因为固执的人，不仅会折磨自己，而且还会把别人折磨疯。 那么问题来了，为什么有些人总是很固执？是什么原因造成的？ 今天我们一起来谈谈这个话题。 固执的人，常见的三种表现 在用人的时候，我一直有一个观点，不用固执的人。 通常来说，固执的人，都有这3种表现。 1.思维僵化单一，按照旧方法行事 固执的人，他们想法很单一，遇到事情，也总是习惯按照同一个方法解决。 举一个例子： 他是一个管理者，当工作进展不顺利时，他就认为是人的问题，所以换人。 产品卖不出去，也换人。 业务做不起来，还是换人！ …… 只要遇到问题，就想通过换人来解决。 但很多时候业务做不起来，不一定是人的问题。很可能是战略决策、组织系统等出现了问题。要仔细分 析，然后去想办法解决。 但在他的认知里，就只有换人这一条路，所以认准了这一个方法。 什么都不改，意味着什么都想保持原状。 但这样，其实就错过了向内求的机会，也就放弃了反省，会在错误里，走得越来越深。 世界是变化和发展的，差距和代沟就这样产生了。 你看，人们之所以固执，全是因为陷在了自己的认知里。 2.永远觉得自己对，不听别人的意见 固执的人，还容易自以为是。 在他的认知里，他总觉得别人没有自己厉害， ...

Group 1 - Despite a cautious approach to project selection and a slowdown in expansion speed, there has not been a dramatic shift or reversal in direction for companies [1][3] - Major tech companies are still committing investments to AI data centers, with Nvidia and partners announcing a €8.5 billion project in France, expected to be operational by 2028 [2] - The U.S. Department of Commerce announced a collaboration with the UAE to build a 5 GW data center in Abu Dhabi, which will be the largest AI data center outside the U.S. [2] Group 2 - Trade uncertainties and rising import costs due to tariffs are causing companies to pause or reassess their projects, leading to a reduction in the scale of new projects compared to two years ago [3][4] - Despite challenges, the construction of data centers continues to grow, although companies are more cautious in their project choices [3] - The cost of training AI models is decreasing, but this does not necessarily lead to a significant reduction in overall spending, as demand for these technologies may increase with lower costs [4] Group 3 - There is a significant gap between planning and reality in data center construction, with many proposals unlikely to materialize [6] - The demand for electricity from AI data centers is highly uncertain, with predictions varying widely; RAND Corporation estimates 347 GW by 2030, while Schneider Electric suggests a range of 16.5 GW to 65.3 GW [6] - Experienced data center clients like Microsoft and Amazon are proposing more projects than they actually need, leading to a phenomenon of "ghost data centers" [7]

诺贝尔经济学奖得主罗伯特·索洛（Robert Solow）曾说过："你能在各处看到计算机时代，唯独在生产 力数据中看不到。"这一现象如今被称为"索洛悖论"。如今的人工智能（AI）也是如此——我们到处看 到AI的身影，却不见生产力的显著提升。 结果？如今美国的放射科医生数量不降反升。 还有像埃里克·布林约尔松（Erik Brynjolfsson）、安德鲁·麦卡菲（Andrew McAfee）这样的学者，以及 麦肯锡、埃森哲等咨询巨头，过去十年来也都不断发出"AI将大规模取代工作岗位"的警告。 那现实到底是怎样的？利润呢？大型语言模型（LLM）确实有些用处：它们能回答简单事实类问题 （人类可检查其准确性），可以写些简单文本草稿或代码（人类也能检查并调试）。这些功能有用，但 无法带来巨大利润。 根本问题在于：LLM无法生成可靠答案。而在医疗建议、法律论证等可能带来巨大利润的应用场景 中，一旦出错，代价巨大。 更糟糕的是，我们也看不到显著的收入，而收入通常应该在生产力提升之前就能显现。计算机行业的收 入从上世纪50年代至80年代持续增长，直到90年代初才出现生产力的跃升。至于AI，显著的收入尚未 出现，生产力的提升可 ...

