海外政策周报 独木难支：为什么英伟达财报暂未打消市场对 AI 的疑虑？ 海外政策周聚焦 核心结论 摘要内容 英伟达财报数据亮眼，市场短暂上涨，随后大跌。有两大原因：美联储降息 预期破灭与 AI 估值泡沫担忧。 英伟达积极财报暂未平息 AI 泡沫担忧。市场质疑 AI 投资回报率与循环融资。 类似于加州淘金热期间，追踪铁锹、鹤嘴锄和筛金盘的销售数据，并不能真 正证明淘金热的繁荣。当前英伟达的销售业绩，或许也不一定能巩固 AI 革 命的合理性。 更深层的，AI 叙事的基本假设也在被质疑。 2022 年 ChatGPT 诞生以来，市场的 AI 叙事有两条基本假设：1）当年智 能手机催生了全新的科技生态，人工智能会复制同样的故事；2）AI 会提升 生产率，最终受益者大概率是那几家美国的科技巨头。但这两条逻辑正在面 临挑战。 AI 创新与智能手机存在两大差异，基建投资规模更大，开源模型竞争导致 赢家盈利水平更低。因而 AI 投资回报率或低于智能手机。 历史上看，索洛悖论表明，技术进步加快不一定表现在全要素生产率提高上。 因为信息技术投资与生产率增速之间存在统计落差。 我们认为，AI 对劳动生产率的提升效果，是类似电力等新 ...

Core Insights - The current AI investment boom is characterized by both revolutionary potential and speculative bubbles, with experts suggesting that the true bubble lies in unrealistic macro narratives rather than the technology itself [1][7]. Group 1: AI Investment Trends - A significant portion of the U.S. economic growth this year is attributed to AI investments, with predictions indicating that over 90% of this growth is linked to AI [1]. - The concentration of market value in the U.S. stock market is notable, with over 30% of the S&P 500 index value held by the top seven tech companies [2]. - The AI investment trend is described as a "rational bubble," where the costs of under-investment are perceived to outweigh the risks of over-investment [2]. Group 2: Historical Context and Future Outlook - Historical patterns show that disruptive technologies often come with significant investment bubbles, which are difficult to avoid [3]. - The development of AI in China is aimed at breaking supply-side growth constraints through productivity improvements, especially in light of an aging population [3][4]. - The "Solow Paradox" is referenced, highlighting the discrepancy between technological advancements and actual productivity gains, emphasizing the need for AI to enhance productivity across various sectors [4]. Group 3: Practical Applications and Market Dynamics - The AI landscape is expected to evolve significantly by 2025, moving beyond basic content generation to deeper industrial applications [5][6]. - The Chinese government has set ambitious goals for AI integration across various sectors, aiming for over 70% application penetration by 2027 [6]. - Startups focusing on vertical applications of AI are seen as more viable than those attempting to develop foundational models without clear market needs [7]. Group 4: Addressing the AI Bubble - The notion of "squeezing" the bubble through genuine market demand and solving real problems is emphasized, with a focus on practical applications of AI technology [7]. - The importance of aligning AI development with actual human needs is highlighted, as seen in projects aimed at creating assistive technologies for individuals with disabilities [7].

