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.
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.
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 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.
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!!!"
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.
That’s one of the purposes of arxiv is to allow others to review or comment on the work.
(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)
https://www.scientificamerican.com/article/controversial-phy...
https://www.nytimes.com/2023/09/29/science/superconductor-re...
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)).
I think that given the subject, if this pans out there is no chance they'll get rejected.
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)
What is the pressures required for the effect? Are there other requirements that limit it's practical usage.
[1] pun definitely intended.
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?
https://en.wikipedia.org/wiki/Superconductivity#/media/File:...
Serious Q: Could a housing between racks/nodes within a POD in a large DC be holding connecting "plates" in a housing wrapped with AeroGel insulation such that horizontal connections could be passed through plate-connectors of these materials, cooled with piping just as we do with anything else in a DC, and allow for ~0 resistance horizontally between racks in a pod?
Or imagine a membrane that can be fed super cooling for connectors, such as that insane machine with 800,000 cores and the rubber connection membrane? (cant find link)
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.
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.
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.
Is that because neither of them have any electronic devices? They both have payphones and cars that don't look like today?
He also was very confused by the “Ronald Reagan? The ACTOR??” joke but immediately thought that BTTF2 Mogul-Biff was supposed to be Trump.
- it was a joke (not typically acceptable on HN)
- it is a dated reference
I decided it was worth it…
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.
Thinking of the 11 rubes who paid $10k each for their own uncontrollable hoverboards that only work over conductive surfaces.
https://www.kickstarter.com/projects/142464853/hendo-hoverbo...
Infrared photons are thermal energy. Thermophotovoltaics and thermoelectrics convert work due to the thermal gradient into electricity.
Laser cooling: https://en.wikipedia.org/wiki/Laser_cooling
Cooling with LEDs in reverse: https://issuu.com/designinglighting/docs/dec_2022/s/17923182 :
"Near-field photonic cooling through control of the chemical potential of photons" (2019) https://www.nature.com/articles/s41586-019-0918-8 https://scholar.google.com/scholar?cites=8589611114160282602... :
> 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.
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.
Ie. currently it can't be formulated as a search problem entirely on a computer.
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.
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.
So obviously if true there's a better explanation of why those early experiments failed.
i.e. LK-99 w/ sulfur contamination was originally blamed for the magnetism. This new variant is explicitly sulfur-based.
"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 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.