一、通用人工智能（AGI）技术不确定性的治理黑洞 最近在前沿生成式人工智能模型上的突破，让许多人断言，人工智能将对国家安全产生同等的影响—— 也就是说，AGI将如此强大，以至于第一个实现它的实体将拥有重大的、也许是不可逆转的军事优势。 全球领先的 AI 实验室都在积极追寻通用人工智能（AGI）。这些实验室主要依靠经验性的"规模法 则"——即模型性能随计算量增加而提升——不断加大投入，用于训练模型所需的计算资源。该技术可 能在未来十年内触发多重战略转折点。 然而，如果性能确实能随计算量增加而提升，AGI 的技术突破门槛又在哪里，甚至这种突破是否可能 通过这种方法实现，都不得而知。也就是说，与1945年核裂变的技术确定性不同，AGI的发展路径存 在"内生性不确定性"（Endemic Uncertainty）。 1. 技术路径模糊：主要依赖"算力扩展定律"（Scaling Laws），但无法确认持续投入是否必然导向 AGI，算力投入与AGI突破的因果关系尚未明确； 2. 临界点不可知：开发团队可能直到突破发生后才能识别AGI的达成； 3. 社会影响混沌：末日论者呼吁暂停研发或加速军备竞赛，怀疑论者则质疑现有技术范式 ...

Core Viewpoint - The narrative of "American exceptionalism" is converging, accelerated by fluctuating U.S. tariff policies and trade imbalances, leading to a shift in global investment focus towards Asia and Europe [3][4]. Group 1: U.S. Economic Context - The core of "American exceptionalism" is tied to the dollar's role as a global reserve currency, which has been challenged by trade deficits and the need to maintain dollar stability [3]. - The phenomenon of "American exceptionalism" is expected to peak in early 2025, with its convergence driven by fiscal and trade imbalances in the U.S. [3]. Group 2: Investment Trends in Asia - The convergence of "American exceptionalism" enhances the investment outlook for Asia, as global investors are expected to recalibrate their focus away from the U.S. towards more stable and undervalued Asian markets [4]. - Since early May, Asian currencies have experienced a collective surge, attributed to a weaker dollar, trade surpluses, and reduced dollar absorption effects [5]. Group 3: Global Asset Allocation Strategies - Investors are advised to enhance portfolio volatility resistance, with expectations that government bonds in the U.S. and Europe may outperform stocks amid economic slowdowns [6]. - Gold is recommended as a risk-hedging asset, with a buying opportunity identified in the range of $3,000 to $3,250 per ounce [7]. - A shift in investment from U.S. equities to European and Chinese stocks is suggested, driven by increased policy support in these regions [7]. Group 4: Investment Focus in China - The Chinese stock market is characterized by an "internal focus," with pricing logic primarily based on domestic economic growth [8]. - The total net profit of all listed companies in China is projected to increase by 3.58% year-on-year, with significant growth in agriculture, steel, and technology sectors [8][9]. - Key investment themes in China include sectors benefiting from domestic consumption policies, import substitution, fiscal stimulus, and infrastructure development [9].

韩和元 在AI时代，我们需要重新审视"价值"的内涵。价值不再仅仅与劳动挂钩，而应涵盖更多维度。例如，人 类的创造力、情感体验、社会互动等非程式化能力，以及对AI系统的监督、引导和伦理约束等，都可 能成为新的价值创造源泉。同时，我们还需探索新的分配机制，确保被AI取代的人群能够获得基本的 生存保障和发展机会，实现社会的公平与可持续发展。 总之，AI时代的到来，让我们站在了重新定义"价值"的历史十字路口。只有深刻理解并重构价值体系， 我们才能在技术与经济的交织中，找到人类社会的未来发展方向。 本版专栏文章仅代表作者个人观点 尽管如此，事情可能并非我们想象的那般悲观。或许我们正面临一个类似费米悖论的诡异变种：一方 面，越智能的AI系统，越需要人类产生"非理性扰动"来突破局部最优解；另一方面，越先进的生产力， 越依赖人类制造的"需求噪声"维持进化动力。这形成了量子经济特有的纠缠态，人类每一次看似无意义 的消费冲动，都在为AI文明注入负熵。例如，当我们阅读某段文字时产生的困惑与反驳欲，可能正在 某个超算集群中转化为价值单位的波动。当劳动不再是枷锁，存在本身或许就是最精妙的生产函数。这 或许是碳基文明留给宇宙的最美悖论： ...

