OpenAI正在招聘数百名"前沿部署工程师"以赢得大型企业业务。 ...

（文章来源：第一财经） 据报道，OpenAI正在招聘数百名"前沿部署工程师"以赢得大型企业业务。 ...

Core Insights - OpenAI is establishing an advertising integrity team to support its upcoming ChatGPT advertising business, ensuring user trust and commercial compliance [1][3] Group 1: Team Formation and Responsibilities - The new team is categorized as a "0→1" startup team, focusing on designing an advertising safety assurance system, developing a Know Your Customer (KYC) system, and setting rules for ad placement and display [3] - The KYC system will verify advertisers' identities and assess risk levels, serving as a critical risk control measure against fraudulent ads and harmful content [3] - The establishment of this specialized team aligns with industry practices among tech giants to combat ad fraud and ensure brand safety [3] Group 2: Advertising Testing and Financials - OpenAI has confirmed plans to conduct small-scale advertising tests in the U.S. for the free and basic subscription versions of ChatGPT, with users expected to see related ads in the coming weeks [3] - Advertisers participating in the test must invest a minimum of $200,000 (approximately 1.389 million RMB), with OpenAI tracking ad clicks and impressions while exploring additional performance measurement methods [3] - An OpenAI spokesperson has confirmed the advertising test arrangements but has not provided further comments on the recruitment for the advertising integrity positions [3]

Investment Rating - The report does not explicitly provide an investment rating for the industry or companies involved. Core Insights - The report highlights significant advancements in Chinese large models and AI agents, marking a transition from "following" to "keeping pace" with global leaders in AI technology [14][24]. - The competitive landscape shows a clear distinction between domestic and international models, with domestic open-source models gaining substantial ground [23][47]. - The report emphasizes the importance of multi-modal capabilities and the emergence of AI agents in practical applications, particularly in programming and task planning [16][14]. Summary by Sections 1. Key Developments in 2025 - The report outlines three major phases of AI model evolution: the initial competition among models, the explosion of multi-modal capabilities, and the rise of AI agents [14][16]. - Notable models such as Kimi-K2.5-Thinking and Qwen3-Max-Thinking have emerged as leaders in specific tasks like code generation and mathematical reasoning [18][24]. 2. Annual Evaluation Results and Analysis - The 2025 annual evaluation ranks Claude-Opus-4.5-Reasoning as the top model globally, followed by Gemini-3-Pro-Preview and GPT-5.2(high) [23][45]. - Domestic models like Kimi-K2.5-Thinking and Qwen3-Max-Thinking are positioned fourth and sixth, indicating a strong competitive stance [23][45]. - The report notes that domestic models are rapidly closing the gap with international counterparts, particularly in code generation and reasoning tasks [24][48]. 3. SuperCLUE Model Quadrant and Capability Landscape - The report presents a model quadrant that categorizes models based on their capabilities in reasoning and application, highlighting the emergence of "technical leaders" and "practical leaders" in the domestic market [38][39]. - The capability landscape indicates that while domestic models excel in certain areas, they still face challenges in hallucination control and precise instruction adherence [42][48]. 4. Comparative Analysis of Domestic and International Models - The analysis reveals that closed-source models dominate the top rankings, with significant advantages in reasoning and instruction-following tasks [74][80]. - Domestic open-source models are noted for their rapid advancements, particularly in coding tasks, where they have begun to outperform some international models [56][84]. - The report emphasizes the structural differences between domestic and international models, with domestic models showing a strong trend towards open-source development [24][47].

美股频道更多独家策划、专家专栏，免费查阅>> 此外，Similarweb的数据显示，ChatGPT过去一年访问次数由38亿次增至57亿次，增长仍达50%。同期 Gemini增速则高达6.47倍，访问次数由2.68亿次飙升至20亿次。 2月5日，数据分析公司Apptopia的数据显示，2025年1月至2026年1月期间，OpenAI旗下ChatGPT的市场 份额正被蚕食，市占率由69.1%降至45.3%。与此同时，谷歌旗下Gemini的市占率由14.7%攀升至 25.1%，马斯克的xAI旗下Grok份额由1.6%激增至15.2%，两者合共占40.3%份额。AI聊天机器人市场已 从一间独大转向多强争霸。 责任编辑：栎树 ...

Core Insights - The surge in capital expenditures (capex) is driving record-high tech valuations, with investors reacting positively when they believe the spending is justified [1] - Alphabet plans to significantly increase its capex to between $175 billion and $185 billion in 2026, a substantial rise from $91.4 billion in 2025 and $52.5 billion in 2024 [2] - The competitive landscape in AI is prompting major players, including Alphabet, to invest heavily in innovation and infrastructure [3] Company-Specific Insights - Alphabet's Q4 capex investment reached $27.9 billion, reflecting its commitment to maintaining a competitive edge in AI [2] - CEO Sundar Pichai expressed concerns about the challenges associated with the rapid increase in capex, including the time required to convert investments into operational data centers and address power and supply chain constraints [4][5] - Despite the significant increase in spending, Pichai acknowledged ongoing supply constraints that will impact the company's AI and cloud services [5][6] Industry Trends - The tech industry is experiencing a fierce competition in AI, with companies like Meta also announcing substantial increases in capex, indicating a broader trend of escalating investments in AI infrastructure [3][6] - The relentless pace of innovation in AI is driving companies to continuously adapt and invest, highlighting the dynamic nature of the industry [3]

