复盘2025年的科技圈时，我们发现了一个讽刺的现象： 当SaaS厂商绞尽脑汁向客户解释"为什么你需要一个AI副驾驶"，手机厂商还在发布会上演示"一键消除 路人"等不痛不痒的微创新，某些所谓的大厂频繁鼓吹AI战略刺激股价时，游戏行业悄无声息地完成了 AI的商业闭环。 比如外界还在讨论AI何时能落地时，腾讯的广告系统已经用AI把流量变现效率拉升到了新高度；人们 还在调戏对话机器人时，网易的NPC已经开始用情感羁绊让玩家心甘情愿地掏钱。 不客气地说，其他行业还在把AI当作"秀花活"的营销噱头，或者仅仅停留在"内化"工具来辅助写代码 时，国内的游戏公司已经率先把AI变成了利润表上实打实的增长引擎。 本文将剥离技术滤镜，用财务视角还原这场发生在中国游戏产业的静默革命。 01 腾讯的"大象起舞"：AI学会了赚钱 在外界的刻板印象里，腾讯的AI故事往往不如OpenAI那样性感。然而财报数据却讲述了一个完全不同 的故事：腾讯可能是这波AI浪潮中，变现最快、赚钱最狠的中国互联网巨头。 有趣的悖论在于，AI对腾讯游戏业务最大的贡献，并没有直接发生在"游戏开发"，而是发"卖游戏"环 节。 广告，成了AI变现的头号功臣。 根据财报数 ...

我们也许正目睹两个世界的激烈碰撞：高速迭代的人工智能世界VS缓慢运转、资本密集型的核物理世 界。 近期一项针对600余名全球投资者的调查数据显示，63%的受访者现将人工智能的电力需求，视为核能 规划中的"结构性"转变。这并非暂时性的需求激增或投机泡沫——每次大型语言模型(LLM)的查询，最 终都以物理足迹的形式体现在全球资产负债表上。 多年来，能源议题始终一直围绕着"效率"展开——人们被告知更先进的芯片将抵消更高的使用量。然 而，这样的一个时代可能已经结束了。生成式AI不仅仅是使用数据，它还在为了创造数据而"焚烧"能 源。 为何"效率"叙事失效了？ AI领域当前的"逆波兰"现实是：人们把芯片做得越高效，部署的芯片也会越多，模型也将变得越复 杂。这就是著名的杰文斯悖论(Jevons Paradox)——当技术进步提高了效率，资源消耗不仅没有减少， 反而激增。 类似的场景眼下就正在美国北弗吉尼亚州和新加坡的数据中心实时上演…… 当审视超大规模AI数据中心的能耗密度时，人们看到的绝非过往传统办公楼的模式。这些数据中心设 施的耗电量堪比中等规模城市，同时它还有着99.999%时间无缝运行的要求。 135美元的铀价天花 ...

本文字数：2306，阅读时长大约4分钟 作者 | 第一财经 高雅 根据职场咨询公司Challenger, Gray&Christmas的数据，2025年美国约有5.5万次裁员被归因于人 工智能（AI）。包括亚马逊和赛富时（Salesforce）在内的科技巨头裁减了上万个岗位，并将AI列 为裁员的主要因素之一。 与此同时，麻省理工学院（MIT）于上个月发布的一项研究显示，AI已经能够胜任美国劳动力市场约 11.7%的工作，并有望在金融、医疗及其他专业服务领域节省高达1.2万亿美元的工资支出。特斯拉 公司创始人马斯克（Elon Musk）也频繁表示，他预测未来人类的工作并不是必须的，工作会像"运 动或游戏一样"。 然而， 裁员与AI应用之间的关系远比表面情况更为微妙。美国拜登政府时期首任AI科学特使、AI审 计与评估公司"人道智能"首席执行官乔杜里博士（Dr. Rumman Chowdhury）接受第一财经采访 时表示，这是一个远比"每个人都有普遍基本收入"或"未来没有工作"更为复杂的故事。 她解释道："我们看到最底层的工作被自动化淘汰，但随着信息流动得越来越快，新工作也在不断涌 现。这正是AI的作用，它帮助我们 ...

