So cool! I'm sure there's a reason this hasn't been done, but I'm curious if there would be value in doing using this for some type of folding at home distributed optimization problem solving. Converging on actual suggest improvements that might not seem obvious