半导体芯闻·2026-01-23 09:38Core Insights - NVIDIA's CEO Jensen Huang emphasized that the world has only scratched the surface in building the necessary infrastructure for artificial intelligence (AI), with thousands of billions already invested and trillions more needed for development [1] - Huang described a "five-layer cake" of modern AI infrastructure, which includes energy, semiconductors, cloud infrastructure, AI models, and application layers, indicating significant economic benefits will arise from these developments [1] - The demand for high bandwidth memory (HBM) driven by large-scale AI data centers is prompting major investments from memory and NAND flash suppliers like Micron, SK Hynix, and Samsung [1] Investment Trends - Huang noted that 2025 is projected to be one of the largest years for venture capital, with significant funding directed towards AI-native companies due to advancements in AI models [2] - The introduction of reasoning models and open models has instilled confidence in investors to fund large-scale AI infrastructure projects [2] Industry Impact - AI advancements are leading to significant progress in sectors like pharmaceuticals, with companies like Eli Lilly achieving breakthroughs that allow scientists to interact with proteins similarly to ChatGPT [2] - The impact of AI on employment is mixed; while it may create high-paying jobs in skilled trades, its long-term effects on other professions, such as radiology and nursing, are still uncertain [2][3] Healthcare Sector - Increased patient numbers have led to higher hospital revenues, resulting in the hiring of more radiologists and nurses, countering the notion that AI will reduce job numbers in these fields [3] - The financial services and legal sectors may see a reduction in analyst positions due to AI's influence [3] Global AI Development - Huang advocates for every country to participate in building their own AI infrastructure, leveraging their unique languages and cultures to develop localized AI solutions [4]