Recently, NVIDIA CEO Jensen Huang has repeatedly emphasized the concept of ‘Physical AI.’ Physical AI refers to intelligent systems that can understand and act within the physical world—not just process digital data like text or images. Such AI must integrate perception, reasoning, planning, and execution capabilities to interact with real-world environments and perform tasks like operating robots, driving vehicles, or managing industrial machinery. Huang believes the future of AI will shift from ‘digital AI’—models confined to virtual spaces—to ‘physical AI,’ which embodies intelligence through interaction with the real world. This evolution requires not only powerful computing and large models but also sensors, actuators, and an understanding of physical laws. NVIDIA is building the infrastructure for Physical AI through its Omniverse platform, Isaac robotics platform, and AI chip ecosystem. Huang stresses that Physical AI will be a key driver of transformation in robotics, autonomous driving, and smart manufacturing, representing the next major frontier in artificial intelligence.
近期，英伟达CEO黄仁勋在多个场合反复强调“物理AI”（Physical AI）这一概念。所谓“物理AI”，指的是能够理解并作用于物理世界的智能系统，而不仅仅是处理文本、图像等数字信息。这类AI需要融合感知、推理、规划与执行能力，通过与真实环境互动来完成任务，例如操控机器人、驾驶车辆或操作工业设备。黄仁勋认为，未来的AI将从“数字AI”迈向“物理AI”，即从仅在虚拟世界中运行的模型，进化为能与现实世界交互、具备具身智能（embodied intelligence）的系统。这不仅依赖强大的算力和大模型，还需要传感器、执行器以及对物理规律的理解。英伟达正通过其Omniverse平台、Isaac机器人平台和AI芯片生态，构建支持物理AI发展的基础设施。黄仁勋强调，物理AI将是推动机器人、自动驾驶和智能制造等产业变革的关键力量，也是AI下一阶段的重要方向。
原创文章，作者：admin，如若转载，请注明出处：https://avine.cn/16709.html