传媒行业人工智能系列从AI Panic到AI HALO，如何看传媒互联网的投资范式转换

Investment Rating - The investment rating for the media industry is "Outperform the Market" (maintained rating) [2][20] Core Insights - The transition from "AI Panic" to "AI HALO" indicates a shift in market sentiment, where investors are moving away from "light asset/software layers" towards "heavy assets" that have physical barriers [9][16] - The concept of "AI HALO" emphasizes the importance of heavy assets with low obsolescence, as AI cannot replicate certain physical infrastructures [3][5] - The report highlights the potential for long-term value reduction in content due to "content deflation" and the instability of application layers driven by rapid iterations of underlying AI models [10][12] Summary by Sections AI HALO Background - The AI tools have unexpectedly impacted the software industry, leading to a significant decline in global software stocks, termed "AI Panic" [6] - The market is now seeking "moats" that AI cannot replicate, focusing on heavy assets [6] Negative Perspectives - The report discusses the risk of "content deflation," where the abundance of low-cost content generated by AI could devalue traditional content IP assets [10][12] - The "intermediary crisis" suggests that internet platforms may lose their revenue-generating capabilities as AI agents reduce the need for traditional traffic channels [10][12] Positive Perspectives - The report identifies opportunities in the media and internet sectors where value may increase, such as the demand for proprietary data and the physical experience associated with IP assets [13][14] - Companies with strong emotional connections to their IP and physical retail networks are likely to thrive in an AI-dominated landscape [13][15] Investment Recommendations - The report suggests focusing on "high stickiness ecosystem" assets that have been undervalued due to market fears of AI disruption [17] - Specific sectors to watch include mid-to-heavy games, long-form video platforms, and high-sticky advertising platforms [17] Conclusion - The report emphasizes the need to identify assets that are resilient to AI disruption, particularly those with physical infrastructure or strong user engagement [16][17]