Core Insights - The global economy and technological landscape are undergoing significant changes, with artificial intelligence (AI) being a central force driving this transformation [1] - The recent dialogue at the Taihu World Cultural Forum highlighted AI as a key topic of interest among experts [1] Investment Trends - The current "craze" in global stock markets is largely driven by enthusiasm and investment in the digital realm, particularly AI [3] - Major companies are heavily investing in AI model development and related infrastructure, including quantum computing and data centers [3] - Over 30% of the market capitalization of the S&P 500 is concentrated in the top seven tech companies [3] - AI investment is characterized as a "rational bubble," driven by competitive pressures rather than irrational exuberance [5] Competitive Landscape - The gap between the US and China in AI is rapidly narrowing, with both countries increasing their investments to avoid falling behind in strategic competition [3][5] - Chinese innovations are fostering the development of open-source ecosystems and breakthroughs in quantum computing [3] - AI is accelerating scientific discoveries, as evidenced by recent Nobel Prize achievements [3] Societal Challenges - The development of AI presents new societal challenges, including labor market changes and job displacement [4] - There is a growing consensus that the future applications of AI will depend on choices made today, necessitating a balance between automation and human collaboration [4] European Context - Europe lacks globally influential tech giants and is facing challenges in AI innovation due to strict regulatory frameworks [7] - The EU's regulations, such as GDPR and the AI Act, while effective in protecting privacy, may stifle innovation [7] - There is a need for a balanced policy framework that promotes innovation while managing risks [7] Emerging Markets - Emerging economies generally have a more optimistic view of AI compared to developed nations, with AI offering new opportunities for growth [8][9] - The core development tools for AI are concentrated in the US and China, while the application of AI is more accessible to many countries [8] - Countries with stable infrastructure are better positioned to leverage AI technology, while those lacking it risk marginalization [9]

Core Insights - The current wave of AI investment is seen as both a revolution and a potential bubble, but it is unlikely to cool down due to pressing demands for productivity improvements, geopolitical competition, and the necessity for companies to adopt AI to remain competitive [2][3] Group 1: AI Investment Drivers - East Asian countries face challenges such as declining birth rates, labor shortages, and accelerated aging, making AI a critical solution for enhancing labor productivity [2] - Geopolitical tensions have made AI capabilities a key determinant of national power, prompting countries to vigorously pursue AI advancements [2] - Major companies view AI technology as a symbol of technological leadership and market position in the context of strategic competition [2] - Organizations recognize that failing to embrace AI could lead to competitive disadvantages and potential obsolescence [2] Group 2: The Dual Nature of AI - AI is characterized as a "creative destruction," meaning it brings both innovation and disruption, which is inherent in any technological advancement [2][3] - Historical patterns indicate that disruptive technology cycles are often accompanied by investment booms and bubbles, which are difficult to avoid [3] Group 3: Productivity and Economic Growth - The "benchmarking" concept is proposed, emphasizing that AI should be directed towards enhancing productivity to overcome supply-side growth constraints [4] - China's economic growth is increasingly limited by demand-side factors, particularly due to population decline and aging, necessitating AI solutions to boost consumer demand [5] - The Solow Paradox suggests that while new technologies can enhance productivity, actual improvements may not be realized uniformly across different sectors and regions, leading to a widening productivity gap [4] Group 4: Institutional Framework for AI - The successful implementation of AI requires a supportive institutional environment, which can be achieved through reforms that balance the creative and destructive aspects of new technologies [6] - Institutional reforms are necessary to ensure that AI contributes positively to societal needs, such as supporting the elderly and addressing demographic challenges [6]

Core Insights - The application of artificial intelligence (AI) in the financial sector is still in its early stages, with both potential benefits and risks needing careful evaluation [1][9] - AI is expected to bring significant marginal changes to the financial system, particularly in banks [5] Group 1: AI Applications in Finance - AI is deeply integrated into various financial processes, primarily focusing on optimizing business operations and customer service [3] - Key areas of AI application include middle and back-office operations, customer relationship management, and the provision of financial products [3] - AI helps financial institutions reduce costs and improve efficiency while offering more personalized and precise services to clients [3] Group 2: Risks Associated with AI - The introduction of AI brings new systemic risks and potential channels for risk transmission [7] - Risks can be observed from both micro and macro perspectives, including model stability risks and data governance risks at the micro level, and concentration risks and decision-making homogeneity risks at the macro level [7] - Concentration risk arises from reliance on a few strong technology providers, while decision-making homogeneity can lead to synchronized industry decisions, potentially causing a "resonance" effect [7] Group 3: Regulatory and Policy Considerations - The impact of AI on monetary policy requires long-term observation, as AI's influence is not yet significant [10] - AI can affect data collection and processing related to monetary policy decisions, but monetary policy adjustments are generally slow and based on economic cycles [10] - The role of human expertise remains crucial in key areas such as credit, insurance pricing, and actuarial science, despite the advancements in AI [9]

