My favorite part about these class of materials is that the computational cost to design them is so high. We finally have computers powerful enough to design o

My favorite part about these class of materials is that the computational cost to design them is so high. We finally have computers powerful enough to design optical metamaterials with enough precision that they are a reasonably affordable thing to make (each real experiment in making these is on the order of $100,000, and the design space is so large >10^10 that we could never reconcile the two without in-silicon experiments).