Core Insights - OpenAI has entered into a partnership with the U.S. government's General Services Administration to provide ChatGPT enterprise services at a cost of $1 per federal agency starting next year [1][3] - This collaboration follows the inclusion of OpenAI, Google, and Anthropic in the approved AI vendor list by the General Services Administration, streamlining the procurement process for government agencies [3] Group 1 - The agreement allows federal agencies to access AI tools from approved vendors without the need for separate negotiations, significantly simplifying the procurement process [3] - Anthropic is also planning to offer its products to the U.S. federal government at a similar price point of $1, although it remains unclear if other AI companies will follow suit with similar discounts [3] - The General Services Administration's Federal Procurement Service Commissioner, Josh Glomb, has encouraged other U.S. AI technology companies to collaborate with the government [3] Group 2 - In addition to the discounted ChatGPT enterprise service, OpenAI will provide federal agencies with 60 days of unlimited access to advanced model usage [3] - Federal employees will gain access to a new government user community and tailored training resources to help them integrate AI technology into their daily workflows [3]

Core Points - The Japanese government has officially requested Apple to open its iOS system to third-party browser engines, with a deadline for compliance set for December 2025 [1][4] - This initiative is based on the newly released guidelines of the "Mobile Software Competition Law," which expands upon the recently passed "Smartphone Law" targeting Apple's restrictive policies in the App Store [1][4] Group 1 - Apple currently requires that iOS versions of browsers like Chrome, Firefox, and Edge use its proprietary WebKit engine, which has been criticized for limiting market competition and hindering the performance and functionality of other browsers [4] - The Open Web Advocacy organization, involved in legal consultations, stated that Apple's enforcement of WebKit obstructs the development of independent browsers on iOS [4] - The new guidelines prohibit Apple from hindering the adoption of third-party browser engines and from setting unreasonable technical or financial barriers [4] Group 2 - The law ensures that browser developers have full access to critical system APIs, and any alternative APIs provided by Apple must not significantly underperform compared to those used by Safari or WebKit [4] - A notable requirement of the law is that devices must present a browser selection interface immediately upon activation, ensuring users can choose their preferred browser during the setup process [4][5] - Compared to related EU legislation, Japan's regulations are stricter regarding the timing and manner of user prompts, emphasizing the importance of user choice in the browser market [5]

Group 1 - Apple has announced that its next-generation chips will be produced by Samsung's semiconductor factory located in Austin, Texas, marking the first application of this innovative chip manufacturing technology globally and its debut in the United States [1][3] - The chips produced at this factory are expected to optimize the power consumption and performance of Apple's products, including the iPhone sold in global markets [3] - Industry observers speculate that the new generation chip is likely intended for the next iPhone's CMOS image sensor (CIS), although Samsung has not confirmed specific details [3]

Core Viewpoint - BeiGene reported significant financial improvements in Q2 2025, achieving profitability for the first time this year, driven by substantial revenue growth and effective cost management [1] Financial Performance - Total revenue for the first half of 2025 reached 17.518 billion RMB, representing a year-on-year increase of 46.0% [1] - Product revenue amounted to 17.360 billion RMB, with a year-on-year growth of 45.8% [1] - Operating profit was 799 million RMB, and net profit attributable to the parent company was 450 million RMB, both showing a turnaround from losses in the same period last year [1] Profitability and Guidance - This marks the first instance of BeiGene achieving comprehensive profitability in operating profit and net profit attributable to the parent company in 2025 [1] - The company updated its revenue guidance for the full year 2025, adjusting the expected range from 35.2 billion to 38.1 billion RMB to 35.8 billion to 38.1 billion RMB [1] - Gross margin guidance was updated to a mid-high range of 80% to 90% [1] Cash Flow - The net cash flow from operating activities, after deducting capital expenditures for fixed asset purchases, is expected to be positive for the full year [1]

Group 1 - The article highlights the success of Chinese AI company Z.ai (智谱) in developing its flagship model GLM-4.5 despite facing severe U.S. sanctions, demonstrating that U.S. export restrictions have been ineffective [1][2] - Z.ai's GLM-4.5 operates on just 8 NVIDIA H20 chips and meets or exceeds Western standards in coding, reasoning, and tool usage, showcasing China's advancements in AI technology [1] - The article argues that U.S. strategies to limit China's access to advanced hardware have not only failed but have also inadvertently encouraged Chinese companies to innovate and develop their own chips [1][2] Group 2 - China is adopting a top-down approach to integrate AI models into key industries and export AI and hardware solutions to developing countries, with over 1500 models developed, many of which are open-source [2] - Huawei's GPUs are rapidly filling the gap left by U.S. export controls, and research indicates that NVIDIA's global AI market share could decrease by 12% if restrictions remain in place [2] - The article suggests that the U.S. response to its failing strategy has been to continue imposing restrictions, while recent proposals emphasize the need to expand overseas supply and applications of AI [2][3] Group 3 - The export of AI hardware and models from China is viewed as a strategic asset, enhancing the global influence of Chinese software and values [3] - The success of Z.ai serves as evidence that sanctions cannot halt China's progress in AI, with future models expected to be developed more quickly and at lower costs [3]

Group 1 - A record number of Chinese companies are seeking to list in the U.S. this year, despite geopolitical tensions and stricter domestic listing rules [1][2] - In the first half of the year, 36 Chinese companies have already listed in the U.S., with many utilizing Special Purpose Acquisition Companies (SPACs) to expedite the process [1][2] - There are currently over 40 Chinese companies waiting to list on NASDAQ, including a mobile advertising service provider and a traditional Chinese medicine manufacturer [1] Group 2 - More than 100 Chinese companies are listed in the U.S., with a total market capitalization of approximately $1 trillion as of March this year [2] - The trend shows an increasing number of startups, particularly in the technology sector, seeking to go public via SPACs to accelerate fundraising [2] - The number of companies going public through SPACs nearly doubled last year to 57, and has already reached 76 this year [2]

