I don't think anyone is proposing continuous fuel (even fission reactors typically don't work like that - you have to shut them down to swap out the fuel rods periodically). The idea is that you inject (say) some deuterium into a blast chamber, it explodes like a miniature thermonuclear bomb, and you gather the explosive energy in vaguely the same way a internal combustion engine (like in cars) gathers energy from fuel-air bombs in its engine cyliders.

And we already know (roughly) how to build that, it's just that the version we know how to build is so ridiculously, impractically huge that you could fit a small city in its blast chamber... because it runs on actual, non-miniature, thermonuclear bombs. The difficulty is scaling that down to something that doesn't have a upfront cost measured in multiples of the world GDP before it starts generating power (in quantities measured in multiples of current global electrical consumption, and likely multiples of global energy consumption period).

The blast chamber that you chuck nuclear weapons in is not a good analogy, I suspect we don't know how to build such a thing that has a net positive energy yield when you factor in how much energy it takes to build a thermonuclear weapon and the terrible thermal efficiency such a contraption would have. Inertial confinement fusion is basically a more manageable version of this idea and we still haven't hit breakeven there. Magnetic confinement fusion is widely believed to be the more viable path to commercial fusion energy anyway. Right now fusion is largely a game of chasing margins to get things efficient enough that first the plasma is self sustaining (ignition) and then second yields a net positive of energy. ITER or its successor will hopefully achieve the first goal (ITER was originally supposed to but is no longer targeting full ignition, at least initially), then hitting the second goal is what will be needed to be practical as a power source.

Also, just to add to your point there are (basically) continuously fueled fission reactors, see e.g. CANDU as well as some more exotic actually continuously fueled designs that use liquid fuels. Continuously fueled designs would probably be more desirable if they weren't such a proliferation concern but that's a different story.

Wait, are you saying we know how to build a World Wonder that produces infinite energy, but we’re not doing it? What’ll we do if Ghandi builds it first?

Because we don't have enough energy demand to make it worthwhile. Even if we had enough money (and political will) to fund it, we'd be better off spending it on a space elevator, since we can actually use that enough to cover the costs.

With the only issue being that we don’t actually know how to build a working space elevator yet. Last I heard anyway.

Stars and fusion nuclear weapons exist.

And I'm only half joking. The energy yield of fusion is there and the basic physics are pretty well understood. Progress has been agonizingly slow but steady. We generally understand what needs to be done to get there, but it's a slow and complicated process that will take a long time and several more expensive, large-scale tests like ITER to push it along. The payoff is still a ways away and the up front costs are high, so barring investment on the scale of the Manhattan project it is something we will have to continue slowly grinding away at. The upside is fusion research is truly pushing the bounds of engineering, which will continue to yield benefits in other areas along the way.

Edit: I'll also add that we have already built fusion reactors, we just haven't gotten them efficient enough that they generate enough energy to sustain themselves.

>barring investment on the scale of the Manhattan project

The Manhattan project cost about $30 billion in today's dollars, I believe.

We have IPOs that raise close to that amount of money these days (e.g. Alibaba).

It's half a percent of federal spending in 2020 (admittedly not a normal year).

GameStop has lost over $30 billion from its peak valuation.

Your matter of fact claim that there's not enough capital in today's world for fusion power development really makes me wonder.

Once upon a time, I read a comment by someone somewhere that "everyone" (who is "in the know") knows fusion break-even is infeasible, but the research is relevant to maintaining nuclear weapons stockpiles.

Besides, if fusion power was possible...wouldn't Elon Musk be working on it?

Comparing the raw numbers in terms of cost I don't think quite gives the right picture, you have to think about scale of investment. Whole (large) towns and biggest-in-the-world facilities were raised and the brightest minds committed. And it would likely take longer. Maybe a better comparison would be to the Manhattan project and the atomic weapons research apparatus in the decades that followed. Current input to fusion research is a big monetary investment worldwide, but I think in terms of effort it's not quite the same scale especially if you talk about the whole world committing to that degree.

I will add that I'm not sure it would be worth committing this amount of resources solely to fusion. Obviously I think it's worth pursuing long term, but not necessarily at that scale given the costs it would have in terms of other scientific research. I never claimed the amount of capital isn't there nor meant to imply it; it probably is possible (by 2040-2050, 2030 seems impossible to me) but not realistic barring some dramatic changes in the world.

