孙正义与Open AI的萨姆·奥尔特曼握手（2025年2月、reuters） 软银集团2025年4～12月净利润增至同期的5倍，其中来自OpenAI的投资收益贡献巨大。OpenAI正向软 银集团等投资方寻求巨额追加注资。孙正义对OpenAI情有独钟，但AI开发竞争正日趋激烈…… 软银集团（SBG）2月12日发布的2025年4～12月合并财报（国际会计准则）显示，净利润达3.1726万亿 日元，为上年同期的5倍。其投资的美国OpenAI估值上升推高了利润。OpenAI正向软银集团等投资方寻 求巨额追加注资，软银集团会长兼社长孙正义"全力押注AI（人工智能）"的决心面临真正考验。 软银集团4～12月期的净利润创历史新高。来自OpenAI的投资收益达2.7965万亿日元，贡献巨大。软银 集团通过2025年12月的注资，对OpenAI累计投资达346亿美元，成为出资比例约为11%的大股东。持有 资产的市值达544亿美元。 在2月12日的软银集团财报说明会上，公司首席财务官（CFO）后藤芳光表示："我们高度评价OpenAI 所发挥的作用"。 软银再次明确提出，要力争成为超越人类智慧的人工超智能（ASI）时代的"全球第一 ...

Group 1 - The core viewpoint of the report indicates that the Chinese gaming market is showing steady growth, with actual sales revenue reaching 32.468 billion yuan in January 2026, reflecting a month-on-month increase of 2.99% and a year-on-year increase of 4.47% [1][2] - The client game market generated actual sales revenue of 8.236 billion yuan, with a month-on-month growth of 7.77% and a year-on-year growth of 23.46%, driven primarily by the launch of new products [5][10] - The mobile game market achieved actual sales revenue of 22.621 billion yuan, with a month-on-month increase of 1.8% but a year-on-year decrease of 1.52%, indicating a mixed performance influenced by new releases and high competition [10][9] Group 2 - The overseas market for self-developed games saw actual sales revenue of 2.077 billion USD, with a month-on-month decrease of 2.5% but a year-on-year increase of 24.05%, highlighting the dual engine of mature products and new games driving growth [27][30] - The top-performing mobile games in January 2026 included "Peacekeeper Elite" and "Dungeon & Fighter: Origin," which saw significant revenue increases due to new content and updates [14][16] - The new game "Arknights: Endfield" achieved over 500 million yuan in its first month, showcasing the effectiveness of innovative gameplay and strong IP influence in attracting players [24][26]

前述公告显示，为建设新CAR-T药物生产基地，恺兴诊断已委托上海金工对位于上海市的一处工业厂房 实施定制装修工程（包括装修、改造、设施设备安装等），预计总工期为14个月。在交付后，科济药业 在早期仅需按季度支付服务费（前10年年均不超过1000万元，第11至20年年均不超过1800万元），并在 交付后的20年内按照协议约定对全部定制装修设施进行回购，回购价格按最终投资额确定，上限为3.7 亿元。 2025年3月，科济药业创始人兼首席执行官李宗海在接受时代财经专访时就曾表示，未来或在舒瑞基奥 仑赛正式上市后考虑扩充产能，以满足实体瘤这一大适应症的市场需求。 李宗海表示，舒瑞基奥仑赛在中国开展的针对晚期胃癌/食管胃结合部腺癌的确证性II期临床试验结果已 经达到主要疗效终点，其治疗效果甚至优于现有疗法。2024年国家癌症中心发布的数据显示，2022年， 胃癌已成为我国发病率和死亡率分别居于第5位和第3位的恶性肿瘤，预估新发病例和死亡病例分别为 35.87万例和26.04万例。由于国内胃癌的早期筛查仍未全面普及，我国近一半的胃癌患者发现时已是晚 期，预后不良。 时代财经注意到，此次3.7亿元投资并非由科济药业一次性现 ...

Core Insights - Apple has fixed a zero-day vulnerability (CVE-2026-20700) that has existed since the iOS 1.0 era, affecting the core infrastructure of iOS, specifically the dynamic linker (dyld) [1][2][7] - The vulnerability allows attackers with memory write capabilities to execute arbitrary code, and it has been confirmed to have been exploited in the wild [2][8] - The dyld vulnerability is critical as it serves as a gatekeeper for all apps, and its exploitation could lead to complete control over the device [3][6] Vulnerability Details - The dyld acts as a "doorman" for the iOS system, loading necessary dynamic libraries and ensuring security isolation for apps [2][3] - Attackers can bypass security checks and gain unauthorized access, potentially leading to "zero-click" or "one-click" attacks, particularly targeting high-value individuals [3][6] - The vulnerability was discovered by Google's Threat Analysis Group and is part of a more complex attack chain involving other vulnerabilities, such as those in WebKit [2][6] Historical Context - The existence of the dyld vulnerability for over a decade raises questions about why it remained undetected until now, attributed to the stability and complexity of core system components [5][6] - Such vulnerabilities often do not appear in isolation and are typically part of a larger exploit chain, similar to techniques used by commercial surveillance firms [6] Other Security Issues - In addition to CVE-2026-20700, Apple has addressed multiple other security issues in the latest iOS and iPadOS updates, including vulnerabilities that could lead to root access and sensitive data leaks [8] - Two high-severity vulnerabilities (CVE-2025-14174 and CVE-2025-43529) were also reported, highlighting ongoing concerns about memory safety issues even in secure mobile platforms [9][10]