来源：环球时报 解读言语，促进康复 当《环球时报》记者走进北京脑科学与类脑研究所的大楼时，发现不同实验室里的装备大相径庭——既 有闪烁的机房，也有正在测试的机械臂，还有生化实验室。这也反映了脑机接口的跨学科特性——它同 时涉及医学、机械、通信和计算等不同领域。北京脑科学与类脑研究所于2018年成立，并根据北京市前 瞻布局的"智能脑机系统增强计划"，牵头成立了北京芯智达神经技术有限公司（简称芯智达）。 芯智达业务发展总监李园在接受《环球时报》记者采访时介绍说，大脑活动最根本的工作机制是通过细 胞之间电信号的发放和传导，实现人体从最基础的吃饭、睡觉等生存动作，到产生复杂的记忆和情绪。 基于这个认识，如果通过工程技术手段，把大脑活动产生的电信号精准地采集出来，然后进行非常低延 时的传输，再通过外部的计算设备实现解码，就能解读出大脑电信号的含义。这是脑机接口最根本的原 理。 目前外界关注的脑机接口技术主要用于医学复健领域，例如可利用读取的大脑信息帮助瘫痪病人控制其 他设备（操作机械臂或控制电脑光标）。美国Neuralink公司的演示视频显示，手术后的瘫痪病人可以在 不移动任何四肢或手指、完全依靠神经信号的情况下移动光 ...

多年来，徐星在恐龙演化领域持续深耕，发表了300余篇论文，他和同事命名的恐龙新物种超过80个。 他的研究成果创造了多项世界第一，并于2023年当选中国科学院院士，他也被亲切地称为"恐龙院士"。 "有人说，我们是远古世界的探险者，能够了解、揭示前人知之甚少的远古世界发生的故事，能够找到 巨大的乐趣。"徐星表示，自己关于恐龙和鸟类关系的第一篇论文就是来自云南，"做基础科学的都有这 种体会，就是科学发现的乐趣，当然这个过程可能是艰苦的，但越是这样的环境，越可能有一些重要的 科学发现。这也是我们学科的魅力之一，就是深入大自然，了解大自然。我们为什么要做科学？最终的 目的就是和自然、社会融为一体。" 来源：人民日报客户端 8月6日上午，2025未来科学大奖揭晓。季强、徐星、周忠和三位科学家凭借"发现鸟类起源于恐龙的化 石证据"获得本年度"生命科学奖"。发布会现场，当颁奖人现场连线中国科学院院士、中国科学院古脊 椎动物与古人类研究所所长徐星时，他正戴着一顶大草帽，在云南省楚雄州元谋县山区进行野外考察。 记者注意到，为了信号更好些，徐星不得不站在地势比较高的山梁上，手里还捧着一枚化石。"这是我 们经常的装扮。"徐星说，在大 ...

来源：环球网 【环球网科技综合报道】8月6日消息，据forbes援引iPhone-Ticker报道，苹果公司将于 9 月 9 日举办 iPhone 17 系列发布会，与 iPhone 16 的发布日期一致。这一消息得到了至少一家当地运营商内部文件的 证实，预示着苹果秋季发布会进入倒计时。 消息称，iPhone 17 系列的发布会将于 9 月 9 日星期二举行。苹果公司预计将在 9 月 12 日（星期五）开 放预订，并于一周后的 9 月 19 日开始发货。这一时间表与过去八年的发布模式相符。 此次发布会预计将推出 iPhone 17、iPhone 17 Air、iPhone 17 Pro 和 iPhone 17 Pro Max 四款机型。其 中，iPhone 17 Air 的厚度预计为 5.5 毫米，测试机厚度为 5.44 毫米，使其有望成为全球最薄的手机。相 比之下，苹果 M4 iPad Pro 的厚度为 5.1 毫米，而 Tecno Spark Slim 概念手机的厚度为 5.75 毫米，三星 Galaxy S25 Edge 的厚度为 5.84 毫米。 iPhone 17 系列预计将在设计和功能上进行多项升 ...

此外，Genie 3还引入了"可提示世界事件"功能，用户可以通过简单的文本指令动态修改虚拟世界，例 如添加一群鹿或改变天气条件。 外媒称，Genie 3的发布被DeepMind视为迈向通用人工智能（AGI）的重要一步。该模型不仅为AI智能 体训练提供了更广阔的模拟空间，还为游戏开发、教育和创意设计等领域带来了新的可能性。例如，机 器人可以在模拟仓库中学习应对不可预测的场景，而无需真实世界的试错成本。 尽管Genie 3在技术上取得了显著突破，但仍存在一些局限性。例如，模型当前仅支持数分钟的连续交 互，远未达到数小时的理想状态。此外，AI智能体在模拟环境中的交互能力有限，复杂多智能体交互 仍需进一步探索。谷歌DeepMind表示，Genie 3目前以研究预览形式向部分学者和创作者开放，旨在进 一步优化模型并评估潜在风险。（青云） 【环球网科技综合报道】8月6日消息，据PANews报道，谷歌DeepMind今日宣布推出其最新一代世界模 型Genie 3。Genie 3是一款通用型世界模型，能够根据文本提示实时生成多样化的交互式虚拟环境，支 持以24帧/秒的速度生成720p分辨率的交互式3D环境。 来源：环球网 ...