Nuclear fusion can’t produce net energy, so is it really a solution to AI’s growing energy demands?
nuclear power produces long-lived radioactive waste, which needs to be stored securely. Nuclear fuels, such as the element uranium (which needs to be mined), are finite, so the technology is not considered renewable. Renewable sources of energy, such as solar and wind power suffer from “intermittency”, meaning they do not consistently produce energy at all hours of the day.
fusion technologies have yet to produce sustained net energy output (more energy than is put in to run the reactor), let alone produce energy at the scale required to meet the growing demands of AI. Fusion will require many more technological developments before it can fulfil its promise of delivering power to the grid.
If the tech brows wanna dump money into developing renewable energy systems, detaching themselves from our main power grid they currently destabilize. Let them!
No they don't; this is literally the first thing I've ever read claiming that. Tech bosses are perfect happy to power AI with nuclear fission and don't give the slightest fuck about the waste.
(As well they shouldn't, TBH, since it really ought to get reprocessed anyway. But that doesn't excuse them for wanting to waste the power on bullshit.)
That turns out to not be true, at least not with the tokamak reactors most groups are pursuing.
You see, at some point you need a shield around the reactor to actually absorb all the high energy particles released, and turn that energy into heat. That's the whole point of the reactor, to generate heat and run a turbine. You absorb those high energy particles with a "blanket", that's just what they call the shield around the reactor.
Here's the issue, absorbing all those high energy particles necessarily results in transmuting the material absorbing them. That blanket becomes brittle and eventually needs to be replaced. Not coincidentally, that blanket is also now radioactive, because you've bombarded it with protons and neutrons and it's now partially made up of unstable, radioactive elements.
So while fission reactors have radioactive fuel rods to dispose of, fusion reactors will have radioactive blankets to dispose of. Who knows if this is an improvement.
If it ends up working though it's not a waste of power is it? And if it doesn't work then, oh well.
Big tech companies do a lot of cramp, but this one I actually don't really mind. You never know we might actually get the Star Trek utopia we've always wanted from this, it's unlikely but it's not impossible.
I don't see why we are hating on the waste like this. Yes it's very dangerous waste, but the amount is quite small, and if we store them safely, as shown in Tom Scott's video on Nuclear Storage in Finland, it's actually a very good solution for the time being.
"if we store them safely" - here's the problem with the entire argument. Nobody wants to pay for it, so they won't unless they are forced to. Carbon capture is a viable technology but it costs money to implement at a net financial loss, so nobody uses that if they don't have to either. The problem is the same as always - nobody who stands to lose money gives a damn. The planet dying is somebody else's concern tomorrow, and profits are their concern today.
Maybe Tom Scott should make a video about the Asse salt mine. It's where the "yellow barrel == nuclear waste" meme comes from look here a picture.
This stuff is the driving factor behind nuclear energy being a political no-go in Germany: We just don't trust anyone, including ourselves, to do it properly. Sufficiently failure-proof humans have yet to be invented. Then, aside from that: Fission is expensive AF, and that's before considering that they don't have to pay for their own insurance because no insurance company would take on the contract.
Fusion OTOH has progressed to a point where it's actually around the corner, when the Max Planck institute is spinning out a company to commercialise it you know it's the real deal. And they did.
They've seen it being always reached in computer games like Civilization
They think the hard part is in becoming the big boss to decide things. The civilization part is easy, just direct resources where you need the "cool thing completed" notification to appear.
I was under the impression that the major two advantages of fusion were exponentially more power output, provided we can actually sustain it, and no radioactive byproducts....
The easiest fusion reaction is deuterium/tritium - two isotopes of hydrogen. The vast majority of the energy of that reaction is released as neutrons, which are very difficult to contain and will irradiate the reactor's containment vessel. The walls of the reactor will degrade, and will eventually need to be replaced and the originals treated as radioactive waste.
Lithium/deuterium fusion releases most of its energy in the form of alpha particles - making it much more practical to harness the energy for electrical generation - and releases something like 80% fewer high energy neutrons -- much less radioactive waste. As a trade-off, the conditions required to sustain the reaction are even more extreme and difficult to maintain.
There are many many possible fusion reactions and multiple containment methods - some produce significant radioactive waste and some do not. In terms of energy output, the energy released per reaction event is much higher than in fission, but it is much harder to concentrate reaction events, so overall energy output is much lower until some significant advancement is made on the engineering challenges that have plagued fusion for 70+ years.
Renewable sources of energy, such as solar and wind power suffer from “intermittency”, meaning they do not consistently produce energy at all hours of the day.
If only we had some way of storing energy for use later. Oh well.
We do not currently have the battery tech to have a fully renewables-powered grid where batteries are used for the regular dips in production wind and solar have.
We likely won't have infrastructure like that in place for decades.
Do you know what they do in Norway with out-of-use old mines? They lift a load when there's energy to be stored. They lower it when there's energy to be spent. I'm sure you know how electric engines work and that the conversion is symmetric.
