I wish they talked more about the architecture though. I don't really understand how the information actually flows, in particular how they get all-to-all connectivity and how the output is measured at room temperature (no need for photon number resolving photodetectors?).

Might need to read the supplementary part. Anyway here's the Nature paper for those who didn't find the link: https://www.nature.com/articles/s41586-025-09838-7.epdf?shar...