Some background: there were two Chinese teams publicly pursuing LK-99-derived room temperature superconductor, which I arbitrarily named "north China team" and "south China team". North China team was headed by Hongyang Wang (who lives in Beijing) and south China team was headed by Yao Yao (who lives in Guangzhou). They used different synthesis and different analysis, i.e. north China team used hydrothermal synthesis and used SQUID measurement, while south China team used solid state synthesis and used EPR measurement.
This is a joint paper of both teams. They reproduced results of each other (this is unclear in the paper, but stated in their behind-the-scene posts) and measured a clear sign of superconductivity. It is "near room temperature", because they are sure about 250 K (hence "near"), but not sure about 300 K. As for "possible", the behind-the-scene post makes it clear it is false modesty.
If you are interested, you definitely want to read behind-the-scene posts. Read them here: https://www.zhihu.com/question/637763289 (they are in Chinese). Hongyang Wang is 真可爱呆 and Yao Yao is 洗芝溪.
> As for "possible", the behind-the-scene post makes it clear it is false modesty.
It's not false modesty to withhold a conclusion that is unwarranted without more evidence. The last time this happened there was also "possible Meissner effect" that turned out to be diamagnetism.
There are no downsides to being conservative until more evidence is acquired.
> There are no downsides to being conservative until more evidence is acquired.
There are, but I'd agree that there aren't downsides we should be concerned with. The downside is that if you're trying to publish the work that it can make it easier for the work to be rejected. I'd agree this is dumb, but it is a thing I've seen happen, and be not too uncommon. Just comes down to metrics: academics are judged by citations and number of papers published, thus papers are written to reviewers as opposed to peers (not necessarily the same thing, but assumed they are), also incentivizes flashy results to generate more publicity, or overselling the novelty of work (sometimes even by mistake). But otherwise I agree, and I think it should be encouraged to take a more tempered approach (I think it'll also really help build back social trust in sciences (again, small part of a larger pie)).
For this particular example, sure. I mean for this particular subject you'd have to hand in a really bad paper to get rejected, like past unintelligible, because the breakthrough is big. But I thought we were talking in general. There's also very clear downsides to this particular example by not being exceptionally cautious, because even a hint of hype can make them look like incompetent fools if they aren't correct. Though that wouldn't be good either because they clearly aren't. It's easy to make mistakes in science and even as I frequently told my friends last time, there was something different here and it'll probably end in failure (which it did) but that we'll learn something important (which looks likely to be true given this paper).
I was reading the paper and thought "Okay, these folks took the LK-99 lesson to heart about being overly enthusiastic in your experimental observations." :-)
Given that the LK-99 theoretical underpinnings were pretty solid (if not the implementation) I've been watching papers in this space. (on google scholar you can just track papers that cite the LK-99 paper as new follow-ups)
>There are no downsides to being conservative until more evidence is acquired.
That's true, but one upside here is that this is not from the guy at Rochester who already got burned twice for publishing false superconductor discoveries. This is at least a report that can't be dismissed immediately.
I don't know who you're referring to "from Rochester", but the excitement from last summer around LK-99 came from two South Korean researchers. Which, mind you, all turned out to be faulty interpretation of their data.
To be clear, I don't mean to cast unnecessary shade on these current results, I'm just saying the sane and prudent thing to do, especially given LK-99's recent history, is to hold off on any champagne popping, and that appropriate restraint shouldn't be characterized as "false modesty".
> Which, mind you, all turned out to be faulty interpretation of their data.
Which we should be critical of, but understandable, since that's one of the main reasons to publish (and why I think it is weird we say arxiv papers aren't peer reviewed. That was probably one of the most peer reviewed works in the last decade) and since we're all human. Science is full of mistakes, and is unsurprising when a lot of it is literally trying to do things that humans have never done before (much more to science than this too).
What I thought was really cool about LK-99 is that it isn't too often that people get a first hand look at what goes on inside the science communities. An abnormal amount of attention and openness, but illustrative. Just not sure this is the takeaway people got. But I saw science working in action, and it was really cool.
> So why do they publish if they're not 100% confident ? Why not wait to be certain ?
There are very good reasons to publish before one is "100% certain". It can get lots of other scientists to evaluate your information where they can try to reproduce it or poke holes in your theories. It just makes sense to do that with a tone of "We got some interesting results..." as opposed to what happened in the summer which was more like "We've made one of the biggest discoveries of mankind!!!"
Yeah, there’s a very good argument I’ve read from researchers that instead of papers we should be publishing continuously what our results are to encourage this early and often community interaction/feedback (ie more like code reviews). The reason it doesn’t happen is publish/perish + needing to be first to publish (ie someone taking your work and beating you to the punch and getting all the credit).
Getting other people to inspect your work sounds great for you but it costs those people a time and money.
So, the incentives for misuse are really high and the net result is that shortcutting the process is a net drag on progress. People remember the bigs stuff like FTL neutrinos, cold fusion, NK-99, etc but arguably this also shows up as part of the ‘reproducibility crisis’ in many fields.
