Hacker News Reader

Comment by floxy | Hacker News Reader

Lots of people are saying there should be a quick turnaround time here for trying the replication. I wonder how much it would cost to try and replicate this, in materials, labor costs, and vacuum furnace time/costs. My naive thought is that that labs in general aren't just going to drop everything to try an replicate every instance of someone declaring they've discovered room temperature superconductivity. Seems more like a "there's a lull in paying work, so hey, why don't some of you junior technicians try this out" thing? Anyone have insight into this? I suppose a university lab might have more leeway here to try out spur of the moment experiments? Maybe this is a "I'm personally interested in this and I'll work in the evenings to test it out, instead of on company/university time"? Or are the authors high profile enough in the community that it meets a certain threshold of credibility that it seems more worthwhile than the average?

View on HN

If i make an analogy to p=np.

If someone claimed to have a complex proof, i dont think people would stop and drop everything.

If someone claimed to have a constructive solution to p=np, claimed to have implemented it, and put the code on github - yeah i think lots of people would drop everything to run it and see if it works.

Im not a physicist, but maybe with the relative ease on making this material (according to other comments) it is more like the second situation.

For real; the original paper is kinda shitty in terms of data presentation, and there's significant value to make a follow up paper to create data and try to get your name out there.

The author of the paper apparently says it wasn’t supposed to be leaked and that it needs more work, but he stands by his results and will assist anyone trying to replicate, which sounds pretty steadfast to me.

They have patents… this is the sensible approach to take with your intellectual property well defended. Between the patents and the published paper establishing his role and thus Nobel eligibility…

At this point assisting anybody who wants to help prove him correct (and I know that I’m making the assumption that he genuinely believes he has made what he claims in the paper with the mechanisms and methodology as outlined) will just help secure get rich faster and make his Nobel prize even more secure.

Polycrystaline materials are kinda finicky so more samples helps eliminate the error bars. So obviously why didn’t they do heaps of samples and have a longer paper? … Money! They got funding for quantum physics based sensor/detector research and this appears to have been a happy accident, so naturally they kept their heads down, got their patents (they have multiple patents each of which costs money) and did a little bit more work to be absolutely sure, then as soon as it was smart, they published their results and are now going to let the entire world excitedly double check their results… because the results if true, are fucking awesome…

Of course the world wants to replicate this, no one wants to be waiting for someone else to confirm a breakthrough this big… I bet there’s even some corporate labs doing a little work on replication too, so some smart managers in various industries can factor room temperature superconducting materials into their designs and future plans sooner, because otherwise their competitors might be! (Obviously there’s not going to be a lot of this since most companies don’t have internal science groups but the Lockheed-Martins of the world, the Daewoos, the Mitsubishi Heavy Industries, the Ratheons … I’d be surprised if they didn’t already have someone somewhere in the world who is still currently making batches and putting them into furnaces because the material is cheap and you can get a lot more certainty from additional samples if your are prepared to pay for the work. I’d be dammed sure someone at General Dynamics is making some because it would massively help them convince the US Navy to give them more funding to continue work on railguns!

Fleischmann and Pons' famous cold fusion result was announced on March 23 1989 and a spate of replication attempts followed in the ensuing weeks; by April 30 1989, the NYT ran an article calling cold fusion dead due to the large number of reported negative reproduction attempts.

Scientists can move far quicker than you expect, and a positive or negative replication of this superconductor claim should be a very easy publication to grab. It would definitely be worth alloting a few weeks of grad student time to replicating this ASAP to get that publication.

The fact that there are other (competing?) scientists doing superconductor research suggests there will be organized attempts to verify it.

That said, I remember reading about shockley having ideas about the transistor that he only fessed up to as his other more open collaborators started understanding the phenomenon.

> My naive thought is that that labs in general aren't just going to drop everything to try an replicate every instance of someone declaring they've discovered room temperature superconductivity. Seems more like a "there's a lull in paying work, so hey, why don't some of you junior technicians try this out" thing? Anyone have insight into this?

Labs all over the world are dropping everything to replicate this. It's simple self interest.

It only takes 2-3 days and the people involved in the paper are credible.

If it's real, your lab would be at the forefront of the biggest revolution since the transistor. If it's real, there will be hundreds of papers on the topic by next year. Your lab could be the one to discover a variant that can carry more current, that could have patent on the material of the future power grid. The papers that provide the basics for how this works and the data that everyone will be using will get 100k+ citations in a few years. The PhD students that become experts in this will be in high demand everywhere.

