
NVIDIA CEO Jensen Huang outlines the company’s strategic expansion in South Korea, emphasizing massive partnerships with SK Group and Naver to secure essential high-bandwidth memory (HBM) and scale AI infrastructure. The semiconductor industry faces a fundamental shift as computing transitions from human-centric use to supporting billions of AI agents and robots, requiring a tenfold increase in capacity over the next decade. Despite current supply chain constraints in power, land, and memory, the industry must prioritize a distributed, open-model ecosystem to ensure security, innovation, and technological sovereignty. While closed models offer efficiency for general tasks, open-weighted models provide the necessary control for specialized, regulated, and self-defensive applications. This dual approach to AI development is essential for navigating the complex, resource-constrained future of global AI deployment.
Sign in to continue reading, translating and more.
Open full episode in Podwise