7 ms·
> This is more than double what was achieved in similar tests back in 1997. The experiment produced > 11 megawatts of power and at Jet > two 500 megawatt fl
by dfdz 5y ago
> This is more than double what was achieved in similar tests back in 1997.
The experiment produced
> 11 megawatts of power
and at Jet
> two 500 megawatt flywheels are used to run the experiments
log_2(2*500/11)*(2022-1997) + 2022 = 2185
At this rate we will have fusion by 2185 I guess?
- DrBazza 5y agoIt's 30 years time. It's always 30 years time. https://www.discovermagazine.com/technology/why-nuclear-fusion-is-always-30-years-away https://www.discovermagazine.com/technology/why-nuclear-fusi...
- actionfromafar 5y agoI think it used to be in 20 years time. https://bigthink.com/videos/fusion-really-is-20-years-away-2/ https://bigthink.com/videos/fusion-really-is-20-years-away-2...
- giantrobot 5y agoInflation.
- ambrozk 5y agoThis is a very dumb meme that people use to avoid learning anything about the actual hurdles that are currently being faced by fusion researchers. At some point, fusion will be much less than 30 years away, and at that point, I guarantee you that lazy people will still be repeating this joke, because it is literaly the only thing they know about the subject.
- khafra 5y ago"Fusion is always 30 years away" is a heuristic that almost always works. https://astralcodexten.substack.com/p/heuristics-that-almost-always-work https://astralcodexten.substack.com/p/heuristics-that-almost...
- mcswell 5y agoWe did nuclear fusion in 1952. Controlled nuclear fusion--and producing more power out than goes in--that's another matter.
- Wiseacre 5y agoWe haven't even been splitting atoms for a full century yet. Give science some goddamn time to work.
- andrepd 5y agoNot to mention we haven't invested a fiftieth of what we should have into fusion research.
- sroussey 5y agoMore money goes to subsidizing almonds.
- vbezhenar 5y ago20 century progress was so crazy that people got all the kinds of unrealistic expectations.
- kaibee 5y agoNot reading the article in the link you're replying to is a heuristic that almost always works ;)
- orangepurple 5y agoThis is an excellent read, thanks for sharing
- NoGravitas 5y ago"Assuming anything Scott Alexander Siskin says is wrong" is also a heuristic that almost always works.
- DrBazza 5y agoIf only I didn’t do my phd in a department with plasma guys 30 years ago back in the 90s eh?
- soperj 5y agoIt was 30 years away at current levels of funding. Funding dipped considerably. This is as dumb as when they ask you for an estimate at work, then change the scope of the project and then retort with "well it was your estimate".
- twobitshifter 5y agoEven if it takes 200 years it is likely still worthwhile. Unfortunately we, as a culture, have a problem seeing and planning across generations. https://longnow.org/ https://longnow.org/
- m_fayer 5y agoI'm happy to hear someone make this point. Rather than sneering "vaporware" when decades fail to crack a problem, I would prefer us to keep in mind that if we never try multi-generational projects, we will never taste the fruit of multi-generational projects, and those are some sweet fruit indeed.
- kiba 5y agoI am not sure if we have a project that spans multiple generations? It would be easier to do a project that provide some kind of immediate benefit while having long term multigenerational long term effect. Science is kinda that way. We get immediate knowledge with long term unknown payoff.
- Out_of_Characte 5y agoThe Netherlands would like a word. Holding back the sea was a monumental project that will definitely last generations to come. Maybe not as exiting as high temp superconductors and fusion but still a nationwide unique product.
- HWR_14 5y agoIt may last for generations to come, but did the original development take generations?
- kuschku 5y agoOld-school land reclamation worked by putting woven fences in the water where waves would deposit sand over decades and centuries, slowly growing the land bit by bit. Those versions were already multi-generational projects. (I'm citing the techniques used in the north frisian wadden sea, I'm unsure if the same techniques were used in west frisia as well)
- willis936 5y agoThis is a disingenuous argument. JET was not designed as a power plant. Those flywheels are used exclusively to transiently power the copper confinement coils. Since superconduction was discovered, no reactor study has had non-superconducting confinement coils because they are (very obviously) impractical.
- SantalBlush 5y agoRemember, the easiest way for uninformed people to appear scientific is to lob high-level criticisms about the topic. In this case, it's "But it only doubled the output!" or something similar. We see this pattern on every single piece of science news that comes across the front page of HN.
- dfdz 5y agoThe only purpose of my comment was to provide context for the numbers in the article via a thought experiment involving doubling. Since the calculation involves taking log_2, even if the estimate for the "used power" is off by a factor of 4 the result will only change by 50 years. Can you make your qualitative comment quantitative and update the numbers in my doubling thought experiment? I would be curious about what you think is more realistic.
- willis936 5y agoA quantitative approach is no virtue here, just a more convincing way to lie. ITER will use roughly the same amount of power as JET, produce 10x the power 40 years later. Even using poor metrics such as Q you have enough data points to p-hack whatever incorrect timeframe model you want. ITER isn't even using HTS coils. If you really want a quantitative projection of MCF performance over time, here you go. Don't lie with numbers if you don't know what you're talking about. https://en.wikipedia.org/wiki/File:Fustion_triple-product_diagram_Horvath.webp https://en.wikipedia.org/wiki/File:Fustion_triple-product_di...
- Neil44 5y agoThat's a co-insidence, 2185 is also the year of Linux on the desktop.
- dghughes 5y agoOr it could be like Zeno's Dichotomy Paradox.