Core Viewpoint - The application of artificial intelligence (AI) in the financial sector is still in its early stages, with potential benefits in efficiency and risks that need careful evaluation [1][6]. Group 1: AI Applications in Finance - AI is deeply integrated into various financial processes, primarily focusing on optimizing business operations and customer service [3]. - Key areas of AI application include middle and back-office operations, customer relationship management, and the provision of financial products [3]. - AI helps financial institutions reduce costs and improve efficiency while offering more personalized and precise financial products and services to clients [3]. Group 2: Opportunities and Changes - AI provides new development opportunities for the financial system, particularly in banking, leading to significant marginal changes [5]. - The financial system has a solid foundation for AI applications due to the vast amounts of data accumulated over time, which can be utilized for machine learning and deep learning [4]. Group 3: Risks Associated with AI - The introduction of AI brings new systemic risks and new channels for risk transmission, enhancing both the monitoring capabilities of regulators and the potential impact of risks [7]. - Risks can be observed from both micro and macro perspectives, including model stability risks and data governance risks at the micro level, and concentration risks and decision-making homogeneity risks at the macro level [7]. - Concentration risk arises from reliance on a few strong technology providers, while decision-making homogeneity risk may lead to synchronized decision-making across financial institutions, potentially causing a "resonance" effect [7]. Group 4: Impact on Monetary Policy - The influence of AI on monetary policy requires long-term observation, as its effects are not yet clearly defined [9][10]. - AI can impact monetary policy decisions through data collection and pattern recognition, but monetary policy adjustments are generally slow and influenced by economic cycles [10].

Core Viewpoint - The application of artificial intelligence (AI) in the financial sector is still in its early stages, with potential risks and regulatory issues being widely discussed. Experts emphasize the need for careful evaluation of the benefits and drawbacks associated with AI in finance [1][5]. Group 1: AI Applications in Finance - AI is deeply integrated into various financial processes, primarily focusing on optimizing business operations and customer service. Key areas of application include middle and back-office operations, customer relationship management, and the provision of financial products [2]. - The intelligentization of middle and back-office operations is already widely adopted in financial institutions, covering data collection, processing, information identification, and customer assessment [2]. - AI applications in providing financial products yield dual benefits: internally, they help reduce costs and improve efficiency; externally, they enable financial institutions to offer more personalized and precise products and services to clients [2]. Group 2: Risks Associated with AI - While AI enhances efficiency, it also introduces new systemic risks and channels for risk transmission. The potential impact of these risks is significant, necessitating careful monitoring [5]. - From a micro perspective, individual financial institutions face model stability risks and data governance risks. From a macro perspective, the industry faces concentration risks and decision convergence risks [5]. - Concentration risk arises from the reliance on a few technology providers with strong capabilities, potentially increasing market concentration. Decision convergence risk occurs when institutions use standardized models and data, leading to homogeneity in decision-making across the industry [5]. Group 3: Impact on Monetary Policy - Despite the rapid development of AI, its application in finance remains auxiliary and cannot replace human decision-making. Human expertise is still crucial in key areas such as credit, insurance pricing, and actuarial science [6]. - The influence of AI on monetary policy is not yet significant, as monetary policy adjustments are slow variables that respond to economic cycles rather than immediate changes [7]. - Further observation and research are required to understand the long-term effects of AI on monetary policy, as AI's impact on data collection and processing may not translate into immediate policy changes [7].

