Core Insights - The launch of the SAIC Roewe M7 DMH marks the introduction of the "Doubao Deep Thinking" model, which enhances the AI capabilities of the vehicle, enabling it to perform deep thinking and understand user intentions more effectively [1][3]. Group 1: AI Capabilities - The Roewe M7 DMH utilizes the Doubao Deep Thinking model to interpret vague user commands and deduce their true intentions, showcasing advanced semantic understanding [3][5]. - The vehicle features the highest industry coverage for vague semantic understanding, encompassing 15 key driving scenarios, transitioning from command-driven to intention-driven interactions [5]. - The model's memory capabilities allow it to retain user preferences and historical interactions, enhancing the personalization of the user experience while ensuring privacy [5]. Group 2: Complex Control Functions - The Roewe M7 DMH can execute multiple vehicle control tasks through a single user command, significantly improving the interaction experience [8]. - The vehicle boasts the highest number of control ports in the industry, facilitating a more integrated and intelligent user experience [8]. - The collaboration between SAIC Roewe and Volcano Engine aims to merge advanced AI capabilities with the vehicle's underlying systems, addressing the evolving user demand for smarter, more intuitive vehicles [8].

字节的 Agent 产品来了。 4 月 18 日晚间，字节跳动扣子空间开启内测，定位通用 Agent。与其他类似产品如 manus 一样，扣子空间采用了邀请码制。 平台上，用户可以选择精通各项技能的「通用实习生」，也可以选择行业的「领域专家」，通过与 AI 的互动完成工作任务。 就在前一天，火山引擎刚刚面向企业市场推出 OS Agent 解决方案及 AI 云原生推理套件，要帮企业更快、更省地构建和部署 Agent 应用。 扣子有探索和规划两种模式，如果想让它一步到位输出，可以选择探索模式；如果想亲自把控每个步骤，可以选择规划模式。 虽然采用了邀请码制，但扣子显然不是走饥饿营销的路线。用户激活获得的邀请码后，创建并完成一个新任务即可获得 5 个邀请码，邀请码激 活后，还可获得更多邀请资格，多创建多获得多邀请。 上线的第一时间，极客公园就上手实测了扣子空间的 Agent 功能。可以看到，在执行的三个任务中，制定旅游攻略和一周穿搭的任务顺利完 成，但另一个专家助手的任务，却出现了 Python 脚本调用失败、 API 权限异常等 Bug，一个晚上都没能执行成功。 字节迈出了走向 Agent 时代的第一步，但距离完美 ...