Comment by bartlettD

Link to the actual paper for those interested.

https://www.nature.com/articles/s41467-024-48534-4 (Open-Access)

I had wondered why they weren't doing this in simulation but didn't realise that Finite Element Analysis gets weird in the regions they're exploring.

Reminds me a lot of the "Chemputer" concept for automated chemical synthesis expriments. https://www.chem.gla.ac.uk/cronin/