[smusamashah]

> That is what makes the finding so striking. Manganese is usually not viewed as a friend of stainless steel corrosion resistance. In fact, the prevailing view has been that manganese weakens it.

> "Initially, we did not believe it because the prevailing view is that Mn impairs the corrosion resistance of stainless steel. Mn-based passivation is a counter-intuitive discovery, which cannot be explained by current knowledge in corrosion science. However, when numerous atomic-level results were presented, we were convinced. Beyond being surprised, we cannot wait to exploit the mechanism," said Dr. Kaiping Yu, the first author of the article, whose PhD is supervised by Professor Huang.

This is the Cannot be explained bit

[MSFT_Edging]

> which cannot be explained by current knowledge in corrosion science

The three stooges effect I see. Too many corrosive elements, they stop each other from getting through the door.

[wnevets]

I was thinking of the Mr Burns sickness bit which references the stooges [1]

[1] https://www.youtube.com/watch?v=aI0euMFAWF8

[MSFT_Edging]

I think people are taking this seriously but in reality I'm just referencing this bit.

[janderson215]

This is great. Any other dynamics in which The Stooges Effect is present?

[PowerElectronix]

High entropy alloys come to mind. There are so many possible metallic combinations when you mix 5 o 6 metals in equal proportions that they all end up being in a single, homogeneous combinantion with about one atom of each per crystal molecule.

[chamomeal]

Oh wow I thought this was a simpsons reference to when Mr. Burns has too many diseases, that none of them can take him down. Cool that it's a real thing

[adrian_b]

A bonus is that manganese is one of the cheapest metals, so this method for increasing the corrosion resistance of stainless steel in salted water and oxidizing conditions is very inexpensive.

[hcknwscommenter]

Will be interesting to see if, in hindsight, there are clues to this in the preexisting literature that we just hadn't quite recognized.

[nradov]

This is one of the areas where AI will accelerate scientific research, by scanning through the journal archives to make connections that no one had noticed before.

[alwa]

See, my thought would have been the opposite: in a situation like this—where nobody tries the thing because “everybody knows” it’s counterproductive—I’d expect AI literature surveys to confidently assert received wisdom.

It sounds from the quote like even the researchers thought it was a mistake at first… and that on the basis of the literature PLUS their collective professional wisdom. Now, obviously, they did in fact try the thing, so maybe the idea was not quite so wacky as they paint it for the article.

But the point feels similar here as with LLMs and writing: they can do what’s come before pretty well, and they can exhaust a well-specified problem space through sheer muscle; but they seem to be less good at evolving the frontiers of the domain, and I see no mechanism by which to expect that to change.

So I tend to take the opposite lesson: surprises like this renew my hope that there will remain a place in science long into the AI era for meatbags and serendipity and the spirit of curiosity.

[nradov]

LLMs don't rely completely on received wisdom. The training process works in a higher dimensional space so some data points that might seem unrelated to humans end up being clustered close together if there is a hidden or unrecognized relationship.

[alwa]

True, but the clusters do come from somewhere—namely the training set and the input. The more technical the literature, the sparser the prior work; the more answers depend on labwork, the less the bottleneck is purely symbolic reasoning or data-retrieval-at-scale.

To hear it from the researchers, this feels like the sort of finding that, even in retrospect, is non-obvious from existing literature.

I remember hearing scientific progress described in terms of punctuated equilibrium: some Big New Idea, then a bunch of work generalizing that new idea to the rest of the problem space. I could see AI tools speeding up the second type of work: taking a new framework and chewing through everything that came before, in that new light.

But I have a hard time thinking about how the AI techniques could produce novel, surprising outcomes like this one—ones where it’s not just a permutation of existing knowledge, but where it turns out reality actually cuts against the accumulated written knowledge that came before. The “there is magic to be explained here!” aspect of science.

[stevenhuang]

> But I have a hard time thinking about how the AI techniques could produce novel, surprising outcomes like this one

The comparison is not equivalent as a human isolated from the environment and unable to perform experiments would fair the same.

It is through conducting experiments that we make discoveries.

Once AI can hypothesize and run experiments from start to finish I see no reason why novel discoveries won't be made this way.

[ccimmergreen]

Yes, that's only if AI would ask you to validate all prior assumptions to avoid being led by a false premise. I don't see AI or humans bothering to do that.

[abduhl]

>> Mn-based passivation is a counter-intuitive discovery, which cannot be explained by current knowledge in corrosion science.

This statement sounds like the type of language one uses when trying to get a patent.

[adrian_b]

The part with "cannot be explained by current knowledge" is an exaggeration, because the abstract of the research article explains it very well, based on the current knowledge.

However, "a counter-intuitive discovery" is true. Manganese is frequently used in stainless steels, but only as a cheap substitute for nickel, when this is considered as giving up the superior resistance to corrosion provided by nickel in exchange for the low cost provided by manganese.

The counter-intuitive result of the research is that there are circumstances when manganese provides improved corrosion protection, not only a lower cost.

The reason why this has not been discovered earlier is that manganese alone does not protect against corrosion, but only in an appropriate combination with chromium, when chromium protects both the steel and the manganese at lower electric potential differences, while manganese protects both the steel and the chromium at higher electric potential differences.

[colechristensen]

A paper is hardly worth publishing if "cannot be explained by current knowledge" isn't true.

It is, however, incredibly tacky to talk about your research like this.

[wolfi1]

you should know how to spin it. Take Müller/Bednorz for example, their paper just said "_Possible_ high Tc superconductivity in the Ba−La−Cu−O system" and they got the Nobel Prize in Physics one year later