Hacker News Reader

Comment by jonny_eh | Hacker News Reader

Makes sense, on the other hand, the recipe is apparently fairly simple, so why would it be so hard to replicate?

It's posts like this that make me think of Roosevelt's Man in the Arena speech. It's easy for folks on the sidelines to criticize and say "what's so hard?" It's for the best who the folks rolling up their sleeves, doing the hard work don't see/hear it. There are tons of teams working at this from multiple angles, both trying to prove and disprove it.

It's only been a few days! The fact that cutting edge science from a pre-release version of a paper isn't going perfectly doesn't mean much. Give them time. Eventually we will have more solid answer.

I am a material scientist. Everyday, I see failed attempts to replicate samples. Slightly different compositions, slightly different heat treatments, subtle differences in crystallographic texture or the arrangement of inclusions that unexpectedly change the properties of your material in mysterious ways.

And I know nothing about superconductors, I am talking about steel, a material that we have worked with for centuries, it's in the first chapter of every textbook and is practically everywhere.

Things are hard.

YES! This is exactly my experience. Even if I do stuff myself it is often very hard to get to consistency, there are almost always more variables in play than the ones that I'm initially aware of. I've seen seasoned pros driven to despair by things that 'worked in the lab' but that they could not replicate outside of it for want of a simple oversight. Sometimes the line between success and failure is hair-thin, you can be almost there and never realize it.

Because if it is superconducting they only have tiny tiny superconducting chunks in a much larger sample that isn't. A couple of superconducting grains of sand in a rock, basically. Separate them out, try to test them on their own while ignoring all the rock parts, well that's difficult.

I suspect that what they actually have is some incredibly tiny superconducting crystals suspended on a material that is almost identical except for the placement of a few atoms. Very hard to work with.

Or it's just a regular diamagenetic material.

The unit cell apparently has:

1 Copper atom 25 oxygen atoms 6 Phosphorous atoms 9 lead atoms

There are many ways to arrange these constituents.

It’s hard to create the environment where they want to go exactly where you hope over a large volume. Sometimes this level of precision requires atomic control of layer growth and extreme temperature management. That’s why the methods list days between steps to get the atoms diffusing, and that’s for a few micrograms of material.

I guess whoever figures out a reliable process for scaling up production gets a second nobel prize?

or just billions upon billions of dollars.

I should patent a process then figure out the details later, or sue anyone who does.

Thankfully the patent system isn't that broken.

The steps are simple, but it seems to be quite finicky and there are almost sure to be sensitive to important factors that we don’t have clarity on yet. If it turns out to be definitively shown to be real

so simple that original authors fail 9 out of 10 times it seems

It is simple as in doesn't require expensive materials or tools or a particularly long process. It can nevertheless be tricky to execute correctly and depend on luck

It's too simple - it's simplistic. It doesn't control for some (yet unknown) things that make it fail or succeed, which makes it hard to replicate.

It seems to be, so far, a stocastic process in such that they don't now what's causing it in the 1/10 times it works afaict.

The recipe for Shakespeare is English, paper and a pen.

The recipe IS easy... Relatively, it's just not reliable/high yield

Simple is not synonymous with easy.

I dunno, but if it's so simple, why don't you try and then let us know if it's actually easy or not, as well as the results you get!

Reply on news.ycombinator.com