在2024年NVIDIA GTC大会上，黄仁勋提出‘物理AI’（Physical AI）概念，重新定义了数字员工的内涵。他指出，未来的AI不仅限于处理数据和生成内容，更将具备感知、推理与行动能力，能够与物理世界互动——这正是‘物理AI’的核心。黄仁勋强调，数字员工不再是虚拟助手或聊天机器人，而是能通过传感器、执行器和具身智能，在现实环境中完成复杂任务的智能体。例如，在工厂中自主巡检设备、在仓库中搬运货物，甚至协助人类进行精密操作。为实现这一愿景，NVIDIA推出了新一代AI基础设施，包括Project GR00T通用人形机器人基础模型、VIMA多模态操控框架，以及Isaac平台的全面升级。这些技术共同构建了一个从模拟训练到真实部署的闭环系统，使AI能在动态、不确定的物理环境中学习和适应。黄仁勋认为，物理AI将开启人机协作的新纪元，大幅提升生产力，并重塑制造业、物流、医疗等多个行业。通过赋予AI‘身体’和‘行动力’，数字员工将真正从屏幕走向现实世界，成为人类不可或缺的合作伙伴。
At NVIDIA’s GTC 2024 conference, Jensen Huang introduced the concept of ‘Physical AI,’ redefining what digital workers mean. He emphasized that future AI will go beyond processing data or generating content—it will possess perception, reasoning, and actuation capabilities to interact directly with the physical world. According to Huang, digital employees are no longer just virtual assistants or chatbots; they are intelligent agents equipped with sensors, actuators, and embodied intelligence capable of performing complex tasks in real-world environments—such as autonomously inspecting machinery in factories, moving goods in warehouses, or assisting humans in precision operations. To realize this vision, NVIDIA unveiled a new generation of AI infrastructure, including Project GR00T (a foundational model for general-purpose humanoid robots), the VIMA multimodal manipulation framework, and a comprehensive upgrade to the Isaac robotics platform. Together, these technologies form a closed-loop system that enables AI to train in simulation and deploy effectively in dynamic, uncertain physical settings. Huang believes Physical AI will usher in a new era of human-robot collaboration, dramatically boosting productivity and transforming industries like manufacturing, logistics, and healthcare. By giving AI a ‘body’ and the ability to act, digital workers will step out of screens and into the real world as indispensable partners to humans.
原创文章，作者：admin，如若转载，请注明出处：https://ktez.cn/10025.html