Are brain waves the next unlock for physical AI?
Startups are betting that the next bottleneck in humanoid and warehouse robotics won't be model architecture but the scarcity of real-world physical training data, and some are now going as far as measuring workers' brain waves to generate richer datasets. Encord, a company based in San Leandro, California, that traditionally built data-annotation tools for AI, is trialling headsets from German neuroscience startup Zander Labs that track brain activity — indicating states such as error, intent and surprise — while human "pilots" perform physical tasks like disassembling a Jenga tower, in the hope this produces more useful training signals for robot-learning models.
Encord's head of robot learning, Vineeth Velmurugan, a veteran of OpenAI's robotics lab and Berkshire Grey, says the scale of data needed is roughly five times the size of YouTube's video corpus, and that such data "simply does not exist" yet for physical manipulation tasks, unlike the vast text corpora that trained large language models. The brain wave work is currently just a trial: Encord plans to build an initial tagged dataset, test it against customer robotics models, and only then decide whether to scale it up. More broadly, robotics firms are sourcing training data through two main routes — "egocentric" video captured by workers wearing cameras, and remote-operated robot rigs — with Encord using its San Leandro facility to experiment with new approaches, including brain-wave tracking and skill-specific data collection.
- Encord tests brain-wave headsets to improve robot training data quality.
- Real-world physical data, not model design, is robotics' key bottleneck.
- Needed dataset scale estimated at five times YouTube's video corpus.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Proponents argue that physical AI systems are fundamentally limited by the poor quality of existing training data, since video alone cannot capture the intention, anticipation, and fine motor planning that precede human movement. Brain wave recordings could offer a richer signal that links intent to action, potentially accelerating robots' ability to learn dexterous, context-aware behaviour far faster than passive observation allows. From this view, pursuing new data modalities, however unconventional, is simply good scientific practice when the goal is building AI that can genuinely understand and replicate embodied human action.
The case against
Sceptics caution that neural signals are noisy, highly individual, and difficult to standardise, meaning the practical benefit for training generalisable AI models may be far smaller than enthusiasts suggest, while the hype risks outpacing the science. There are also serious concerns about the privacy and consent implications of harvesting brain wave data, given how sensitive and potentially revealing such information is compared with video footage. For these critics, resources might be better spent refining existing data pipelines and annotation methods rather than chasing a speculative new frontier with uncertain returns and significant ethical exposure.