Regarding why wouldn't Elon Musk be working on it... let's ignore that several other prominent billionaires have invested in fusion. But currently fusion is not an appealing private investment outside of philanthropy. There are a lot of up front costs and uncertainty for a payoff that is way down the road. Whether it's worth investing in is up to him, but "whether or not Elon Musk is doing it" is not a good measure of viability.

Regarding the comment that everybody in the know thinks breakeven is infeasible, that's certainly not the prevailing nor even a common view with the people I know working on fusion. Opinions are mixed on how far we are from breakeven and how far we are in terms of the technology required to achieve it, but as I noted previously the basic energy yield of fusion is there and the route to get there is mostly clear, if long. ITER has little cross-application with nuclear weapons, that would be facilities like NIF which aren't the main focus of research towards power production.

I'll add that I have no skin in this game, my research is unrelated to fusion I just happen to talk with those guys a lot and took some classes in grad school. I see none of those sweet gubbment fusion research bucks.

>you have to think about scale of investment

That's what I was doing...

>currently fusion is not an appealing private investment outside of philanthropy. There are a lot of up front costs and uncertainty for a payoff that is way down the road

As opposed to the Mars rocket business? As opposed to money-losing biotech? Bitcoin, now representing around $1 trillion in total? Is the disinterest in long term investments why the US government is thinking about issuing 50 and 100 year bonds, as well as the 30 year that already exist?

In 2021, saying that significant capital is only available to non-speculative investments with short term payoffs sounds like something from another dimension. One without the GameStop hearings.

Elon Musk isn't literally the sole litmus test of whether fusion is feasible; he just represents the general zeitgeist. If a few billion dollars would make it happen in the near future, how could it be that everyone is disinterested? Especially now that taking climate change seriously seems to be snowballing.

> As opposed to the Mars rocket business?

SpaceX talks a lot about Mars, but so far they've mostly stuck to Earth orbit, because that's where the money is. They are starting to do some Moon-related stuff – because both NASA and space tourists will pay to go there. Even Starship, which was designed as a vehicle to get to Mars, most of its actual use is going to be Earth orbit or lunar, it is going to be very many years before its Martian use cases outnumber its Earth/Moon ones.

SpaceX will probably fund a few unmanned trips to Mars out of their own pocket – land a couple of Starships containing supplies that might be useful to future colonists. But I'm sure straight after doing that they'll be knocking on NASA's door suggesting that NASA pay them for some Mars missions... imagine how many (and how big) rovers you could fit on a Starship.

Musk has said he wants SpaceX to develop the necessary transport technologies to support colonisation of Mars – and they'll try to use non-Mars-related revenue sources to pay for that development. But he's also said the role of actually setting up a Martian colony (as opposed to just providing transportation to get there) won't fall to SpaceX.

I have a theory – which could just be wild speculation on my part – at some point Musk is going to set up something like a "Mars Colonisation Foundation" as a non-profit tasked with colonising Mars. And when he dies he's going to leave it the vast bulk of his assets to it (I'm sure he'll leave his family enough to live comfortably, but they don't need billions to do that). And the "Foundation" will end up owning a big chunk of SpaceX, and will be a major customer of SpaceX, but will remain separate from it. It will actually try to pay for the colonisation of Mars, not SpaceX themselves. And it will stick at it as long as it takes (even if it takes a few lifetimes to pull it all off.)

It's not a few billion dollars, it's hundreds, and probably decades before it pays off. Private investors have dabbled with alternative approaches to fusion that promise lower costs, but none of them have really panned out and it's pretty much all fallen to government sponsored research.

I expect private investment will pick up eventually, when the technology is a little closer to market. I'm not at all disagreeing that we should be investing more in it and that it's looking increasingly urgent, it's still a relatively small effort compared to other things.

Well, ITER will cost 22-65bn. It's much more ambitious than the Manhattan Project IMHO.

> GameStop has lost over $30 billion from its peak valuation.

This isn't how market cap works.

I know what market cap is, and that they didn't really have the opportunity to sell the company for that much, but it still provides a sense of scale.

A very misleading one.