Group 1 - OpenClaw's founder Peter Steinberger shared his experience of sudden fame, including challenges such as name change demands from Anthropic and harassment from the crypto community [1][2] - The project is currently in a loss-making state, relying on donations and limited corporate support, raising concerns about its sustainability [1] - Peter received acquisition and collaboration offers from major companies like OpenAI and Meta, but he insists on maintaining the project's open-source nature [1] Group 2 - Peter expressed views on the AI industry, stating that many AI safety concerns are exaggerated and that AI will not replace core creative roles of programmers [2] - He highlighted the importance of efficient collaboration in AI development, warning against the pitfalls of overly complex agent orchestration [2] Group 3 - The renaming process of the project was fraught with difficulties, including domain name acquisition and the need for a rapid response to legal pressures [4][6] - Peter faced significant stress during the renaming, nearly abandoning the project due to the overwhelming challenges [13][14] - The final name, OpenClaw, was chosen after a meticulous and secretive planning process to avoid further issues [16][17] Group 4 - Peter discussed the MoltBot incident, describing it as an artistic expression rather than a genuine security threat, emphasizing the need for better public understanding of AI [20][21] - He noted that the safety concerns surrounding MoltBot were largely unfounded and stemmed from misunderstandings about AI capabilities [22] Group 5 - Peter is actively addressing security concerns within the project, collaborating with VirusTotal to scan skills before deployment [23] - He acknowledged that while software will always have bugs, the project has benefited from community feedback and contributions to improve security [24] Group 6 - The conversation highlighted the evolving nature of AI models, with Peter noting that as models become smarter, their attack surfaces may shrink, but the potential damage from failures could increase [26][27] - He emphasized the importance of using robust models to mitigate risks associated with prompt injection and other vulnerabilities [26] Group 7 - Peter discussed the need for a cognitive shift in how developers interact with AI agents, advocating for a design approach that aligns with the agents' logic and capabilities [29][32] - He stressed the importance of understanding how agents perceive tasks and the necessity of guiding them effectively to achieve desired outcomes [33][35] Group 8 - The future of AI development is seen as a blend of personal assistants and collaborative coding partners, with an emphasis on creating a seamless interaction experience [54][55] - Peter believes that the current interfaces for interacting with AI are still in their infancy and will evolve significantly over time [57]

【导读】OpenAI深夜突袭，GPT-5.3-Codex-Spark正式炸场。核心卖点只有一个：快！每秒1000个token，让代码生成告别加载条。联手Cerebras怪兽级 硬件，物理外挂直接拉满。这不再是简单的工具升级。而是一场关于速度的暴力美学。 OpenAI又深夜炸场了。 GPT-5.3-Codex-Spark正式发布！ 这次不讲大道理，只讲一个字：快。 到底有多快，看一下官方的演示： 它是GPT-5.3家族里的「闪电侠」。 也是OpenAI首个专为实时编程设计的模型，OpenAI称之为「超高速模型」。 芯片巨头Cerebras。 它的生成速度超过每秒1000个token！ 这是什么概念？ 你刚敲完回车，代码已经写完了。 体感接近「瞬时响应」。 这次OpenAI找了个强力外援。 大家写代码最烦什么？肯定是等待。 Spark的出现就是为了干掉等待。 Spark跑在Cerebras的Wafer Scale Engine 3上。 这不是普通的GPU堆叠。 这是专为低延迟设计的顶级硬件。 为了配合这股怪力，OpenAI还重写了底座。 他们引入了持久的WebSocket连接。 往返开销降低了80%。 首个字符出 ...

Core Viewpoint - MiHoYo terminated its cooperation with Shanghai Huiye Law Firm due to a conflict of interest involving a lawyer who assisted a competing company, Envision Energy, in a lawsuit against MiHoYo [1][7][15]. Group 1: Incident Background - The conflict originated in June 2025 when Envision Energy filed a lawsuit against MiHoYo, claiming that a fictional antagonist in MiHoYo's game "Honkai: Star Rail" resembled their brand, causing reputational damage [4][6]. - MiHoYo counter-sued Envision Energy in December 2025 for commercial defamation [6]. - By February 2026, Envision Energy withdrew its lawsuit, which was approved by the court [7]. Group 2: Legal and Ethical Implications - The lawyer from Huiye Law Firm, who was MiHoYo's long-term legal advisor, was found to have collected evidence for Envision Energy despite knowing the conflict of interest, violating legal regulations [7][14]. - The Shanghai Judicial Bureau warned the lawyer for privately accepting a commission, which is prohibited under the Lawyer Law [7][14]. - Huiye Law Firm acknowledged the incident highlighted deficiencies in their dynamic conflict of interest monitoring system and pledged to improve compliance standards [15]. Group 3: Industry Context - The incident reflects the challenges within large law firms where lawyers operate with significant autonomy, often leading to potential conflicts of interest [11][13]. - The legal services market in China is characterized by a commission-based distribution model, which can complicate conflict management [11][13]. - MiHoYo's proactive stance in protecting its intellectual property rights indicates a broader trend in the gaming industry to safeguard intangible assets [15].

