March 30, 2026 – At the 2026 China Internet Media Forum, Wang Xingxing, the founder and CEO of Unitree Technology, delivered a captivating speech titled “When Robots Take Over Our Screens”.
During his talk, Wang addressed the current state of embodied artificial intelligence (AI), stating that it has not yet reached its tipping point. He went on to offer his personal definition of what he calls the “GPT moment” for embodied AI.
He illustrated this concept by describing a scenario where a robot, when brought into an unfamiliar environment, can successfully complete 80% to 90% of tasks through simple voice commands. According to Wang, the true “GPT moment” for embodied AI is still two to three years away. However, he emphasized that significant technological advancements are on the horizon, stating, “It could happen sooner or later, but there will definitely be major breakthroughs this year or next.”
Wang also shared insights from his friends in Silicon Valley, who believe that the “ChatGPT moment” for embodied AI could arrive in as little as 18 months. He reiterated that the coming year or two will be pivotal for the field, with substantial progress expected.
Previously, Wang has publicly emphasized Unitree’s development philosophy of “advancing through both movement and practical application.” He argued that mobility is the fundamental prerequisite for robots to perform real-world tasks. Once humanoid robots achieve sufficient agility and can execute a wide range of motions, he explained, they can be programmed to perform various tasks by combining these movements.