Core Insights - The rapid rise of artificial intelligence (AI) highlights the unpredictability of technology [1] - There is a possibility of experiencing a "Solow Paradox," where significant investments in large computers do not lead to productivity improvements [1] - It may take more time for AI to translate into a leap in productivity, necessitating a cautious evaluation of the benefits and drawbacks associated with AI [1]

Core Viewpoint - The article emphasizes the urgent need to address the sustainability of the pension system in China, driven by factors such as aging population, labor market challenges, and the potential for increased productivity through artificial intelligence. It argues that the issue is not a lack of material wealth but rather inadequate institutional arrangements to support social security [4][5][10]. Group 1: Factors Affecting Pension Sustainability - The first factor is demographic, with a rapidly increasing aging rate. By 2032, over 21% of China's population will be aged 65 and above, indicating a significant aging society while income levels remain relatively low compared to developed nations [7]. - The second factor is the labor market, characterized by structural employment issues, high youth unemployment (17.8% for ages 16-24), and the challenges faced by older workers nearing retirement [8][9]. - The third factor is labor productivity, which has the potential for unlimited growth due to advancements in artificial intelligence. The expected annual growth rate of the "supporting productivity" for the working-age population is projected at 5.55%, outpacing the growth of the elderly dependency ratio [9][10]. Group 2: Institutional Arrangements and Recommendations - Current social security arrangements are insufficient to share the benefits of increased productivity, necessitating reforms in the pension system to ensure equitable distribution of wealth generated by productivity gains [12][19]. - The article suggests establishing a universal social security system that includes a "living wage" and "unconditional basic income" to address the challenges posed by artificial intelligence and ensure comprehensive coverage for all citizens [19][20]. - It also advocates for a reconsideration of nominal account systems, emphasizing the need for a record-keeping approach that does not require actual funding but ensures the sustainability of the pay-as-you-go pension system [20][21].

Core Viewpoint - AI provides a historic opportunity for China's economy, aiding in overcoming the middle-income trap and addressing the challenges of an aging population. The transition from a policy-driven phase to a performance-driven phase is underway, indicating a significant economic transformation ahead [1][31]. Group 1: Economic Growth Paradigm - China's economy is entering a new growth paradigm, with a shift in capital market dynamics towards "innovation-efficiency" as evidenced by the rise of companies like Cambricon [2]. - The government's strategic focus on AI development has been solidified with the release of the "AI+" action plan, marking AI's elevation to a national strategy [2][5]. Group 2: Market Validation and AI Impact - Current market trends suggest that the assumptions made in AI models are being validated, with a recognition that both "dreams" and "reality" are being traded in the market [5]. - AI's penetration is expected to significantly mitigate the decline in potential economic growth rates, with projections indicating that a 20% AI penetration could sustain growth at around 5.8% by 2035, compared to a baseline of 4.6% [6][7]. Group 3: Capital Structure Transformation - The transition from "land finance" to "computing power finance" is a profound and irreversible trend, reshaping local government financial structures [10][11]. - The sustainability of this transition relies on increasing computing power utilization and the ability to generate revenue from AI-related assets [12][14]. Group 4: Addressing the Solow Paradox - AI has the potential to address the Solow Paradox, where technological advancements do not immediately translate into productivity gains. Key indicators include the ratio of AI capital expenditure and the revenue-to-cost ratio [15][16]. - A systematic measure called Elasticity of Compute-to-Output (ECO) is proposed to assess AI's impact on productivity, with a threshold of ECO greater than 0.25 indicating effective productivity enhancement [16]. Group 5: Market Valuation and Pricing Models - Traditional valuation metrics are inadequate for AI companies, which are often priced based on future earnings potential rather than current profitability [19][20]. - A more robust valuation approach involves using compute rent discount models and focusing on cash flow from AI-related revenues [20]. Group 6: Application and Commercial Viability - The most promising areas for AI applications that could achieve commercial viability include AI in financial services, industrial software, and biopharmaceuticals, with financial services expected to lead in generating cash flow [22][23]. - The criteria for identifying sectors likely to achieve commercial success include rigid demand, quantifiable ROI, and established data barriers [23]. Group 7: Strategic Outlook and Market Signals - The goal of achieving over 70% penetration of new intelligent applications by 2027 is aimed at creating a substantial domestic market for AI, fostering competition and profitability [25]. - Key signals to monitor for potential market overheating include financing ratios, regulatory attitudes, and insider selling behaviors [26][27].