If we do the fusion in zero g then we have solved the confinement issue. The problem is creating conditions for fusion in zero g. The simplest way would probably be aggregating enough material to a single spot that gravity itself creates conditions for fusion. But then the power plant becomes too energetic for earth so it has to be at an enormous distance away to be safe. And with that of course you have the problem of transmitting the power back to earth. But I think photons could be gathered at a safe distance from this fusion, to harvest it without having to be so close.
> This was a 25% improvement on the previous record time achieved with EAST, in China, a few weeks previously
I applaud this nuclear arms race. 22 minutes is really impressive for a technology that’s always been “20 years away”. I think I will do a deep dive on the technical challenges of fusion.
Specifically they were able to maintain a tokamak plasma (presumably at fusion temperatures) for 1337 seconds, using two megawatts of heating. 1337 is not a joke; presumably the "leet" reading is coincidental.
>> In the H-mode, a calm edge without turbulence reduces how much heat and how many charged particles the plasma loses. This leads to a sharp increase in pressure across the entire volume of the plasma, including the core where the conditions that can lead to fusion occur. The reduced energy and particle losses also minimize damage to the material surfaces surrounding the plasma.
One issue I see for applying prediction markets to things like “there is a commercially successful fusion power plant before the year 2070” is the long time until resolution. Now, of course, one can hope to sell your shares in “yes” or “no” 5 years from now, but there may not be enough liquidity?
Suppose we had one prediction market M_1 for “On January 1st 2070, resolves ‘yes’ if there has been a commercially successful nuclear fusion power plant, and otherwise resolves ‘no’”, and then another market M_2 that, maybe it resolves in 5 years as ‘yes’ if the price of M_1’s ‘yes’ is greater than 30%? Or… hm, that seems problematic because people could just buy a bunch of M_1’s “yes” right before M_2 resolves? Or maybe that’s a self-correcting problem because people could… no, still seems like a problem..
Well, what if instead of a prediction market about the future value of another prediction market, it was futures contracts for the shares in a prediction market? Like, the right to buy or sell shares in “yes” or “no” at a particular price?
So like, if you’re confident that the prediction market will assign probability p or higher on a particular day 5 years from now, then if you bought futures which, on that day each of the futures could be used to sell a share in “no” at the price (1-p), then…
well, if the probability assigned to “yes” on that day is indeed p or higher, then the price of “no” would be (1-p) or lower, so one buy a share in “no” at a price less than (1-p) and then sell it at (1-p)..
Hm, issue there is one still needs to buy the “no” in order to sell it, so that doesn’t seem to really fix the “what if there is no liquidity in 5 years?” issue?
I guess one could spend 1 to create a share of “yes” and a share of “no”, and then sell the “no”, and be left with the share in “yes” which is ostensibly worth at least p, and then like, sell it a bit later when there’s more liquidity or something?
I probably don’t know what I’m talking about about this.
For people more aware of the fusion industry, what is it that stopped the plasma at 22 minutes (or lower times in alternate tests)? Did they just stop injecting power to maintain the heat as they achieved their benchmark?
Is this something where it's on the precipice and small tweaks bridges from 22 minutes to basically indefinitely?
Related: PBS Space Time just did a neat episode on plasma, focusing on what the requirements are for confinement, called "The Final Barrier to (Nearly) Infinite Energy"
I do enjoy sharing this kind of news with all the fusion haters online. Fusion tech is legitimately cracking away on their "perpetually X-years away" stigma. That perpetual barrier can very reasonably be viewed as a normal technology barrier now.
Challenges persist, notably in developing materials that can endure prolonged exposure to extreme temperatures and neutron radiation within fusion reactors. Addressing these material durability issues is essential for the realization of continuous, commercially viable fusion power.
I fail to find this info: how much energy did they get out? How much did they pump in? How long does it have to burn until there's energy in-out equilibrium?
Fusion will not be commercial ever at least not for power generation.
Fission has potential for far cheaper fuel cost and can be done with less capital cost.
We are spending a crazy amount of money researching fusion while we have only explored like 1% of the potential of fission. If only part of this money was invested in fission we could have a competition for multiple advanced fission reactors.
It still eludes me how we are going to be able to reproduce the same temperature and pressure as in the core of a star considering the humongous amount of mass (hence energy) required to create those conditions.
AFAIU, no existing tokamaks can handle sustained plasma for any significant period of time because they'll burn down.
Did this destroy the facility?
