编辑｜冬梅 北京时间昨日深夜，OpenAI 正式发布了新一代智能体编码模型 GPT-5.2 Codex，并在官网同步发布 技术博客，对该模型的定位、能力改进及部署方式进行了说明。 OpenAI 重磅发布 GPT-5.2 Codex 据介绍，GPT-5.2 Codex 基于通用模型 GPT-5.2 构建，并针对"智能体编码"（Agentic Coding）场景 进行了专门优化，主要面向复杂的软件工程任务。相较此前版本，新模型在长程任务执行、大规模代 码变更、Windows 原生环境支持以及网络安全相关能力等方面进行了系统性改进。 在工程能力层面，OpenAI 表示，GPT-5.2 Codex 通过引入原生的上下文压缩（compaction）机制， 提高了对超长上下文的理解与利用效率，使模型在跨文件、跨模块的长期编码任务中具备更稳定的表 现。同时，该模型在代码重构、迁移等涉及大规模改动的场景下，整体可靠性和一致性有所提升。 安全能力也是此次更新的重点方向之一。OpenAI 在博客中提到，随着模型在推理与工具调用能力上 的增强，其在网络安全领域的适用性也随之提升。 官方披露，就在上周，一名安全研究人员使用 GPT- ...

Group 1 - OpenAI is planning an unprecedented capital operation, aiming for financing of up to $100 billion to strengthen its dominance in the AI field and cover high model training costs [1] - The company is reportedly discussing a valuation of approximately $750 billion with investors, with potential financing reaching hundreds of billions, possibly up to $100 billion [1] - If successful, this financing would represent a 50% increase over OpenAI's last valuation from October of the previous year, further enhancing its cash reserves for AI model training and operations [1] Group 2 - Amazon is in deep discussions to invest at least $10 billion in OpenAI, which would support OpenAI's commitment to rent Amazon Web Services (AWS) servers for $38 billion over the next seven years [2] - The agreement would involve OpenAI using Amazon's proprietary Trainium chips, positioning Amazon as a significant investor alongside tech giants like Nvidia [2] - This financing activity highlights the escalating arms race in the generative AI sector, indicating a complex competition among Silicon Valley giants regarding computing power, exclusive agreements, and market share [2]

Core Viewpoint - The article emphasizes that electricity has become a critical constraint on the development of artificial intelligence (AI), with the demand for power surging due to the exponential growth of AI models and data centers [2][14]. Group 1: AI Power Consumption - ChatGPT consumes approximately 2.9 watt-hours per response, which is nearly ten times the energy used by a traditional Google search, leading to a daily consumption of over 500,000 kilowatt-hours [3][9]. - The energy required for training AI models has significantly increased, with GPT-3 consuming 1,287 megawatt-hours for a single training session, enough to power 3,000 Tesla vehicles for 200,000 miles [9][10]. - The projected AI computing power in the U.S. is expected to require about 1,269 terawatt-hours by 2030, accounting for 22% of the total electricity consumption [10][12]. Group 2: Electricity Supply Challenges - The U.S. electricity grid is aging, with 70% of transformers exceeding their 25-year design life, leading to a low load reserve margin of only 20% [10][11]. - The average outage duration for U.S. users reached 662.6 minutes in 2024, a year-on-year increase of 80.74%, with states like Virginia and Texas experiencing even longer outages [11]. - The demand for electricity from AI is characterized by "pulse-like" spikes, which poses significant challenges to grid stability [10][11]. Group 3: Renewable Energy Solutions - Solar power combined with energy storage is viewed as the most viable solution to meet the growing electricity demands of AI, given its economic advantages and shorter construction timelines compared to other energy sources [15][17]. - The cost of solar power generation is the lowest among various energy sources, with prices ranging from $0.038 to $0.078 per kWh, making it an attractive option for data centers [17][18]. - Major tech companies like Google, Microsoft, and Amazon have set ambitious goals for achieving 100% renewable energy for their data centers by 2030, indicating a strong commitment to sustainable energy solutions [20][21].

在这一轮大规模融资计划背后，软银集团已同意向OpenAI投资300亿美元。为了资助这一赌注，软银在 上个月出售了价值58亿美元的英伟达股份。预计到今年年底，OpenAI将从软银处获得剩余的225亿美元 计划融资。 OpenAI正计划在一轮新的融资中募集高达1000亿美元资金，以支持其雄心勃勃的增长计划。 据《华尔街日报》19日援引知情人士透露，这轮融资目前处于早期阶段。如果OpenAI能顺利募集到全 额目标资金，其公司估值可能高达8300亿美元。该初创公司计划最快在明年第一季度末完成这轮融资， 不过交易条款仍可能发生变化。 目前尚不清楚是否有足够的投资者需求来通过这一目标。此次融资将是OpenAI在公开市场对人工智能 支出的狂热情绪消退后面临的最大考验之一。尽管对人工智能泡沫的担忧已经对许多相关科技股造成压 力，OpenAI仍需积极寻求大量资本以保持竞争力。 OpenAI首席执行官Sam Altman此前已在全球范围内寻找投资者以建立资本池，据《华尔街日报》此前 报道，该公司正在权衡潜在的首次公开募股（IPO）。为了在快速变化的市场中构建其人工智能模型， OpenAI需要极其充裕的资金支持。 巨额注资与战略布 ...