It doesn't require the kind of equipment that you have to justify to accountants and run at full capacity. And while the reactants are things that I believe most labs won't have at hand, they are the kind of thing anybody can buy on the web to arrive this week.

Whether people will stop what they are doing to verify it is literally a matter of people stopping what they are doing. University labs are usually quick on that.

One of the elements used here is phosphorus (P) which is regulated by the DEA in the United States, and can't just be purchased online

If you are at a materials science or physics lab, it can't be difficult to get.

Hell, Sigma Aldrich will sell you LSD. Established labs have no trouble with lightly restricted drug precursors, they've done all the paperwork.

> One of the elements used here is phosphorus (P) which is regulated by the DEA

I would have figured ATF. What drug can you synthesize with phosphorus? (EDIT: Apparently meth [1].)

[1] https://www.federalregister.gov/documents/2000/09/25/00-2455...

Probably ATF, not DEA. DEA regulates controlled substances. ATF regulates explosives.

Wonder element phosphorus can be used for both!

Seriously… elemental phosphorus is controlled by the DEA? The ATF I could maybe understand but the DEA… that’s just whacky.

From what I recall of the original paper, 1-2 hours of actual labor with equipment and reagents many labs already have on hand--although it required 24-48hours in the furnace at each step which is the only reason for the delay and why the article author mentions we should see results tomorrow. So, trivial enough most people in similar labs already spent more time discussing the implications than actual labor/investment to replicate.

If it really is as easy as you say then surely every YouTuber with a metal furnace dropped everything to make that 100M view video.

I'm hoping they are. I want to watch them this weekend.

Finally found someone who's liveblogging it: https://nitter.net/andrewmccalip

It's a reasonable enough paper, and replicating should be pretty easy (an undergrad could do it). Because the implications are so massive if true, people are going to jump on it.

From the linked article:

> You can bet that furnaces in solid-state materials labs around the world have been cooking yesterday and today to try to reproduce its synthesis and the properties, and we should be hearing about the results of these experiments very soon. The first samples should be coming out of the quartz vessels. . .sometime tomorrow, perhaps? Depends on what was available around the lab!

Yes, that quote from the article makes it seem like everyone is dropping everything and jumping right on it. I'm wondering how likely that is we'll find out this Friday vs. in the next couple of months, as people get around to it.

Friday is more plausible than a couple months. If the radio silence ends up extending well-into next week, people are having trouble replicating and want to cross their ts before calling it out as fraud.

It's an easy thing to attempt to replicate, and it's probably _the_ single biggest goal in a particular field of science in at least decades.

It's impossible that at least some people didn't drop everything to try it.

That depends on how easy it is to replicate, and if it is true. If it is easy (the 24 hour in the furnace is really long enough for everything to get to temperature - 10 minutes is enough in ideal cases) then we will get lots of success Tomorrow. If it is hard, well we will get a lot of failures, but a few will try again an in a month as they refine technique will report a success.

If it is false we may never get a final report, just everyone gives up on reproducing it. (though with the hype we may see it like cold fusion were a few not very good researchers keep it alive via badly done experiments).

Only time will tell.

I’m sure this is on Applied Science’s (a YouTube channel) radar. He recently went on a deep dive on making YBCO superconductors so it’s right up his alley.

Would love to see Ben Krasnow from Applied Science try to replicate this!

https://nitter.net/BenKrasnow/status/1684431462643437569#m

A livestream of this would be awesome

It’s all joever by now.

Why? I see a livestream happening here: https://www.twitch.tv/andrewmccalip

Under a week of work to publish potentially the highest impact paper of your PhD (especially if you're first)? I'm guessing most students would take that chance.

Not much, see article (~1 day), or article + thread from yesterday. https://news.ycombinator.com/item?id=36864624

I don't think, at least generally, work is as rigid as indicated. At least at white collar jobs, and especially a research lab, and most especially a university research lab.

$10, $100, $1,000? 1 hour, 10 hours of labor?

If you have the equipment, some $100s on consumables, some 10 hours of labor, and some 60 hours of waiting.

Just testing the samples already made in an independent laboratory would be replication enough!

The synthesis outlined in the paper takes about 90 hours, but basically all of it is just waiting for the material to bake in a furnace.

I'd expect some lab out there to report a result by the end of the week. The synthesis really is super simple, just grinding powders together and baking in a furnace. Granted, the final product has to be baked under vacuum, but that's not super difficult to achieve. The authors sealed the material in a quartz tube and baked in a standard furnace.

I think if it isn't replicated by tomorrow, it's dead.

Reply on news.ycombinator.com