Xinghai Map co-founder Gao Jiyang proposed a new business model for embodied AI, shifting from selling hardware to selling "physical world tokens." This model envisions robots as low-margin hardware carriers to deploy more terminals, with the core value migrating to the embodied AI's "brain." Similar to how OpenAI charges for text tokens, future embodied AI giants will profit from each effective action or task completed by a robot in the physical world. Xinghai Map is developing its embodied AI brain using a unified autoregressive architecture to handle zero-shot generalization, general grasping, and long-range tasks, aiming to reduce training costs and increase efficiency. AI
IMPACT This new model could shift the industry focus from hardware manufacturing to AI brain development, potentially lowering hardware costs and increasing robot deployment.
RANK_REASON The article discusses a proposed future business model for embodied AI, rather than announcing a new product or research breakthrough.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
