Unitree CEO Wang Xingxing said the robotics industry is edging toward a "ChatGPT moment" where robots can act on simple voice or text instructions [1].
This potential breakthrough would allow humanoid robots to operate in unfamiliar environments and perform most tasks without extensive re-programming. By integrating large-language-model AI, robots could process natural-language commands to adapt to new settings in real time [1, 2].
Wang, the founder of the Chinese humanoid-robot unicorn, said the shift is a fundamental change in how "robot brains" function [3]. Current robotics often require rigid coding for specific actions, but the proposed evolution would enable a more fluid interaction between human intent and machine execution [1].
Despite the optimism, Wang provided varying timelines for when this level of capability will be fully realized. In one instance, he said it could still take up to five years [4]. However, in another discussion, he said it could be 10 years before the industry reaches that milestone [5].
The development is centered in Beijing, where Unitree operates as a leader in the global humanoid market [3]. The goal is to move beyond pre-programmed movements toward a system where a robot can hear a command, and determine the necessary physical steps to complete the task independently [1, 2].
Wang said, "We are edging towards a ChatGPT moment for robot brains" [3]. This suggests that the intersection of generative AI and physical actuators is the final hurdle for widespread humanoid utility [1].
“"We are edging towards a ChatGPT moment for robot brains."”
The transition toward natural-language control represents a shift from 'automation,' which follows a set script, to 'autonomy,' where a machine interprets goals. If Unitree and its competitors achieve this, humanoid robots could move from controlled factory floors into unpredictable domestic and commercial spaces, significantly lowering the technical barrier for human-robot interaction.
