Good morning CT. AI has learned from billions of images, videos and text. But there's one thing the internet can't fully teach it: how the physical world actually looks and changes in 3D. A robot doesn't just need to know that a chair exists. It needs to understand where the chair is, how far away it is, what's around it, and how that environment changes. That's where @vangrid_io comes in. VanGrid turns everyday smartphones into a distributed sensor network. People capture real places from multiple angles, and VanGrid reconstructs that footage into structured 3D data instead of simply storing raw video. The interesting part? No expensive dedicated hardware. No giant centralized mapping fleet. Just phones contributing real-world ground truth. The bigger bet is simple: more captures → more spatial data → better perception for AI and robotics. Still early, but I think the idea is worth watching.