People argue about whether putting stuff on arxiv constitutes "publication". I like the loose definition of publish (it's available for a wide audience to read, with at most, pay barriers) but other folks do not. There is still a huge cachet in being able to point at a peer-reviewed paper in a prestigious journal, no doubt, and many people will consider arxiv a preprint server, not a publication.
(in the old days, after your article was accepted, but before it was published, they'd send you a few copies labelled "preprint" that you could distribute informally)
"The first law of superconductivity: stay away from theoretical physicists." What should I say, the rules are all used to be broken?
I haven't been drunk for many years. Last Friday, I always stayed there to test and kept sending me photos and live broadcast the real-time measurement results. Every time I sent one, I couldn't help drinking a drink. I was directly broken. I was carried back by the students. It's embarrassing~~
This is LK-99 derived? Now I'm even less interested. Because if you believe this one, you need to believe that despite LK-99 being bogus, that somehow, trying to make a new LK-99 variant was lucky enough to find the one compound out of countless attempts that winds up working - as opposed to coming out of a completely different, still viable, line of research. Stranger things have happened in the history of science & technology... but not that many.
The hard part is that people have already tried really hard to make and measure a superconductor in the LK-99 system, and these researchers have once again gotten close, but not succeeded. There is some precedent for this: one of the highest-temperature superconductors is two-dimensional iron selenide supported on strontium titanate, which superconducts at 100 K, while the bulk iron selenide superconductivity is a measly 8 K at normal pressure (38 K under pressure). At this point, the most plausible way that superconductivity could be occurring in the LK-99 system is if it's in a metastable or nanostructured (possibly two-dimensional) phase that doesn't like to or can't exist as a uniform bulk material.
That scenario, if true, would surpass any other in the history of science. It's akin to the Monty Hall problem, but with a trillion doors, where you don't change your choice after all but two are shown not to have the superconductor.
So obviously if true there's a better explanation of why those early experiments failed.
Yes this is LK-99 derived. (See the paper's reference 4 and 5.) Eh, of course it is unlikely LK-99 is bogus and this one is not, but then the correct conclusion to draw is that LK-99 is not bogus?
Perhaps LK-99 is bogus, and they were triying to reproduce the result, they failed succesfully and they got another similar compound that is not bogus.
That’s possible but far less likely than the alternative explanation that either this is bunk too or LK-99 contained signal that people dismissed due to a hole in our methodology.
Wait how are all the flaws in the paper and all the failed reproductions of LK-99 and all the signs that LK-99 is just diamagnetism compatible with LK-99 being a proper superconductor?
The thing is.... all the failed reproductions of LK-99 were based on the leaked version of the paper, which according to the authors is missing some information/steps for production. They said they would publish the complete/final version of the paper in 2024, waiting to see that.
Jokes like that are going to get a lot of resistance. Oh well, time to amp it up.
Best chart I could find. Looks like there's another at 250 K from 2019. But what I don't have here is the Temperature / Pressure / Timeline, though...? What is the Pressure in the article? Is it STP?
250 K at ambient pressure is still revolutionary, as it can be reached by dry ice. The highest critical temperature at ambient pressure had been something like 150 K: still above boiling point of liquid nitrogen, and won Nobel Prize in Physics in 1987.
I grew up in Minnesota, and I have to laugh that 250 K is -9.67 F, and yeah, we had a lot of days like that. It's amazing to think that future childhood toys could be superconducting outside on those cold Minnesota days.
Recently shared the bttf series with my 11yo. The third doesn't hold up too well, but the first two are great. The funny part was that she saw the 80s and the 50s as basically the same.
I think it's an issue of the fact that they're both "before my time". I treat the 20s and 50s as the same, I wouldn't be able to tell you any difference, even though they must be massively different.
Yeah, they both just looked "old fashioned" to her. Obviously she could pick out some differences, but when they first went back to 1955, she didn't really notice that the cars, outfits, signage, etc. were particularly more dated than they had been in the 80s version.
Movies of the 80s and 90s feel a lot more modern compared to the “old movies” we had living in the 80s. Black and white movies with cardboard acting sure felt quaint for the tv generation.
Today though I’m constantly surprised by the number of young people who recognize things from 80s movies and especially music. I’d say that number is higher compared to our generation.
The 80s and 90s produced a lot of movies that became pop culture classics (Back to the Future among them).
Millenials and Gen Z grew up in an era with much easier access to older media than previous generations. First was the video store - while Gen X had this too, it really took off in the 90s. I remember when I was a kid in the late 90s and early 2000s, it was $5 to rent a new release or 3 for $5 for old releases. This meant that we were basically encouraged by our parents to watch older stuff, and of course the fact that they lived through the 80s themselves meant they tended to recommend movies to us from that era.
Of course, after the video store came VOD services like Netflix. Old movies are a great way to pad out a VOD catalogue, so that increased the access to 80s/90s movies even more.
It also doesn't hurt that, as you've pointed out, many of these films still hold up pretty well today.
We started to get a lot of the classic movies in the 70s (Taxi Driver, Godfather, Star Wars, Halloween, Blazing Saddles, Rocky, Alien, Clockwork Orange, Exorcist, Jaws, Apocalypse Now just to name a few) and the 80s went absolutely wild and - particularly - far more broad. The kids films from the 80s didn't really exist before then (outside of Disney).