智通财经APP获悉，中金发布研报称，2025年全球大模型技术能力向前演进，逐步攻克生产力场景，在 推理、编程、Agentic以及多模态等能力方向取得明显进步，但模型通用能力在稳定性、幻觉率等方面 仍存在短板。展望2026年，该行认为大模型在强化学习、模型记忆、上下文工程等方面将取得更多突 破，从短context生成到长思维链任务，从文本交互到原生多模态，并向实现AGI长期目标更进一步。 中金主要观点如下： 强化学习重要性提升，成为解锁模型高级能力的关键 强化学习的引入提高了模型的智能上限，让模型可以更有逻辑、更符合人类偏好进行思考和推理，其本 质是"自我生成数据+多轮迭代"，强化学习的关键在于大规模算力+高质量数据。海外OpenAI、Gemini 等模型厂商对于强化学习非常重视，国内DeepSeek、阿里千问等也在跟进，该行预计2026年海内外模 型厂商强化学习占比将进一步提升。 持续学习、模型记忆、世界模型等新路线将迎来核心突破 持续学习和模型记忆本质上是解决大模型"灾难性遗忘"问题，让模型具备选择性记忆机制。Google提出 的Titans、MIRAS、Nested Learning等算法和架构核心是让模 ...

2月4日，现货黄金重回5000美元关口，现报5039.66美元/盎司。国内黄金饰品价格对比显示，多家黄金珠宝品牌当日公布的境内足金首 饰价格有所上调，周生生报价1600元/克，较前一日1498元/克涨102元；周大福报价1566元/克，较前一日1495元/克涨71元；老凤祥报价 1576元/克，较前一日1518元/克涨58元；老庙黄金报价1571元/克。（智通财经） 今日头条 多家中小银行宣布上调存款利率 随着春节脚步临近，多地中小银行近期宣布上调存款利率，例如湖南新晃农商行、陕西镇巴农商行、山西临县农商行等，均在1月下旬 将部分期限存款利率上浮。不仅如此，部分银行还推出"限时专享"的新年专属存款产品。不过，利率上调并非统一趋势，在市场利率 下行的背景下，亦有部分银行近期选择下调各期限存款利率。有业内人士指出，考虑到净息差处于低位并且仍然有下行压力，银行需 要降低负债成本稳定净息差等因素，未来存款利率大概率稳定在低位或者有小幅下调。（每日经济网） 国内品牌金饰克价普遍上调，周生生克价1600元 义乌世界杯订单生产进入"冲刺期" 2026年美加墨足球世界杯将在6月启幕。从2025年5月开始，义乌市场上的世界杯订 ...

Market Dynamics - The U.S. stock market showed mixed results, with the Dow Jones up by 0.53%, while the S&P 500 and Nasdaq fell by 0.51% and 1.51% respectively, with the Nasdaq breaking below its 100-day moving average [1] - Southbound capital in Hong Kong saw a net inflow of approximately 13.373 billion HKD, marking the highest single-day inflow since January 22, with Tencent and Alibaba receiving net purchases of about 2.231 billion HKD and 1.171 billion HKD respectively [4] - The latest data indicates that the number of newly opened margin trading accounts in January 2026 reached 190,500, a 29.5% increase from December 2025, and a 157.09% increase year-on-year [4] Commodity Prices - WTI crude oil futures rose by 3.05% to $65.14 per barrel, while Brent crude increased by 3.16% to $69.46 per barrel [2] - COMEX gold futures increased by 1.04% to $4,986.4 per ounce, and silver futures rose by 5.36% to $87.765 per ounce [2] Industry Developments - Domestic private commercial rocket company Dongfang Space is expected to complete the first flight of its "Gravity II" medium-sized liquid reusable launch vehicle by mid-2026 [5] - The National Space Computing Center, part of the "Star Computing" plan by Guoxing Aerospace, aims to establish a super space computing network consisting of 2,800 satellites, with the first group already launched [5] - Tesla's Shanghai factory delivered 69,129 vehicles in January, marking a 9% year-on-year increase, continuing a trend of growth in wholesale sales for three consecutive months [16] Technological Innovations - The brain-machine interaction and human-machine integration laboratory has developed the first brain-controlled acupuncture and neuro-rehabilitation equipment platform in China, which is currently undergoing clinical trials [9] - Tencent's AI Infra team launched an open-source high-performance LLM inference core operator library, achieving a 30% improvement in model inference performance [7] Regulatory and Policy Updates - The People's Bank of China emphasized the need for structural monetary policy tools to support key areas such as expanding domestic demand and technological innovation [10] - The Ministry of Industry and Information Technology highlighted the importance of AI in manufacturing, aiming to strengthen application foundations and key technologies [11]

本文字数：2725，阅读时长大约4分钟 作者 | 第一财经 樊志菁 随着投资者关注到人工智能技术的快速发展或颠覆某些行业商业模式，同时相关板块的热潮可持续性担忧再次涌现，本周美股部分成长性板块面临巨大 考验，纳指和标普500指数进一步脱离历史高位。 与此同时，资金开始撤出转向工业、运输等避险板块，等待形势进一步明朗。 2026.02. 05 市场抛压加剧 受软件板块与大型科技股走弱拖累，美股本周面临巨大抛售压力。 导火索在于人工智能新贵Anthropic推出全新合规模型对部分行业商业模式的冲击和 威胁。 投资者对大型科技股的集中度风险，以及市场对人工智能的狂热情绪推高板块估值、导致估值泡沫的担忧有所升级。软件股遭遇抛售潮，道琼斯市场数 据显示，道富标普软件与服务交易所交易基金ETF两天下挫近8%，近两周跌幅扩大至15%。 据华尔街两家大型基金的消息人士透露，对冲基金正加大对软件类股的做空押注，成为今年以来该板块遭遇惨烈抛售的一大推手。S3 Partners的数据显 示，随着软件行业总市值蒸发1万亿美元，卖空者今年迄今已从软件类股的交易中斩获240亿美元的暴利。 英伟达CEO黄仁勋在思科的一场活动中驳斥AI引发 ...