Data-Centric AI

The more time I spend thinking about robotics and multimodal AI, the more I suspect that one of the biggest long-term bottlenecks is not just models. It is data infrastructure. A lot of progress in…

One thing I’ve learned working with real-world multimodal datasets is that “having data” is very different from having data you can actually train on well. A dataset can be rich, interesting, and…

One thing I’ve learned working with real-world multimodal datasets is that “having data” is very different from having data you can actually train models on. The original SCOUT dataset contains rich…