What duration of sustained fusion plasma can tokamaks like EAST, WEST, and ITER withstand? What will need to change for continuous fusion energy to be net gained from a tokamak or a stellerator fusion reactor?
well done on the french achieving yet another extraordinary feat of engineering and research while still bearing the stigma of being shit at everything for some reason
I hope this international race ends up bearing fruit in a few decades, we need it
Nuclear fusion: WEST beats the world record for plasma duration
(cea.fr)482 points by mpweiher 18 February 2025 | 494 comments
Comments
I applaud this nuclear arms race. 22 minutes is really impressive for a technology that’s always been “20 years away”. I think I will do a deep dive on the technical challenges of fusion.
Still people make jokes about fusion research, some things just take time.
I recommend this excellent review of the even more excellent book „The future of fusion energy“
https://www.astralcodexten.com/p/your-book-review-the-future...
https://www.energy.gov/science/articles/science-close-develo...
>> In the H-mode, a calm edge without turbulence reduces how much heat and how many charged particles the plasma loses. This leads to a sharp increase in pressure across the entire volume of the plasma, including the core where the conditions that can lead to fusion occur. The reduced energy and particle losses also minimize damage to the material surfaces surrounding the plasma.
Suppose we had one prediction market M_1 for “On January 1st 2070, resolves ‘yes’ if there has been a commercially successful nuclear fusion power plant, and otherwise resolves ‘no’”, and then another market M_2 that, maybe it resolves in 5 years as ‘yes’ if the price of M_1’s ‘yes’ is greater than 30%? Or… hm, that seems problematic because people could just buy a bunch of M_1’s “yes” right before M_2 resolves? Or maybe that’s a self-correcting problem because people could… no, still seems like a problem..
Well, what if instead of a prediction market about the future value of another prediction market, it was futures contracts for the shares in a prediction market? Like, the right to buy or sell shares in “yes” or “no” at a particular price?
So like, if you’re confident that the prediction market will assign probability p or higher on a particular day 5 years from now, then if you bought futures which, on that day each of the futures could be used to sell a share in “no” at the price (1-p), then… well, if the probability assigned to “yes” on that day is indeed p or higher, then the price of “no” would be (1-p) or lower, so one buy a share in “no” at a price less than (1-p) and then sell it at (1-p)..
Hm, issue there is one still needs to buy the “no” in order to sell it, so that doesn’t seem to really fix the “what if there is no liquidity in 5 years?” issue?
I guess one could spend 1 to create a share of “yes” and a share of “no”, and then sell the “no”, and be left with the share in “yes” which is ostensibly worth at least p, and then like, sell it a bit later when there’s more liquidity or something?
I probably don’t know what I’m talking about about this.
I'm not sure how this works, how are they confident enough that they can make it produce net power?
Is this something where it's on the precipice and small tweaks bridges from 22 minutes to basically indefinitely?
An easier (more fun) version of this with some context is here: https://x.com/olshansky/status/1892069988707729614
https://www.youtube.com/watch?v=nAJN1CrJsVE
Nuclear fusion: New record set at Chinese reactor EAST https://news.ycombinator.com/item?id=42917662 03-feb-2025
China's artificial sun burns for 1000 secs, creates record in fusion research https://news.ycombinator.com/item?id=42854306 28-jan-2025
If you're a programmer, that is. Could mean something else in physics, right?
https://physicsworld.com/a/chinas-experimental-advanced-supe...
Fission has potential for far cheaper fuel cost and can be done with less capital cost.
We are spending a crazy amount of money researching fusion while we have only explored like 1% of the potential of fission. If only part of this money was invested in fission we could have a competition for multiple advanced fission reactors.
I have also read that achieving productive fusion will require temperatures above 100 million degrees.
Most of my sources are pop-sci, so correct me if I'm wrong.
AFAIU, no existing tokamaks can handle sustained plasma for any significant period of time because they'll burn down.
Did this destroy the facility?
What duration of sustained fusion plasma can tokamaks like EAST, WEST, and ITER withstand? What will need to change for continuous fusion energy to be net gained from a tokamak or a stellerator fusion reactor?
I hope this international race ends up bearing fruit in a few decades, we need it
* Fusion's potential is enormous.
* Health software, like fusion, needs breakthroughs.
* Are we ready for the data deluge?
* Can we build it fast enough?
Is this a joke and reference to internet culture or a coincidence? Probably the latter, but i found it entertaining.