No battery tech involved.
Battery tech is in general mostly relevant for autonomous devices we carry, for airplanes and ships, for cars.
For the central grid the ways to store energy are almost inifinite.
Really shitty scaremongering article. I'd like to know how exactly increased investment in fusion could potentially make it unsuitable for public use, as the article claims?
So maybe they will invest to get it further. It's not a 9 women can make a baby in a month .. but sufficient funding for next gen nuclear and fusion will help progress.
The best way I have heard it described is that Fusion is going to happen next year but probably not in the next 12 months.
We think Fusion must be coming soon because we understand all of the fundamental principles around how it works, so what we need to do is put those principles into practice. For some reason though that doesn't quite work what we end up with is a machine that makes a lot of noise but doesn't really achieve anything
That's been the history of tokamaks because they're dealing with an inherently unstable situation. It's like balancing a ball on another ball, saying "yep ok I've figured that out", scaling it up and discovering that between those two balls were actually five other that now that the system is bigger have quite a relevant impact.
Contrast with stellerators, which are more like balancing a ball in a bowl. Long considered impossible because the magnetic field just has a too complex geometry the Max Planck institute proved that they work as the theory says, and they're currently working on commercialisation.
Maybe AI can help us break the fusion hurdles. Oh. It's still telling people to eat rocks, just used to create waifu porn and as a mass spy application? Nothing else, really? Well shit.
I know you're being reflexively downvoted by who hate everything AI, but this is the sort of thing AI should be most useful for, which is finding patterns within large problem spaces with many variables.
Probably. Proxima fusion is using simulation-driven engineering to pave their way through the design space, no matter how you approach it it's gotta involve dimension reduction in some way and that's ML. They speak of AI but well it's a press piece.
LLMs or diffusion models? Nah, don't think so. This is actual engineers throwing statistics at a particular problem to identify what prototypes they should build, not techbros throwing shit at the wall.
It's even bad at porn. Very limited means of describing the process, forgetting that there are no bed sheets in a park, same repeating metaphors. Boring.
I think people still don't understand what the problem is with fusion. The problem is not that it doesn't work, it will work, and soon. The problem is that everyone seems to think fusion means cheap limitless energy, and that couldn't be further from the truth. When fusion does finally work, it will be the most expensive form of energy available. That's going to be a gamebreaker, right out of the gate.
So far, the only method we know of to guarantee that your reactor will be energy positive is to make it truly enormous. Let me tell you, going truly enormous is not a good way to keep costs down. But let's say you just spent 8 years building a cutting edge fusion power plant and you want it to work smoothly. Well you better hire a large team of nuclear physics PHDs to keep that reactor working, they must be a dime a dozen, right? You'll need them for all the maintenance of your cutting edge reactor, get ready for those maintenance costs to mount up. And be prepared to continue paying for all your staff and facilities even while the reactor is (frequently) power down for maintenance.
Also, you do have an economical way to dispose of nuclear waste right? Because fusion reactors are probably going to generate a significant amount of nuclear waste... That's one of the side effects of actually turning the high energy particles released by the reactor into heat. Those free protons and neutrons get absorbed by a physical shield around the reactor called the blanket. That blanket becomes radioactive over time and needs to be replaced. Congratulations, your clean fusion energy is now producing radioactive waste, and your back to the exact same problems we have with fission.
My advice, keep using the nuclear energy we understand really well at this point, fission. Also, renewables and storage are already a cheaper solution to do exactly the same thing fusion will eventually do. I'm certain that fusion will be a fantastic technology for large spacecraft someday, but I make no promises it will ever become the first choice for general terrestrial power generation.
It does, it produces low and intermediate level waste, the waste decays over about a hundred years, not thousands. So it's better, but still an issue requiring management.
Your comment doesn't stand up. It seems you've got something against fusion energy for some reason.
On cost: it's a best guess, since we don't yet have a working fusion reactor. The error bars on the cost estimates are huge, so while it is possible fusion will be more expensive, with current data you absolutely cannot guarantee it. Add to that the decreasing costs as the technology matures, like we've seen in wind and especially solar over recent decades.
On nuclear physics PhDs: that's no different to any energy generation, you need dozens of experts to build and run any installation.
On waste: where are you getting this info on the blanket? The old beryllium blanket design has been replaced with tungsten and no longer needs to be replaced. The next step is to test a lithium blanket which will actually generate nuclear fuel as the reaction processes.
This is the important fact that you have omitted, for some reason.
Nuclear fusion reactors produce no high activity, long-lived nuclear waste. The activation of components in a fusion reactor is low enough for the materials to be recycled or reused within 100 years
And that is why it's so important this technology is developed. It's incredibly clean and, yes, limitless.
As for your advice, there was a time not long ago when we didn't understand how to build fission plants either, and it cost a lot of time and money to learn how. I wonder if people back then were saying we should just stick to burning coal because we know how that works.