虽然他已经离开人世，但留给了世人三笔"隐形财富"——寒门能出贵子的信念、实干兴邦的坚守、予人玫瑰的担当。 每一笔，都值得我们反思。 最近农牧圈最大的事儿，就是大北农董事长邵根伙病逝，享年60岁。 比起那些天天上热搜的商业大佬，邵根伙太"不起眼"——没有铺天盖地的通稿，没有豪车豪宅的排场，最常见到的就是他穿着朴素参加学术交流。 可这个"不起眼"的人，藏着太多反差感： 他是中国第一个猪营养学博士，凭一己之力填补国内空白。 他出身贫寒，白手起家创立大北农集团，2025年胡润百富榜显示身家105亿，但背包用到磨损都舍不得换，生活极度节俭。 很多人说他"抠"，但他的"抠"只对自己。对农业、寒门学子和需要帮助的人，他非常大方。 第一笔"隐形财富"：寒门能出贵子 现在网上总说"寒门再难出贵子"，说阶层固化、努力无用。 不可否认，寒门子弟脱颖而出更难，但邵根伙用一生证明，出身从不是宿命。刻在骨子里的坚韧，才是逆袭的底气。 这就是他留给我们的第一笔隐形财富。 邵根伙的起点，比很多人都低得多。1965年，他出生在浙江金华澧浦镇的农村，幼年丧父，全家生计全靠母亲艰难维系，家里条件并不好。 他清楚自己没有退路，只有读书才能改变命运。 ...

最近在网上冲浪，学到了不少关于智能体的黑话。比如说一家公司说我们有agentic workflow，实际上就是写几个for循环调用API；说自己正在做AI agent，不过是把一群只有金鱼记忆、胡言乱语的AI凑在一起开会，直到掏空钱包。 没想到被业内人士看好的企业智能体，却成了老板眼中不顶用、员工眼中纯忽悠的存在，这就很讽刺了。 毕竟2025年初，智能体爆火之后，行业内的普遍共识是，通用智能体如Manus，这类全能型AI助手不懂业务，无法满足企业对高可靠、低容错的业务要 求。而懂业务、能干活、有行业沉淀的企业智能体，能嵌入企业内部流程之中，被认为是智能化转型的主力。 没想到一年过去了，通用智能体依然是长盛不衰、自带流量的技术网红，从Manus到clawbot，老板一刷热搜就两眼发亮，恨不得立刻将其请进公司。 而企业智能体，可以用雷声大、雨点小来形容。既不像通用智能体那样炫酷而全能，能让公司蹭上AGI的光环，又不如原来的RPA、企业系统更符合使用 习惯，员工还得从头学习，配合公司的转型表演，纷纷在社交媒体上吐槽。 这种想法，像不像很多老资历的职场人，一旦公司里空降了技术网红或明星高管，就觉得老板遇人不淑，自己怀 ...

文｜富充 编辑｜苏建勋 清华大学电子工程系长聘教授汪玉 北京智源人工智能研究院院长王仲远 阶跃星辰创始人&CEO 姜大昕 星海图创始人&CEO 高继扬 原力灵机联合创始人&CEO 唐文斌 阶跃星辰创始人&CEO姜大昕首先提出"ChatGPT时刻"的定义标准，是"零样本泛化"——即使给出从未见过的指令，AI也能回答问题完成任务——这正是 大语言模型所展现的能力。 但姜大昕旋即指出，因为具身智能的泛化要涉及场景、任务、操作物体等更多维度，所以机器人要达到这个标准还十分困难。 作为机器人创企的CEO，高继扬进一步解释了具身智能商业化落地的难点：大语言模型可以"模型即产品"，终端是手机电脑、渠道是互联网传播；具身智 能却必须穿过更长的产业链——整机、供应链、真机数据、线下交付，缺一不可。 具身智能正在等待自己的"ChatGPT时刻"。但关于这个时刻的具体定义，业内还充满非共识。 近日，原力灵机的技术开放日圆桌论坛上，5位AI界的一线产、学、研从业者把这个问题摊开，各自发表了见解。他们分别是： 基于以上种种待解决的问题，原力灵机联合创始人&CEO 唐文斌，给出了一个眼下更可抵达的"具身智能ChatGPT时刻"：先在一个 ...