to hardware that exists, a contract that is signed and a site under active construction.

[00:07:49]That's either the beginning of the largest civilization scale shift since the industrial revolution or the most expensive lesson in energy history.

[00:07:57]Either way, we're watching. I'll drop a follow-up video the moment Helion confirms net electricity, positive or negative. Subscribe and hit the bell so you're there when it happens.