对于硅谷的裁员，乔杜里分析称，这一现象已持续约三到四年，其背后并非完全源于纯粹的AI创 新。"在某些领域，比如开发者群体中，AI确实显示出替代效应。它取代初级开发者的能力已相当先 进，且这类案例正在增多。"她随后强调，"但同样值得关注的是，所有这些公司都投入了数千亿美元开 发一项尚未实现盈利的技术，因此它们必须在其他地方节省开支。我不能完全确定所有这些裁员都是纯 粹由自动化驱动的。" AI自动化研究所创始人贝格（Hamza Baig）在一篇名为《美国企业大规模裁员背后：AI洗白还是经济 现实？》的文章中写道，当前许多裁员"实质上是将传统重组包装成AI驱动的创新"，但企业需要区分真 正的AI转型与以自动化为名、掩盖运营低效的行为。 IBM首席执行官克里希纳（Arvind Krishna）也在本月公开承认，近期的裁员更像是一场"自然修正"。 他认为人们在就业上"暴饮暴食"了，当前的调整主要是为了解决过度招聘的问题，而非完全由AI导致。 近期硅谷更像是用AI来洗白裁员的决策，长期来看，AI会带来更多工作岗位。 根据职场咨询公司Challenger, Gray&Christmas的数据，2025年美国约有5.5万次裁 ...

Core Viewpoint - The global AI industry has transitioned from a model performance competition to a "life-and-death race" for the large-scale implementation of intelligent agents, where cost reduction is no longer optional but a critical factor for profitability and industry breakthroughs [1] Group 1: Cost Reduction Breakthrough - Inspur Information has launched the Yuan Brain HC1000 ultra-scalable AI server, achieving a breakthrough in inference cost to 1 yuan per million tokens for the first time [2][3] - This breakthrough is expected to eliminate the cost barriers for the industrialization of intelligent agents and reshape the underlying logic of competition in the AI industry [3] Group 2: Future Cost Dynamics - Liu Jun, Chief AI Strategist at Inspur, emphasized that the current cost of 1 yuan per million tokens is only a temporary victory, as the future will see an exponential increase in token consumption and demand for complex tasks, making current cost levels insufficient for widespread AI deployment [4][5] - For AI to become a fundamental resource like water and electricity, token costs must achieve a significant reduction, evolving from a "core competitiveness" to a "ticket for survival" in the intelligent agent era [5] Group 3: Historical Context and Current Trends - The current AI era is at a critical point similar to the history of the internet, where significant reductions in communication costs have driven the emergence of new application ecosystems [7] - As technology advances and token prices decrease, companies can apply AI on more complex and energy-intensive tasks, leading to an exponential increase in token demand [8] Group 4: Token Consumption Data - Data from various sources indicates a significant increase in token consumption, with ByteDance's Doubao model reaching a daily token usage of over 50 trillion, a tenfold increase from the previous year [13] - Google's platforms are processing 1.3 trillion tokens monthly, equivalent to a daily average of 43.3 trillion, up from 9.7 trillion a year ago [13] Group 5: Cost Structure Challenges - Over 80% of current token costs stem from computing expenses, with the core issue being the mismatch between inference and training loads, leading to inefficient resource utilization [12] - The architecture must be fundamentally restructured to enhance the output efficiency of unit computing power, addressing issues such as low utilization rates during inference and the "storage wall" bottleneck [14][16] Group 6: Innovations in Architecture - The Yuan Brain HC1000 employs a new DirectCom architecture that allows for efficient aggregation of massive local AI chips, achieving a breakthrough in inference cost [23] - This architecture supports ultra-large-scale lossless expansion and enhances inference performance by 1.75 times, with single card utilization efficiency (MFU) potentially increasing by 5.7 times [27] Group 7: Future Directions - Liu Jun stated that achieving a sustainable and significant reduction in token costs requires a fundamental innovation in computing architecture, shifting the focus from scale to efficiency [29] - The AI industry must innovate product technologies, develop dedicated computing architectures for AI, and explore specialized computing chips to optimize both software and hardware [29]