责任编辑：王永生 日经指数上涨0.9%，至49，443.48点，追随华尔街隔夜的涨势，该涨势受美国通胀数据低于预期提振。 IG的首席市场分析师Chris Beauchamp表示："投资者当然热切希望看到支持美联储明年将继续降息这一 论点的数据。"这位首席市场分析师补充说，该CPI数据"很好地印证了这一点，但风险在于，围绕价格 趋势的不确定性将不利于明年进一步降息的希望。"在日本基准股指中，软银集团、武田制药和马自达 汽车涨幅居前，分别上涨3.5%、3.3%和3.35%。 日经指数上涨0.9%，至49，443.48点，追随华尔街隔夜的涨势，该涨势受美国通胀数据低于预期提振。 IG的首席市场分析师Chris Beauchamp表示："投资者当然热切希望看到支持美联储明年将继续降息这一 论点的数据。"这位首席市场分析师补充说，该CPI数据"很好地印证了这一点，但风险在于，围绕价格 趋势的不确定性将不利于明年进一步降息的希望。"在日本基准股指中，软银集团、武田制药和马自达 汽车涨幅居前，分别上涨3.5%、3.3%和3.35%。 责任编辑：王永生 ...

Market Overview - The S&P 500 index has experienced four consecutive declines, primarily driven by large technology stocks[1] - Major U.S. indices collectively fell, with the Nasdaq down nearly 2%[9] - The S&P 500 closed at 6,800, reflecting a decline of 1.16% for the day and a year-to-date increase of 14.28%[3] Hong Kong Market Performance - The Hong Kong stock market showed a significant recovery in the afternoon, with major indices rebounding after two consecutive days of decline[9] - Large technology stocks such as Meituan and Kuaishou rose nearly 2%, while Tencent and Alibaba increased over 1%[9] - The financial sector saw a collective rise, with China Life Insurance up over 4% and CITIC Securities up over 3%[9] Sector Highlights - Precious metals stocks strengthened, with China Silver Group rising over 7% due to a weaker dollar and expectations of interest rate cuts, pushing silver prices above $66 per ounce[9] - The aviation sector saw gains, with China Southern Airlines up over 5% as ticket bookings surged ahead of the New Year holiday[9] - The lithium battery sector also rose, driven by a significant increase in lithium carbonate futures prices, with Tianqi Lithium up 5.83%[9] Economic Indicators - Japan's November exports grew by 6.1% year-on-year, exceeding market expectations, with a trade surplus of 322.2 billion yen[13] - The U.S. Treasury yields have shown fluctuations, impacting market sentiment and investment strategies[18]

（原标题：亚太股市普跌！AI医疗概念股集体走强 商业航天全线爆发） 【导读】又来找原因啦 大家好，不好意思，又来找下跌原因了。 隔夜美股科技板块大跌，泰勒以为这次我们会走出独立行情，没想到还是想多了！ A股下跌 12月18日，A股指数普遍下跌，仅沪指一枝独秀。截至收盘，沪指涨0.16%，深成指跌1.29%，创业板指 跌2.17%。 两市共2845只个股上涨，73只个股涨停，2416只个股下跌。 AI医疗概念股集体走强，华人健康、塞力医疗等涨停。消息面上，近日蚂蚁集团正式推出AI健康应 用"蚂蚁阿福"，月活跃用户规模超1500万，并登上苹果应用商店下载榜第二位。市场解读为AI技术与健 康管理场景的深度融合有望开启新增长空间，从而提振了板块整体情绪。 商业航天全线爆发，盛洋科技、天箭科技等涨停。有股民调侃称，"犒赏经济"重磅提出，商业航天迎来 重大利好。 零售、IP经济等消费股拉升，百大集团6连板，上海九百等涨停。 下跌方面，算力硬件股下挫，生益科技、景旺电子等跌超5%。 有分析师指出，近几周一个更清晰的叙事正在形成：推动本轮牛市的超大市值科技股，可能正在失去单 独"扛起市场"的能力。"市场对该板块的信心正面临挑 ...

