加入 DeepMind 的 Windsurf CEO Varun Mohan 与联合创始人 Douglas Chen 也发布了一份声明，其中表示：「我们很高兴能与 Windsurf 团队的部分成员一起加入谷 歌 DeepMind。我们为 Windsurf 过去四年取得的成就感到自豪，并期待看到它与他们世界一流的团队一起前进，开启新的发展阶段。」 至于达成这笔交易的具体金额，目前还没有人透露。作为参考，此前有报道称，OpenAI 将斥资 30 亿美元收购 Windsurf。 那么，OpenAI 与 Windsurf 的交易究竟出了什么问题？ 机器之心报道 编辑：Panda 就在所有人还在惊叹于月之暗面开源 Kimi K2 模型的同时，谷歌 DeepMind 却宣布截胡了 OpenAI 原本打算收购的 Windsurf。 诺贝尔奖得主、DeepMind 联创兼 CEO Demis Hassabis 与 DeepMind CTO Koray Kavukcuoglu 在 上向 Windsurf CEO Varun Mohan、联合创始人 Douglas Chen 以及 部分加入 DeepMind 团队的研发人员表示了 ...

Core Insights - The article emphasizes the importance of treating AI as a junior developer that requires clear guidance rather than a magical tool, highlighting the need for engineers to adapt their management style to effectively utilize programming agents [1][3][9] - Senior engineers are found to be the quickest adopters of these tools, which can save approximately 80% of time on medium to large tasks [1][8][24] Introduction - The article introduces a practical guide based on two years of experience building Devin, an autonomous programming agent, and aims to share valuable insights from customer feedback and internal practices [1][3] Getting Started: Basics and Daily Applications - Key principles for effective communication with agents include providing specific instructions, indicating starting points, anticipating potential errors, and establishing a feedback loop [10][11][13][15] - The guide suggests integrating agents into daily workflows to enhance personal efficiency, such as handling new requests without interrupting deep work and managing urgent issues on the go [17][19][20] Intermediate: Managing Complex Tasks - For complex tasks, the article recommends having agents draft initial versions and collaborating on implementation plans, while also setting checkpoints to ensure alignment with expectations [23][25][26] - It emphasizes the importance of teaching agents how to validate their work and increasing testing coverage in areas frequently modified by AI [28][29] Advanced: Automation and Customization - The article discusses creating automation templates for repetitive tasks and implementing intelligent code reviews using agents [30][33] - It highlights the need for a unified development environment to enhance agent performance and suggests building custom tools to empower agents [35][36] Practical Considerations: Embracing Change - The article outlines the limitations of autonomous agents, such as their debugging capabilities and knowledge cut-off dates, advising users to manage expectations and time effectively [39][42][43] - It concludes by asserting that the value of software engineers will not diminish, as deep technical knowledge and understanding of business codebases remain essential in the evolving landscape of software development [50]

Core Insights - The article emphasizes treating AI as a junior developer that requires clear guidance rather than a magical tool, highlighting the importance of effective communication with programming agents [1][8][9]. Group 1: Key Methods for Effective Use - Clear Instructions: Specificity in commands is crucial, such as detailing which functionalities to test rather than vague requests [3][16][18]. - Reasonable Expectations: Large tasks cannot be fully automated, but can save approximately 80% of time; checkpoints should be established for planning, implementation, testing, and review [3][27]. - Continuous Validation: Providing a complete CI/testing environment allows agents to discover and correct errors independently [3][19][33]. Group 2: Daily Usage Tips - Instant Delegation: Quickly assign tasks to agents when urgent requests arise [5][21]. - Mobile Handling: Use mobile devices to address urgent bugs while on the go [5][23]. - Parallel Decision-Making: Allow agents to implement multiple architectural solutions simultaneously for better decision-making [5][25]. Group 3: Advanced Applications - Automate Repetitive Tasks: Create templates for recurring tasks to enhance efficiency [5][35]. - Intelligent Code Review: Utilize agents for precise code reviews based on a maintained list of common errors [5][36]. - Event-Driven Responses: Set up agents to automatically respond to specific events, such as alerts [5][37]. Group 4: Practical Considerations - Understanding Limitations: Agents have limited debugging capabilities and should not be expected to resolve complex issues independently [42][43]. - Time Management: Learn to recognize when to stop ineffective attempts and start anew with clearer instructions [46][49]. - Isolated Environments: Agents should operate in isolated testing environments to prevent unintended consequences in production [51][52]. Group 5: Future Outlook - The value of software engineers remains significant despite advancements in programming agents; deep technical knowledge and understanding of codebases are essential [53].