智能体时代，token成本就是竞争力 回顾互联网发展史，基础设施的"提速降费"是行业繁荣的重要基石。从拨号上网以Kb计费，到光纤入 户后百兆带宽成为标配，再到4G/5G时代数据流量成本趋近于零——每一次通信成本的显著降低，都推 动了如视频流媒体、移动支付等全新应用生态的爆发。 当前的AI时代也处于相似的临界点，当技术进步促使token单价下滑之后，企业得以大规模地将AI应用 于更复杂、更耗能的场景，如从早期的简短问答，到如今支持超长上下文、具备多步规划与反思能力的 智能体……这也导致单任务对token的需求已呈指数级增长。如果token成本下降的速度跟不上消耗量的 指数增长，企业将面临更高的费用投入。这昭示着经济学中著名的"杰文斯悖论"正在token经济中完美 重演。 来源：美通社 北京2025年12月25日 /美通社/ -- 当前全球AI产业已从模型性能竞赛迈入智能体规模化落地的"生死竞 速"阶段，"降本" 不再是可选优化项，而是决定AI企业能否盈利、行业能否突破的核心命脉。在此大背 景下，浪潮信息推出元脑HC1000超扩展AI服务器，将推理成本首次击穿至1元/每百万token。这一突破 不仅有望打通智能体 ...

预计2025-2027年全球GDP增速高于市场一致预期，流动性已提前释放且财政接力，大美丽法案逐步实 施正向影响较大，经济深度衰退概率低；美国零售与就业数据稳健，消费动能仍在。对久期债券相对谨 慎，对信用债保持中性，美元长期仍看空，黄金仍看好，继续看好全球股票。中国资产处于外资低配状 态，估值吸引力上升，若情绪趋稳存在回补空间，投资者可考虑减美国资产、增欧洲和亚洲资产，中国 台湾和韩国也有结构性机会。 多资产 债券-利率债&信用债 今年十年国债1.65%-1.90%区间波动，7-9月债大幅度调整，随后小幅走多，目前到了均衡位。市场以看 多和中性观点为主，年末抢跑或导致利率下行但空间受限。央行下场买债，地产与基建实物量持续低于 预期，社融信贷同步走弱，货币宽松未转向，为债市提供下行保护。机构行为来看，银行理财突破32万 亿，固收类理财偏好信用票息；二级债基热销，ETF持续吸金，主动债基赎回明显；公募销售新规未落 地，或有赎回扰动。 地产基建数据持续下行，基建实物量远落后往年且11月季节性下滑，地产投资加速下行，一线二手房挂 牌激增、新房集中处置冲击情绪，30城成交仍低、居民房贷负增长，12月会议难见强刺激。财政 ...

德银认为，当前AI热潮并非单一泡沫，而是由估值、投资、技术三重泡沫交织。公开市场巨头估值有盈利支撑，而私营公司估值已极度高企。天量投资 由现金流驱动，非债务扩张，但复杂循环融资与潜在技术瓶颈埋下风险。AI需求强劲且成本骤降，但能源与芯片供应或成最终制约。 站在2025年12月的时间节点，距离ChatGPT发布仅过去三年，市场对于"AI泡沫"的讨论已至沸点。德意志银行认为，当前AI热潮既不是完全的泡沫，也 不是毫无风险，关键在于区分不同类型的"泡沫"。 12月12日，德银在最新研报中创新性地将AI泡沫分为估值泡沫、投资泡沫和技术泡沫三个维度进行分析。 报告称， 公开市场大型科技公司的估值有盈利支撑，投资增长符合趋势且由现金流推动，技术进步仍在持续。真正的风险集中在估值过高的私营公 司、可能失控的循环融资结构，以及潜在的技术瓶颈和供应限制。 估值泡沫：估值分化揭示真实风险所在 德银的核心观点是当前AI热潮并非单一泡沫，而是由三种不同性质的泡沫构成。 在估值维度 ，报告显示希勒周期调整市盈率（Shiller Cyclically Adjusted Price/Earnings ratio）已超过40，接近2000年 ...

