Itâ(TM)s planned to be 500MW, not 2GW, but Commonwealth Fusion Systems have a solution to the problem youâ(TM)re alluding to on their ARC reactor.

The problem is that fusion gives of neutrons, which smash into the reactor walls, and damage them. Worse, the damage makes the reactor radioactive over time. The bigger problem is that in the most common design (a tokamak), you need to build huge superconducting magnets tight around the outside of the reactor walls, which mean that to do maintenance, you need to take apart the reactor walls and extract them from in between extremely sensitive and expensive magnets, all while not actually standing there, because the area is irradiated. The even bigger problem is that the reactor walls need to be made of some pretty exotic materials because you need them not only to survive high energy neutron bombardment for a decent while, but also cool 500MW of heat on a relatively tiny surface area (after all the whole point is how energy dense these things are), and even more importantly, you need them to take those high energy neutrons convert them to more, lower energy neutrons, and then use them to breed tritium by smashing them into lithium. That means you need a bunch of beryllium to act both as a moderator, and to turn each high energy neutron into two lower energy ones. Meanwhile, the reactor has to resist the forces of enormously powerful magnetic fields trying to crush the thing.

So the tough situation. You have weird materials that need to put up with huge pressures, huge temperatures, high neutron fluxes. They need to breed the fuel, and then they need to be taken apart, and removed by robots through a letter box between huge magnets that canâ(TM)t be damaged. Big problem.

The solution CFS are designing and soon building is that their magnets are made from REBCO tape, and can have âoejumpersâ inserted into them that connect each tape across a gap in the magnet. They are then soldered on which makes the magnets very slightly non superconducting, but not enough so to make them problematic. Once youâ(TM)ve done this, you can split your magnets in half, which lets you crane out the whole of the inner reactor without opening it, set it down in a spent fuel pond, and crane in a new one. You then make your blanket out of molten salt (specifically FLiBe) which you pump through 3D printed cavities in the reactor walls. That means you can continuously refresh your coolant/moderator/breeder by just changing it out as itâ(TM)s pumped around the system. The reactor walls meanwhile can be made from a steel alloy that can survive the heat and embrittlement for a couple of years while conducting the heat into the FLiBe through a complex 3D printed structure thatâ(TM)s both strong and high surface area.

Thereâ(TM)s a couple of good talks about their plans on YouTube. Thereâ(TM)s this one that gives you the overall concept: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fm.youtube.com%2Fwatch%3Fv%3D... and then this one that goes into more details about exactly how you design the reactor walls: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fm.youtube.com%2Fwatch%3Fv%3D...

Both of those are at a more academic level because theyâ(TM)re from MIT PSFC researchers, but this stuff is actively being engineered and commercialised by CFS. Their demonstration reactor is well under construction, and the site for their first power plant is selected and will start construction soon.

What about the cited sources? What makes them unreliable?

Saying "that's a design issue with the seat" isn't going to help you when you're 36000 ft in the air, and have a damaged battery that could easily burst into flames, and not be exinguishable.

Saying "it's alarmist" doesn't change the fact that you can't put out a lithium battery fire with the battery jammed in a seat, or the fact that a fire inside a metal tube that no one can leave is a really really bad scenario.

The idea that you would not immediately land a plane with a damaged battery jammed in an inaccessible place is ludicrous.

You can't put a battery that is jammed in a seat on an aluminium pie plate.

Wow, this is an interesting weird belief that seems to permiate this thread. Go and poke a hole in a lithium battery with a nail, and see what hapens. Hint - if you expose lithium metal to water in the air, BAD things happen very very quickly.

It's not just the potential, it's the problem of containing the fire if it does happen.

First, an iPad with a visibly damaged battery is much more likely to catch fire than one that's plugged in (and this alone would be enough to divert the aeroplane, since generally the rules on such things require a plane to be landed as soon as possible for any damaged lithium cell).

Second, it's jammed in a seat. The only way to stop a lithium fire from burning the whole plane is to put it in a bucket of sand. If you can't do that, you basically gaurentee that the fire is going to spread. Once a fire starts on a plane, it takes only a few minutes before the atmosphere inside it is unbreathable, and before the pilots are unable to see the controls, let alone out the windows. From there it only takes a little while longer before control systems fail, and you have a major disaster. Fire on a plane is about the worst case scenario.

I mean, first of all, âoethe big oneâ isnâ(TM)t expected in California, itâ(TM)s expected in Washington/Oregon, when the cascadia subduction zone next slips.

Secondly, assuming youâ(TM)re thinking of next time the San Andreas fault slips, thatâ(TM)s certainly wonâ(TM)t result in any small part of the state becoming a reef, let alone the whole state. The San Andreas fault is a strike/slip fault, it moves sideways, not up and down.

If people are living a significant distance from any office, the requirement to move to keep your job could indeed be terrible.

Cmnd-`.

Yeh, the idea that gnome is less buggy or more consistently good quality than macOS is frankly laughable.

Yes we do. The only provider at the moment has demonstrated that theyâ(TM)re willing to pull the plug on you, or even actively provide information to your enemies if they decide they donâ(TM)t like your country. We need several more constellations, and we need them from several countries.

I mean, Bezos hasnâ(TM)t been doing Nazi salutes, destroying important parts of our government, and fucking up the system that gives wellbeing to people in their old age, or disability. Donâ(TM)t get me wrong - I donâ(TM)t have any illusions that Bezos is a wonder, in fact, heâ(TM)s probably just smart enough to stay in the shadows while Musk destroys himself, but Musk has shown his hand.

SpaceX is launching roughly once every 4 days at the moment. Amazon though is not using their rockets though, or in fact, any currently reusable rockets. Iâ(TM)d bet that Amazon will be lucky to get 5 stacks of Kuipers in orbit by the end of the year. Blue origin do of course have New Glenn reuse in the pipeline, but I expect itâ(TM)ll take another few launches before they can nail it reliably, and then itâ(TM)ll be another couple of years before they have a fleet of them that can launch with any regularity.

We almost certainly *are* in an epicycular trap, but *even then* each explanation is more correct than the last. Each one models reality slightly better than the last, and is able to make predictions slightly better than the last. It may well be that at some point we'll take a significant step by completely rejigging how the model works, but that doesn't change the fact that each one improves the situation.

Itâ(TM)sa machine powerful enough that if you said you had doom running on it, Iâ(TM)d wonder why you bothered. Itâ(TM)s a couple of orders of magnitude faster (at least) than the machines that originally ran it. How on earth have we got to a point where such a machine takes 5 minutes to load a simple game library?