In the 50s/60s there were less (but still some of per personal faves) and the dominant genres (Westerns particularly) have been out of fashion for at least 40 years now.
In my experience jokes usually do well on HN. Memes and catchphrases don't. That's the main difference. Once something becomes a meme it's not actually funny anymore, but you'll still get upvoted for it on a place like Reddit because you're signalling that you're part of the "in" group.
As you can see, it's not actually that fun to ride. It hovers in every direction, like standing on an ice cube. The reason ice skates and roller blades work is they have low friction only along one dimension, so you can still apply force to the ground along the normal vector.
One would think with all of our crazy AI and supercomputers and quantum computers that a team would give it evolutionary goals of just trying simulations of molecular combinations to reach superconductivity. Sure, it'd be one thing to make it in a computer, and making the materials in the real world is quite another but I'm kind of shocked no one has come forward with something yet. I saw simulations of whole viruses running on a cluster of computers where they test drugs out and how they interact with the virus and simulated human cells so one would think its something with enough effort would be possible?
This is done across many disciplines to try and aide in new discovery paths. Typically you’re limited in exactly what you can simulate and often times solution candidates may be found that are impractical, currently impossible, or perhaps actually impossible to produce. Sometimes you can add search constraints to tie simulations together to narrow down such false positive solutions found but not always. Heck in some cases it’s literally cheaper and more accurate to do the bench science no matter how alluring virtualized renditions may be.
Most fields are still left with piles and piles of potential solutions to sort through. They often select candidates that are the cheapest and most practical to approach or they have high suspicion of success and pursue those. At the end of the day though we don’t have full universe simulators at every scale we’d want, we have very specific area simulators within very specific bounds. You have to go out an empirically test these things.
But this is and has already been going on for decades across most disciplines I’ve interacted with, they just weren’t using DNN or LLMs at the time but domains are adopting these as well to leverage where feasible in the search process.
I work with a variety of people interested in leveraging simulation and everyone wants to take the successes they see in LLMs or say RL from AlphaStar or AlphaGo and apply them in their domain. It’s alluring, I get it, the issue is that we often lack enough real understanding in domains and the science isn’t as airtight and people think it is, its too general or narrow, or on some cases we have good suspicion of how to build better more accurate simulations but there’s not enough compute power or energy in the world to make them currently practical, so we need to take some tradeoffs and live with less accurate and detailed simulation which leads to inaccurate representations of reality and ultimately inaccurate solution suggestion candidates.
Supercomputers (the classical kind) typically don't want to run codes that try lots of combinations; they were designed for, excel at, and cost a lot because of, the need to speed up single runs at a time. This is partly due to history, and partly due to definitions of supercomputers, but every time I proposed runs like this (proteins/drugs) to the supercomputer centers, they told me to bug off because my codes "only scaled to 64 processors" (that's 64 servers, mind you; before SMP was common).
We did what you described using idle cycles at Google (search for "Exacycle") and we got great results doing large scale parameter explorations (either randomly sampled, or sampled based on where the previous sims suggested looking next)- although, nobody actually did material simulations like this, we did proteins.
Realistically, almost nobody does this because it's just not cost effective (the search space is too large, the loss functions aren't accurate enough, and it uses TONS of energy), and more importantly, somebody else is just going to find a way to generate 75% of the results with 25% of the energy, and that person will get published faster.
You far overestimate the state of the art, and even our basic understanding of what superconducting is mechanistically. Simulating a single atom, alone in the universe is still a struggle not quite achieved.
Isn't the whole point that it's something we might not predict from what we understand about the materials so far? Why would it be likely that a simulation would do better than theory at predicting unknown experimental results?
You might hit on some interesting interactions between known properties that haven't been investigated but I would assume the real interesting results are from things we just don't know to model, or how to model.
There's plenty of researchers working on material simulation on a molecular level. It's not easy to just search millions of possible combinations and accurately predict their behavior
I believe that the fundamental physical particle interactions are not yet well enough understood to make a precise simulation, even if you had a supercomputer.
Ie. currently it can't be formulated as a search problem entirely on a computer.
My basic kitchen freezer can do -25. And that is a very basic and cheap freezer, so I think it would be relative easy to push it a few more degrees down.
To be clear, you mean a freezer meant to contain food. Laboratory freezers can easily go lower without being remotely exotic (in design or construction).
> This demonstration of active nanophotonic cooling—without the use of coherent laser radiation—lays the experimental foundation for systematic exploration of nanoscale photonics and optoelectronics for solid-state refrigeration and on-chip device cooling.
This is a joint paper of both teams. They reproduced results of each other (this is unclear in the paper, but stated in their behind-the-scene posts) and measured a clear sign of superconductivity. It is "near room temperature", because they are sure about 250 K (hence "near"), but not sure about 300 K. As for "possible", the behind-the-scene post makes it clear it is false modesty.
If you are interested, you definitely want to read behind-the-scene posts. Read them here: https://www.zhihu.com/question/637763289 (they are in Chinese). Hongyang Wang is 真可爱呆 and Yao Yao is 洗芝溪.