3 ms·

I had a teacher that in the 80s or something made an experiment where he would drop lots of elements/composites that were "good candidates" in a melting pot and

I had a teacher that in the 80s or something made an experiment where he would drop lots of elements/composites that were "good candidates" in a melting pot and hope that some superconductor material would come out from it. He made some important discoveries there but no "room temperature" superconductor came from there. I wonder how AI could be used in the "same way" with data and much less effort.

hinkley 4y ago
It's a bit of a mystery to me too. Are we really sure we understand how superconductivity happens? At one point there were theories, materials that didn't match the theories, and new theories to try to explain the anomalies. But I don't know how that all played out in the end. If we actually understand the process then you could simulate the materials, the way we do fluid dynamics or protein folding. But without that understanding we could end up flagging the eventual winners as unlikely and reduce progress instead of increasing it. And if we can do that, why haven't we used the same thing for battery chemistries? Give me all of the low-dendrite battery formulas sorted by simulated capacity, weight, longevity and engineering difficulty, please.