Microsoft Bets That Fusion Power Is Closer Than Many Think

[burkaman]

Do they really believe that, or are they just trying to look greener and make their carbon negative commitment look more real?

> “We wouldn’t enter into this agreement if we were not optimistic that engineering advances are gaining momentum,” said Microsoft President Brad Smith.

Why not? What's the downside for Microsoft, even if they think this probably won't happen by 2028? They even get paid "penalties" if Helion doesn't deliver.

[asdfman123]

> are they just trying to look greener

Absolutely not! They’re also trying to look cutting edge.

[dotBen]

Given the implications of this technology I wonder how much development is actually being done "in the open". And therefore how close to a breakthrough we are, whether a breakthrough has already occured and we are being fed delayed findings, and ultimately how a private organization like Microsoft would truly be able to access matured tech.

Whether by incentive or by coercion, it feels like the DOD and the DOE would want to have first access to any breakthroughs and also guard knowledge transfers (ie protect any program - commerical or government - from foreign state espionage).

The DOE is basically an entire national security organization centered around nuclear security, it seems unlikely they would be standing by watching R&D in fusion occur without a seat at the table.

[DennisP]

Given climate change, if Helion works then it would be in the interest of the US for it to spread worldwide as quickly as possible. Especially since it uses a mostly-aneutronic reaction.

[aaronblohowiak]

Are most fusion reactor designs a national security concern? I am not a nuclear physicist but from my rudimentary understanding of most modern approaches I don’t see how they would help with bad stuff (other than creating material for a dirty bomb through waste perhaps)

[mensetmanusman]

Disruption outside of the government eye is common, I wouldn’t be surprised. Those organizations focus on nuclear weapons tech.

[gregw2]

My impression is that Gates (and other tech billionaires) have been investing in fusion for a decade, so he probably knows or has leverage/access to people who know what’s going on, ergo Microsoft does too.

There is expertise in DOE and DOD, but there is an also a lot of private capital being invested in companies that don’t need to talk themselves up much. People in the industry generally know who most of the players are; that’s how you or your technical peers/friends get a job after all.

I don’t see a particular need for there to have been a top secret breakthrough to explain this deal.

[deadeye]

Worth watching about Helion:

https://youtu.be/_bDXXWQxK38

[maeln]

The response by Improbable Matter is also interesting in its own right https://www.youtube.com/watch?v=3vUPhsFoniw

[bob1029]

Agreed - I had no idea they were skipping the steam/turbine stuff. This is a compelling concept because it looks like it could actually scale.

[PaulHoule]

Gates might also be the most enthusiastic person about fast breeder reactors as he has put a lot of effort and money into SMRs, breed-and-burn fuel cycles, and making up the for the high cost of a three-stage heat transfer system by using the intermediate coolant to store energy ahead of time.

[steve1977]

Is that why they keep developing stuff with Electron?

[causi]

Microsoft Bets That Fusion Power Is Closer Than Many Think

"Bet" implies a potential for loss. This is a couch-cushion spare change level "bet". Microsoft probably spends more on printer ink every month than they stand to lose on this bet.

[onlyrealcuzzo]

It's a power purchase agreement.

They're going to get ~$50M worth of power. Either they're going to overpay for Fusion or overpay for energy from some other source if Helion can't deliver and has to pay for it via penalty.

It's not really much of a risk for MSFT.

But it's a great headline for Helion to pass on to clueless people.

[chx]

https://www.cnbc.com/2023/05/10/microsoft-agrees-to-buy-powe...

> Helion’s deal with Microsoft is to get 50 megawatts online,

Looking at solar costs https://www.iea.org/data-and-statistics/charts/capital-costs... 50MW in solar would cost 50-100M and I can't imagine this could be much more expensive or it wouldn't be competitive at all. Microsoft annual operating expenses for 2022 were $114.887B so about a 1-2000 times more.

[lisasays]

https://archive.ph/oYpP6

[avalys]

Given that Helios hasn’t demonstrated production of any amount of energy from fusion, this is probably just bullshit, like the United “order” for supersonic jets from Boom a few months ago. Similarly, Boom has not even flown a prototype yet.

[emaginniss]

I'm not sure why you're so negative on a business model that has worked for tech startups for decades now. Many businesses have started by promising to fix a problem experienced by a larger entity. Along with product guidance, the startup often gets capital in the form of investment. The larger company gets a low-risk attempt to solve their problem with the possibility of upside in the case of a larger than expected success.

[asdfman123]

Helion gets to survive 5 more years, Microsoft risks nothing, they both get good PR for associating themselves with one another.

[justrealist]

Helios is building real facilities. Maybe it won't work, but it's not just drafting CAD templates — they are deploying capital into physical reactors. That's not bullshit.

[user6723]

The Safire plasma has self-containing magnetic fields and does not require artificial magnet infrastructure. Safire plasma can stay "lit" at under 240 watts of power input almost indefinitely.

Here is some raw video of plasma in a Safire chamber vaporizing a tungsten rod https://www.youtube.com/watch?v=7y46wMAHnsI

[gumballindie]

We just need to redefine fusion and we have it.

[throwuwu]

Sounds like someone has been talking to the base model

[markus_zhang]

If we get fusion power and ai and longevity up in next 10-20 years it's going to be nightmare for ordinary people. Yeah, nightmare, not paradise.

[partiallypro]

I don't see the fusion argument, but I can see the AI argument at least temporarily making lives worse. What is the argument that fusion power would? It would definitely destabilize countries like Saudi Arabia, but otherwise seems like a win. Even still you have to have petroleum products for most anything and everything.

[throwuwu]

Wide adoption of longevity tech is necessary to prevent population collapse. AI also doesn’t seem like it’s going to stay cooped up for long, there’s a solid chance that open source projects beat OpenAI to the punch as well if scaling is no longer productive beyond this point

[slothtrop]

A rapid transition would necessitate the need for legislative/policy changes. People aren't going to complacently waddle in "nightmares".

Workers already feel priced out of the housing market, just add it to the pile.

[HPsquared]

Not fully automated luxury space communism?

[ThomPete]

Fusion is not just another way to power our lightbulbs it's a completely different kind of energy.

Two unique properties of energy density like Fusion.

1. The ability to hit 10% of the speed of light in space 2. The ability to create fundamental elements like gold, silver etc.

Another good example of the fact that there are no scarcity of resources only scarcity of knowledge. Resources are created not found.

Edit: Why am I being downvoted?

[tasty_freeze]

Fusion releases energy when producing items lighter than iron. It requires energy to create atoms heavier than iron via fusion. It would cost huge amounts of money to produce an ounce of silver or gold that way.