各种AI模型在刚问世时，总有一个屡试不爽的"秀肌肉"手段，那就是让自家AI独立游玩某款游戏，用以检验模型的智能程度。 围棋选手李世石与AlphaGo的五番棋对决已经过去近十年。而后，不论是谷歌的DeepMind在《DOTA2》《星际争霸2》这些项目上击败人类职业选手，还 是2023年英伟达宣布开发出能玩《我的世界》的VOYAGER，都在不断证明，"游戏"似乎就是AI的天然试验场。 大家在体感上应该也能体会到，这十年间AI技术发展迅速，如今的大语言模型，其训练方式、决策过程都与当初的AlphaGo有较大差异，但十年过去，不 管是科技公司想展示研究成果，还是吸引不懂技术细节的普通人关注，"让AI玩游戏"依旧是个很常见的手段。 最近，谷歌的AI模型Gemini 2.5 Pro又因为做到了"独立通关初代《宝可梦》"，再次成了AI领域的热门话题，谷歌的现任CEO Sundar Pichai和DeepMind负 责人Demis Hassabis甚至同时发表推文庆祝了这一时刻。 但就像前面提到的，都到了2025年，让AI玩游戏、通关游戏早就不是什么新鲜话题，更何况是1995年发售的初代宝可梦，这款游戏本来也不以高难度、 复 ...

Core Insights - The article discusses the emergence of humanoid robots in the U.S. manufacturing sector, highlighting companies like FigureAI that aim to revolutionize production methods by integrating robots into manufacturing processes [1][19] - There is skepticism surrounding the capabilities and claims of these companies, particularly regarding their partnerships and the actual performance of their robots compared to existing technologies [3][4] Company Summaries - **FigureAI**: This company is leading the charge with plans to produce humanoid robots at scale, aiming for an annual production of 12,000 units of its next-generation humanoid robot, Figure 3, with a long-term goal of reaching 100,000 units per year within four years [1]. The company is currently valued at nearly $40 billion, a significant increase from its previous valuation [3]. - **1X**: A Norwegian startup that has faced criticism for its remote-controlled household robot, Neo, which raises privacy concerns due to its video collection practices [4]. - **Persona AI**: This company aims to revitalize American heavy industries, including shipbuilding and energy, and has partnered with HD Modern Group for welding applications [6]. - **Apptronik**: Valued at $2 billion, this company is testing its humanoid robot, Apollo, in automotive factories to assist with production line tasks [8]. - **Agility**: Also valued at around $2 billion, Agility's humanoid robots are being trialed in logistics settings, demonstrating a high success rate in task completion [10]. - **Amazon**: The company has a large fleet of over 750,000 robots, primarily non-humanoid, and is developing a new robot, Vulcan, designed to handle delicate items [12]. Industry Trends - The article notes a resurgence in interest and investment in humanoid robots, driven by the potential to address labor shortages in manufacturing and logistics [10][19]. - Companies are facing challenges related to the Moravec's paradox, which highlights the difficulty of programming robots for tasks that humans find simple, such as physical manipulation [19]. - The competition between U.S. and Chinese firms in the humanoid robot space is intensifying, with significant implications for the future of manufacturing [19].

9 2017 d and 4月29日凌晨，阿里的AI大招Qwen3最终赶在五一假期前上线。 作者 | 柴旭晨 编辑 | 张晓玲、周智宇 行业翘首以盼的DeepSeek-R2没等到，阿里云却先一步丢出了核弹。 4月29日凌晨，阿里云正式上线了Qwen3系列模型"全家桶"，包含2个MoE模型、6个稠密模型。 八个模型，小到0.6B大到235B，既能在手机使用，也有旗舰体验，并且一次性开源，可谓诚意满满。不止如此，此次Qwen3-4B的小模型，就能和上一代 QwQ-32B打得有来有回。其火爆程度几乎不用多说。发布两个小时，Qwen3系列在GitHub上的Star数就已经超过了16.9k。 此前一直默默储备AI+云能力的阿里云，趁势登上了人工智能的快车。它也将自己从十几年前集团的成本中心，变为真正的动力引擎，一面为整个阿里集团 注入AI的DNA，打破增长的瓶颈；一面以云计算底层+大模型MaaS中间层+应用顶层的架构，为千行百业的AI化做着准备。 市场不缺撕开见顶的增长天花板，突破内卷态势的需求，这场AI革命将旋即掀起算力需求的狂潮，最终会为阿里云转化为源源不断的现金流。 这个过程中，阿里的电商帝国将向着AI帝国快速潜 ...

