In a warehouse in San Leandro, California, Encord, a company that builds data tooling for AI models, is running a trial with Zander Labs, a German neuroscience startup, to see whether measuring brain activity during physical tasks can create a richer dataset for training robots. Vineeth Velmurugan, Encord’s head of robot learning, estimates that breaking through the data bottleneck will require a dataset roughly five times the size of YouTube’s video corpus. The headset worn by Encord’s pilot, Andrew Ceja, includes sensors that measure his brain waves as he disassembles a Jenga tower. Lucas Gehrke, a Zander neuroscientist, says the amount of brain activity used during a task offers clues for model builders. The goal is to build an initial brain-wave-tagged dataset, run it through customer robotics models, and evaluate performance. Encord is also experimenting with sensors on the forearm to detect electrical signals in muscles, aiming to build a 3D depiction of hand position. Velmurugan estimates dense annotation is worth 100 times as much as ‘junky ego data’ for training specific tasks, though it costs about 20 times more to produce.