Market Overview - Asian stocks ended mostly lower, influenced by the S&P 500 and Nasdaq reaching three-week lows due to concerns over high stock valuations [1] - The dollar remained firm in Asian trade ahead of multiple central bank meetings [2] - Oil prices increased due to supply disruption concerns from Venezuela and Russia [2] Regional Indices - China's Shanghai Composite index rose by 0.16 percent to 3,876.37, as China Vanke initiated a second meeting with bondholders to extend debt payments [2] - Hong Kong's Hang Seng index finished slightly higher at 25,498.13 after a volatile session [3] - The Nikkei average fell by 1.03 percent to 49,001.50, while the broader Topix index decreased by 0.37 percent to 3,356.89 [4] - Seoul's Kospi average dropped by 1.53 percent to 3,994.51, driven by concerns over AI sector profitability [5] - Australian markets managed to recover slightly, ending marginally higher, while New Zealand's S&P/NZX-50 index dipped by 0.29 percent to 13,256.77 despite strong Q3 GDP growth [6] Sector Performance - Semiconductor and AI-related stocks in the U.S. faced significant losses, with the Nasdaq Composite down by 1.8 percent and the S&P 500 down by 1.2 percent [7] - LG Energy Solution's stock plummeted by 8.9 percent following Ford Motor's cancellation of a significant battery agreement [5]

Market Overview - The Nikkei index fell 1.03% to 49,001.5, marking its lowest close since November 25, while the broader Topix slipped 0.37% to 3,356.89 [1][8] - Wall Street's main indexes also closed lower, with the S&P 500 and Nasdaq reaching three-week lows due to concerns about the artificial intelligence trade impacting technology stocks [1][8] AI and Data Center Concerns - Market concerns regarding the profitability of AI-related businesses and data centers in the U.S. have become more pronounced [2] - Japan is experiencing similar issues, with numerous data centers under construction in the country [3][8] Company-Specific Movements - SoftBank Group's shares fell 3.76% following Oracle's 5.4% drop, attributed to a report indicating that Blue Owl Capital will not support a $10 billion deal for Oracle's next data center facility [6][8] - The Stargate project, a significant U.S. data center development initiative involving SoftBank, Oracle, and OpenAI, is central to these developments [6][8] - Chip-related companies such as Advantest and Fujikura saw declines of 3.32% and 3.42%, respectively [6][8] Financial Sector Performance - Financial stocks declined ahead of the Bank of Japan's policy meeting, with expectations of an interest rate hike; Mitsubishi UFJ Financial Group shares fell 1% and Mizuho Financial Group lost 0.92% [7][8] - In contrast, software testing firm Shift rose 5.09%, becoming the top-percentage gainer on the Nikkei, while railway operator Keisei Electric Railway increased by 4.57% [7][8] Index Performance - The Topix's value share index decreased by 0.13%, while the growth share index fell by 0.63% [7][8]

Core Viewpoint - Oracle's stock price has dropped by 5.4% due to concerns over funding issues related to its data center projects, reflecting a shift in market sentiment from future revenue potential to current financial costs and risks [1][2][3]. Group 1: Financial Performance and Market Reaction - Oracle's future order total (RPO) surged to $455 billion, leading to a 36% stock price increase and a market cap rise of $244 billion, but subsequent RPO growth to $523 billion resulted in an 11% post-market stock drop, indicating a change in investor focus [2][3]. - The company's cloud business revenue increased by 52%, and it signed a significant $300 billion agreement with OpenAI, positioning Oracle at the center of AI demand [6][10]. - Despite strong revenue growth, Oracle's capital expenditures have risen significantly, leading to a negative free cash flow of approximately -$10 billion, marking the first negative cash flow since 1992 [9][10]. Group 2: Debt and Financial Risks - Oracle's total debt has exceeded $100 billion, with projections suggesting it could approach $300 billion by 2028 when including long-term lease obligations [10]. - The company has become the largest issuer of investment-grade bonds among non-financial firms, raising $18 billion through a bond issuance, the largest in tech history [10]. - Concerns over Oracle's ability to manage its debt are reflected in rising credit default swap (CDS) costs, which have reached their highest levels since the 2009 financial crisis, indicating increased market anxiety about the company's financial stability [13]. Group 3: Strategic Partnerships and Market Position - The "Gateway to the Stars" initiative, a $500 billion project announced with OpenAI and SoftBank, has transformed Oracle's role in the cloud market, although the execution of such large-scale projects remains under scrutiny [4][5]. - Oracle's reliance on a single major client, OpenAI, raises concerns about customer concentration risk, as the long-term contracts may limit flexibility in adjusting to market changes [13][14]. - Management has attempted to reassure investors by suggesting that even without OpenAI's continued partnership, alternative clients could be found, and that initial project costs would be lower than anticipated [14].