Sounds like Altman used his personal relationship with Microsoft CEO to do this meaningless PR stunt, so he can more easily pursaude stupid money to actually invest in this.
The numbers don’t seem out of this world to me, and it’s probably smarter money than what we are throwing at ITER given time to market.
I mean I kind of agree that the microsoft thing is probably a bit of a stunt give that I think they haven't generated any net electrical output at all yet. But I think they have a reasonable chance of getting it working at some point.
Ha! One of those "only on HN" comments. What would qualify as "difficult"? Mining it on Pluto?
Not criticizing though. I'm glad there are people out there who aren't deterred by these sorts of challenges.
By the time any actual work can be done at ITER, won't the technology have progressed beyond they design they are trying to test?
But in fact the engineering itself is exploratory. And often times, to understand something, you just have to start building.
That's what ITER is about. When the project started, we had gone about as far as theory and design could take us. You can see the same thing with other unproven fusion designs getting underway as well including the one at Lawrence Livermore.
And to be perfectly frank, a lot of it bursts onto the scene with sensational claims, never to be heard from again. At the end of ITER, there will be a thing. And probably, it will do most of what it was projected to do. We can't say the same for almost any of these other advances, compelling as they may sound.
Lastly, I'll point out this. "Breakthroughs" are rarely made in isolation. Are you so sure that without the billions that have gone into Lawrence Livermore, ITER and such that these other "breakthroughs" would even be happening? If there is no market for nuclear engineers, there will be fewer nuclear engineers. There's fewer people to bounce ideas off of and peer review your work. There's fewer people to come up with the small innovations that add up to a big result. There's fewer colleagues to get you that next gig where your team might beat a hard problem.
Helion isn’t vaporware.
Their seventh will start up in 2024, and is supposed to demonstrate net electricity from D-He3.
Their eighth is supposed to be the commercial reactor. If they don't achieve that, they will have to pay penalties to Microsoft.
I don't have anything near the expertise to evaluate Helions approach, but I know enough about fusion to say that just because you have a design that can achieve fusion conditions does not mean the approach could even theoretically lead to a power plant. e.g. electrostatic fusors
https://www.llnl.gov/news/lawrence-livermore-national-labora...
Lasers are really inefficient (poor at transforming electrical energy into light energy). They got more energy out than they put into the pellet, but nowhere near as much energy as they put into the lasers. Which means that turning that experiment into commercial fusion is... "not straightforward".
They seem to scale well, too. In their big shot they increased the laser power by 8% and output went up 230%.
That said, compared to Helion it'd still be a lot harder to make a practical reactor with NIF's approach.
(A la the scene in The Titanic: "You're money won't do me any good at the bottom of the ocean.")