Core Insights - The article discusses the dual role of AI as both an energy consumer and an energy efficiency enhancer, highlighting the potential for AI applications to significantly reduce energy consumption over time despite its immediate energy demands [1][2]. Group 1: AI's Energy Consumption - AI's energy demand is growing rapidly, with data centers projected to consume 1.5% of global electricity by 2024, amounting to approximately 415 TWh, with the U.S. accounting for 45% of this consumption [4]. - The International Energy Agency forecasts that global data center electricity consumption will more than double by 2030, reaching around 945 TWh, driven primarily by AI and other digital services [4]. - In the U.S., data centers are expected to contribute nearly half of the electricity demand growth from now until 2030, surpassing the total electricity consumption of energy-intensive industries like aluminum and cement [4][5]. Group 2: AI's Role in Energy Efficiency - AI can act as a "savings tool" in the real economy by optimizing energy supply systems, improving industrial processes, and enhancing efficiency in sectors like transportation and construction [1]. - AI technologies are being developed to reduce energy consumption during model training and inference, with innovations such as the "Mixture of Experts" (MoE) architecture leading to a 70% reduction in training energy consumption [1][6]. - Companies like Tencent and Google are actively pursuing green energy initiatives, with Tencent aiming for 100% renewable energy by 2030 and Google exploring hourly matching of renewable energy supply [9][10]. Group 3: Innovations in Energy Supply and Consumption - AI is enhancing energy supply systems by improving predictive accuracy and operational strategies, particularly in renewable energy sectors [11][12]. - In industrial applications, companies are using AI to optimize processes, resulting in significant energy efficiency gains, such as a 3% improvement in energy use at ArcelorMittal's Luxembourg plant [14]. - AI applications in transportation and building management are also yielding substantial energy savings, with logistics companies reducing fuel costs by 20% through route optimization [15][16]. Group 4: Future Prospects and Challenges - The relationship between AI's energy consumption and its potential for energy savings is complex, with short-term increases in energy use expected before long-term savings materialize [19][20]. - The development of fusion energy technology is seen as a potential long-term solution for providing zero-carbon energy to support AI's growth [21]. - The article emphasizes the need for a balanced approach to AI deployment, ensuring that energy efficiency gains are realized while managing the immediate energy demands of AI systems [23].

Core Insights - The report titled "State of AI: An Empirical 100 Trillion Token Study with OpenRouter" analyzes the usage of over 300 AI models on the OpenRouter platform from November 2024 to November 2025, focusing on real token consumption rather than benchmark scores [3][5][67] - It highlights the significant rise of open-source models, particularly from China, which saw weekly token usage share increase from 1.2% to 30%, indicating a shift towards a complementary relationship between open-source and closed-source models [2][10][74] - The report emphasizes the transition of AI models from language generation systems to reasoning and execution systems, with reasoning models becoming the new paradigm [18][83] Open-Source vs Closed-Source Models - Open-source models are no longer seen merely as alternatives to closed-source models; they have carved out unique positions and are often preferred in specific scenarios [6][70] - By the end of 2025, it is expected that open-source models will account for approximately one-third of total usage, reflecting a more integrated approach by developers who utilize both types of models [5][70] - The dominance of DeepSeek is diminishing as more open-source models enter the market, leading to a diversified landscape where no single model is expected to exceed 25% of token usage by the end of 2025 [13][77] Model Characteristics and Trends - The report identifies a shift towards medium-sized models, which are gaining market favor, while small models are losing traction [16][80] - The classification of models is as follows: large models (700 billion parameters or more), medium models (150 to 700 billion parameters), and small models (less than 150 billion parameters) [20][85] - The usage of reasoning tokens has surpassed 50%, indicating a significant evolution in how AI models are utilized for complex tasks [18][83] User Behavior and Model Utilization - AI model usage has evolved from simple tasks to more complex problem-solving, with user prompts increasing in length and complexity [27][92] - The concept of "crystal shoe effect" is introduced, where certain models lock in a core user base due to their unique capabilities, making it difficult for competitors to attract these users later [55][120] - Programming and role-playing have emerged as the primary use cases for AI models, with programming queries rising from 11% to over 50% [27][100] Market Dynamics - The report notes that the paid usage share of AI in Asia has doubled from 13% to 31%, while North America's share has fallen below 50% [129] - English remains the dominant language in AI usage at 82%, with Simplified Chinese holding nearly 5% [129] - The impact of model pricing on usage is less significant than anticipated, with a 10% price drop leading to only a 0.5%-0.7% increase in usage [129]