On the evening of November 19, 2025, the Kennedy Center in Washington was brightly lit as the US-Saudi Investment Forum welcomedTesla CEO Elon Musk and NVIDIA CEO Jensen Huang.As the host posed questions, the two AI industry giants engaged in a dialogue about the profound impacts of artificial intelligence, robotics, and future energy development, discussing how AI and robotics will reshape industries, economies, and social structures, propelling humanity into the next intelligent era.
The conversation began with the host asking Musk how he applies the “first principles” approach to disrupt traditional industries. Musk recounted how SpaceX and Tesla used the “first principles” method to reduce battery costs by 90% and make reusable rockets a reality. He emphasized that the disruption of existing industries is more about innovation, which is based on essential improvements to existing products and industries. He believes that the robotics industry is undergoing similar transformations, with humanoid robots expected to be mass-produced and widely applied in the near future. The proliferation of robots will make human work a matter of personal preference.
Jensen Huang, as a seasoned expert in the computing field, reviewed the evolution of human computing models and technological architectures. He believes that the way computers process data is shifting from query-based to generative. He concluded that “the world is transitioning from general computing to accelerated computing.” He stated that the construction of GPU-based computing infrastructure is just beginning, with enormous demand. He expressed that NVIDIA is committed to accelerating the shift from the “CPU general era” to the “GPU factory era,” making low-latency, high-efficiency, distributed computing for real-time tasks like recommendation algorithms, autonomous driving, and AIGC as accessible as the current power grid is to data sources.
The following text will systematically summarize the viewpoints of Musk and Huang from this interview, helping you quickly grasp the latest judgments from the “AI leaders.”
First Principles Thinking Drives Innovation Rather Than Disruption
Musk stated that by starting from the basics with the “first principles” thinking approach, significant reductions in battery costs have been achieved, and the same method is being applied to the fields of robotic actuators and motors, aiming to create truly useful humanoid robots rather than merely improving existing products. Currently, there are no truly practical humanoid robots, but once breakthroughs are achieved, their market size will surpass that of smartphones, becoming the largest global product, widely used in home and personal services as well as industrial scenarios.
AI and Robotics Will Fundamentally Eliminate Poverty
Traditional means alone are insufficient to solve the problem of poverty, but the productivity leap brought by AI and robotics is expected to truly achieve universal prosperity, ultimately making currency less important, although energy and physical resources remain basic constraints. Musk also proposed an interesting viewpoint, suggesting that “work is not a necessity but a choice people make based on their hobbies,” similar to how some people prefer to grow vegetables in their gardens rather than simply buying them at the supermarket.
AI is Changing the Computing Paradigm
Huang reviewed the evolution of computing methods, transitioning from “retrieval-based computing” to “generative computing,” where software content is generated in real-time based on context, featuring high levels of personalization and intelligence. This necessitates a global network of “AI factories” to support real-time inference and training. Consequently, the world is undergoing a transformation from CPU-dominated general computing to GPU-dominated accelerated computing.
AI Increases Productivity but Will Not Reduce Workload
Huang believes that due to the proliferation of AI technology, humans will complete more tasks in the short term due to increased efficiency, potentially becoming busier. For example, AI-assisted radiologists have improved diagnostic capabilities, leading to more patients being accepted and overall industry expansion. From this, we can infer that the productivity leap brought by AI will liberate our imagination, allowing more creative ideas to be quickly realized.
Students’ learning methods and people’s work content will fundamentally change due to AI simplifying tedious tasks, freeing up more time for creative pursuits.
AI Accelerates Scientific Breakthroughs
Saudi scientists have made progress in the fields of chemistry and gene editing using AI models (such as Brock), for example, developing metal-organic framework materials to capture water and carbon dioxide from the air, and creating nanorobots to treat sickle cell disease.
Space is the Ultimate Direction for AI Development
In the long term, energy and cooling conditions on Earth are limited, while in space, continuous power supply can be achieved through solar energy and radiation cooling, enabling more cost-effective AI computing. It is expected that within a few years, space-based AI will outperform ground systems. The Starcloud project, mentioned in my previous article, is deploying such computing satellites, which can not only continuously bathe in sunlight in sun-synchronous orbit with zero water waste for cooling but will also provide low-latency AI inference services to the ground—without occupying land, needing a power grid, or requiring freshwater cooling, effectively breaking the Earth’s computing bottleneck. Starcloud’s space data center: moving AI computing off Earth, with the first satellite equipped with H100 chips already in orbit.
Current AI Investment is Not a Bubble
The last question of the interview concerned the widely discussed “AI bubble.” Huang believes that global computing is undergoing a structural transition from general CPUs to accelerated computing, with data processing and recommendation systems having largely shifted to GPUs, and generative AI developing on this solid and reasonable underlying demand.