36氪获悉，劲仔食品在互动平台表示，公司部分创新升级产品已进入胖东来系统，目前销售情况良好。 扎克伯格为Meta新 "超级智能"AI团队招聘人员 大公司： 劲仔食品：公司部分创新升级产品已进入胖东来系统 Meta平台首席执行官马克・扎克伯格正在组建一个专家团队，以实现所谓的 "通用人工智能"（AGI）， 即能力可匹敌或超越人类的机器。彭博援引消息人士称，扎克伯格在组建这支新AI团队的同时，计划 向Scale AI投资超100亿美元，并补充称，交易完成后，Scale AI创始人亚历山德・王预计将加入该AGI 团队。路透社未能立即核实彭博的报道。Meta在正常工作时间外未立即回应置评请求。（新浪财经） Uber、Wayve宣布计划在伦敦进行无人驾驶汽车测试 Uber、英国初创公司Wayve宣布计划在伦敦进行无人驾驶汽车测试。（智通财经） 天康生物：5月生猪销售收入3.45亿元，同比下降19.95% 越南电动车商VinFast一季度交付量同比大增296%，净亏损近7.12亿美元 越南电动车商VinFast当地时间6月10日公布一季度经营业绩，共交付36330辆电动汽车，同比增长 296%，超过去年上半年交付总量。一 ...

在最新的媒体采访中，GitHub CEO托马斯·多姆克正面battle最近甚嚣尘上的"编程无用论"。 在他的构想中，2025年是属于编程智能体的一年，但未来仍然属于全体人类程序员。 鹭羽 发自 凹非寺 量子位 | 公众号 QbitAI 「不要学习编程」是错误的！ 网友们纷纷表示：终于！看到有人说相反的话了。 与他一同接受采访的CPO马里奥·罗德里格斯，更是直言不讳地说道： 我们将用Copilot增强个人能力，并且我们在这方面相当成功。 在过去的一年里，这家以"帮助程序员编写更好的代码"为使命的公司，先后推出了 Workspace 、 Project Padawan 等多个SWE （软件工 程） 智能体产品，一经发布就引爆互联网讨论热潮。 更是在Microsoft最新的季度收益报告中，CEO指出，Copilot现拥有的 1500万 用户，同比增长 4倍 ，皆是源于免费开放的Copilot层。 而在智能体角逐更加激烈的2025年，GitHub仍然抱有野心： 我们正在进入第二波——下一步是软件开发的下一次演变。 来看看具体讨论细节。 2025年：SWE智能体之年 两位高层一致认为，2025年的SWE演变将完全围绕 ...

机器之心报道 机器之心编辑部 从编程开始，今年智能体要卷飞了！！！ 昨天，OpenAI CEO 奥特曼预告了一项新研究，吊足了所有人的胃口。 就在刚刚，谜底揭晓！OpenAI 宣布， 在 ChatGPT 中引入了 Codex 的研究预览版。 Codex 是一个云端软件工程智能体，可以并行处理多项任务 ，包括编写功能、解答代码库相关问题、修复 bug 以及提交拉取请求以供审核等。并且，每个任务都 在其专属的云沙盒环境中运行，并预加载代码库。 Codex 背后的模型是 codex-1，它是 OpenAI o3 的一个版本，专门针对软件工程进行了优化。 它使用强化学习在各种环境中针对真实编程任务进行训练，生成的 代码能够高度反映人类的风格和 PR 偏好，精确遵循指令，并可以迭代运行测试，直到获得合格结果。 自今天开始，ChatGPT Pro、Team 和 Enterprise 用户即可使用 Codex，Plus 和 Edu 用户也很快可以上线使用。 对于 OpenAI 的新智能体产品，大家的反应似乎非常兴奋。有人表示自己被震撼到了，迫不及待想要体验一番。十年前学习编程时，绝对想不到这会成为可能。 也有人认为，Cod ...