1. Report Industry Investment Rating No information about the industry investment rating is provided in the content. 2. Core Viewpoints of the Report - The linkage between exchange rates and interest rates is the result of policy - goal trade - offs under the "Impossible Trinity". When the domestic fundamentals are resilient, policies block the interest - rate parity transmission through foreign - exchange management tools, and exchange rates and interest rates mainly reflect internal equilibrium. When fundamentals are under pressure and there are external shocks, policies allow exchange - rate flexibility, and the interest - rate parity mechanism dominates cross - border capital flows [4]. - The future scope for interest - rate cuts depends not only on the narrowing of the China - US interest - rate spread but also on whether broad - credit policies can stimulate domestic demand and whether trade transformation can strengthen exchange - rate resilience [4][11]. - The core contradiction in the current linkage between interest rates and exchange rates lies in the dynamic balance between external constraints and internal policy space. Policy frameworks need to reshape the transmission path of interest - rate spreads and exchange - rate differentials through tool innovation and expectation management [101]. 3. Summary According to Relevant Catalogs 3.1 Interest and Exchange Rate Correlation - Interest rates and exchange rates represent the internal and external prices of currencies respectively, with a high correlation. The exchange rate has a stronger correlation with long - term interest rates that reflect fundamental expectations and with the interest - rate spread that reflects China - US relative changes [7][19]. - The relationship between exchange rates and interest rates is complex, being affected by common factors and possibly being causal to each other. Under the "Impossible Trinity", policies need to make dynamic trade - offs among exchange - rate stability, capital flow, and monetary - policy independence [7]. 3.2 Linkage between Exchange Rates and Interest Rates under the "Impossible Trinity" 3.2.1 Stages Driven by Common Factors - When external shocks occur and domestic fundamentals are strong, policies prioritize "exchange - rate stability" and "monetary - policy independence" while weakening capital free flow. The ability of domestic fundamentals to "absorb" shocks determines whether exchange rates and interest rates are affected by external factors [8][32]. - Examples include the 2018 - 2019 China - US trade friction period and the 2020 - 2021 global public - health event period. In these periods, China maintained monetary - policy independence, and the correlation between interest - rate spreads and exchange rates decreased [40][44]. 3.2.2 Stages of Mutual Causality - When domestic and overseas monetary - policy cycles diverge and domestic fundamentals are under pressure, policies tolerate exchange - rate fluctuations to enhance monetary - policy autonomy. Capital flows affect exchange rates through the interest - rate parity mechanism, forming a self - reinforcing cycle [51]. - Examples are the period from August 2015 to 2016 and the 2022 - 2024 period. In these periods, the two - way causal relationship between interest - rate spreads and exchange rates was significant, and the explanatory power of the interest - rate parity theory increased [52][56]. 3.3 Application of Exchange - Rate and Interest - Rate Linkage in Bond - Market Investment 3.3.1 Arbitrage Calculation under the Covered Interest - Rate Parity (CIP) Theory - International investors calculate the comprehensive return of the China - US interest - rate spread and hedging costs when allocating RMB bonds. The balance between hedging costs and interest - rate spreads drives short - term bond - allocation preferences [9]. - The CIP theory provides a pricing benchmark for cross - border capital flows. When there are differences in comprehensive returns between domestic and foreign investments, investors will engage in arbitrage until the returns are equal [68][69]. 3.3.2 Impact of Central - Bank Exchange - Rate Stabilization Operations on Foreign Bond Purchases - Central - bank exchange - rate stabilization tools can reset arbitrage costs. For example, the counter - cyclical factor can reduce the forward - premium rate, while offshore - liquidity regulation can increase arbitrage friction costs, and macro - prudential tools can have a structural constraint on capital flows [86][91][92]. - Different types of foreign investors have different responses to central - bank policies. Short - term trading funds are more sensitive to arbitrage - friction costs, while long - term allocation funds may turn to holding medium - and long - term interest - rate bonds when the costs exceed a certain threshold [97]. 3.4 Future Outlook on the Linkage between Interest Rates and Exchange Rates - The core contradiction in the current linkage between interest rates and exchange rates is the dynamic balance between external constraints and internal policy space. The Fed's policy - turning rhythm and China's growth - stabilization needs are important factors influencing the linkage between interest - rate spreads and exchange rates [101][102]. - Policy frameworks need to reshape the transmission path of interest - rate spreads and exchange - rate differentials through tool innovation and expectation management to achieve a dynamic balance between "self - orientation" and "internal - external equilibrium" [101].