In recent news, Microsoft has placed a purchase order for a fusion power plant.

Yes, you read that right, nuclear fusion.

The novel order came from a company called Helion. Founded in 2013, the company is working to build the world's first fusion power plant by 2028.

In other words, Microsoft is not so much buying electricity as it is investing in controlled fusion technology.

In its announcement, the company said:

This collaboration marks an important milestone for Helion and the controlled fusion industry as a whole.

We are very proud to have Microsoft as our first customer. This partnership advances the timeline for bringing commercial fusion energy onto the grid.

It's worth noting that,Helion's biggest investor is OpenAI CEO Sam Altman. He is also Chairman of Helion's Board of directors.

According to the Wall Street Journal, Altman has invested $375 million in Helion, his largest investment to date.

Ten months ago, the world's most high-profile CEO posted a job AD for Helion on his personal blog.

A fusion startup that Ultraman has invested in

So where is this controlled fusion startup that OpenAI's CEO is bullish on?

In terms of funding, the most recent public announcement was that Helion closed a $2.2 billion funding round in late 2021. The round was led by Altman, along with Meta co-founder Dustin Moskowitz, Peter Thiel's Mithril Capital, and Capricorn Investment Group, a well-known investor.

In terms of technology, Helion's technical route is unique.

At present, in the field of controlled fusion, there are two common technological routes:

One is to use magnetic confinement to achieve controlled fusion."Tokamak", as you often hear, is one of these technologies.

The other is inertial constraints. In general, this is achieved by laser ignition: a highly powerful laser beam is fired at the target (the target shot), causing the fuel in the shot (deuterium and tritium) to form a plasma. In a very short time, the plasma particles are compressed to a hot, dense state before they can fly around due to their inertia, and nuclear fusion occurs.

In December, scientists achieved the first net energy gain in a nuclear fusion reaction, using a laser ignition scheme.

Helion chose the magnetically confined route, but unlike the Tokamak, their proposal was a pulsed non-ignition fusion system.

Specifically, Helion is developing a plasma accelerator about 12 meters (40 feet) long that will heat deuterium and helium-3 to 100 million degrees Celsius, creating a plasma. A pulsed magnetic field is then used to compress the plasma until fusion occurs.

Such an approach could make fusion devices smaller and make it possible to adjust the power output as needed, officials explained.

Helion himself also mentioned that Helium-3 is a difficult isotope of helium to obtain. But it's cleaner and has a higher octane rating, which helps keep the system small and efficient. In order to obtain helium 3 steadily, Helion has been granted a patent to produce it using an efficient closed fuel cycle that fuses deuterium in its plasma accelerator.

The first nuclear fusion power station will be built in 2028

As mentioned above, Helion's goal is to have its first fusion power station by 2028, producing at least 50 megawatts of electricity (current fission plants produce gigawatts).

The planned power output is small, but it still sounds incredible.

But David Kirtley, Helion's founder and CEO, told The Verge that after signing a deal with Microsoft, "there will be financial penalties" if they don't get it right by then.

We promised to be able to build a system and sell it to Microsoft.

Interestingly, Helion employees are said to have started using ChatGPT to improve their productivity. If Helion's plans come to fruition five years from now, ChatGPT will, at some point, be able to use the power it helped produce (manual dog head).

So, do you think this is gonna work?