Super productive software week! Wish I got a bit more done on the admin and strategy side, but we had some pressing manufacturing/voro processes to improve. First, the voro work I did last week improved things, but in testing them more robustly, it wasn’t as consistent of a benefit as we need it to be, so work continued to drive up the density, drive down the number of links that need to be adjusted for manufacturing, and get as close to hex close packs as possible. We had a great meeting with a group that has a lot of experience developing medical devices, and may work with them on ours, but it’s early days there. The new voro search space traversal technique is absolutely the most robust we’ve ever had, and the models really speak for themselves. I think we’re approaching the local maxima for this whole paradigm, which feels very exciting.
September 4, 2026