Core Viewpoint - The article highlights a significant breakthrough in AI-assisted surgery by a Chinese team, demonstrating the feasibility of embodied intelligence in real-life dynamic environments, marking a shift from following to leading in the field of surgical robotics [1][2]. Group 1: Breakthrough Achievements - The research team from Chinese University of Hong Kong and Conbot has achieved three world firsts: the first autonomous surgery in a clinical setting, the first verification of multi-task surgical automation on live animals, and the first general-purpose multi-task automated AI surgical robot system [1]. - The embodied intelligence system successfully performed various surgical assistance tasks on live pigs, relying solely on visual feedback without additional sensors, indicating a major breakthrough in bridging the "simulation-reality gap" [2][3]. Group 2: Training and Efficiency - The open-source surgical training platform SurRoL has been adopted by multiple global institutions, with over 100 citations in Google Scholar, significantly improving the training speed of novice surgeons by nearly 100% under AI-assisted training [2][10]. - The AI-assisted training system allows for dynamic generation of optimal paths based on different surgical scenarios, providing personalized guidance to trainees, which is crucial given the shortage of experienced surgeons [10][12]. Group 3: Technical Innovations - The VPPV framework, which includes visual parsing, perception regression, policy learning, and visual servo control, enables the AI to abstract complex surgical scenes into understandable states for the robot, achieving zero-shot transfer from simulation to reality [3][4]. - The system's reasoning speed reaches real-time levels, with target segmentation taking 40 milliseconds and strategy prediction only 7 milliseconds, indicating that AI's computational speed is no longer a bottleneck [6]. Group 4: Future Implications - The "supervised autonomy" model proposed by the research team enhances surgical efficiency and safety, allowing surgeons to focus on complex tasks while the AI handles specific operations, potentially becoming a standard in future operating rooms [14][15]. - The development of a 3D Gaussian splatter scene reconstruction technology allows for rapid creation of realistic virtual surgical environments, significantly reducing the cost of simulation environment construction [12][13].