Group 1 - The article discusses the concept of "halo" assets, which are seen as foundational and resilient in the face of AI advancements, suggesting that certain heavy asset industries will benefit from AI rather than be replaced by it [1][5] - It highlights the decline of online course platforms, noting a 50% drop in users following the release of GPT-3.5, indicating that light internet platforms may struggle in the AI era [3][5] - The article argues that industries with unique vertical data and strong operational know-how will be less affected by AI, while those without such advantages will face erosion [5] Group 2 - It emphasizes that heavy asset industries, characterized by low obsolescence rates and strong barriers to entry, are likely to thrive and even monopolize in the AI landscape [5] - The discussion includes the idea that companies like OpenAI cannot replace heavy asset industries, as they lack access to critical industry data and cannot build infrastructure like steel mills or oil transport [5] - The article concludes that "halo" assets are positioned between AI infrastructure and edge AI, representing valuable clients that benefit from AI advancements [5]
都在说halo 资产
小熊跑的快·2026-02-28 00:57