📄 中文摘要
近期 AI 领域的发展主要聚焦于实际应用层面，而非单纯的性能基准提升。OpenAI 已开始在美国针对免费和 Go 订阅的成年登录用户测试 ChatGPT 广告。这些广告明确与模型回答分离，并被标记为赞助内容，OpenAI 声称广告不会影响模型回答，但其选择会基于对话主题、历史聊天记录和过往广告互动。广告商无法直接访问用户数据，但可根据广告互动情况进行归因。Discord 正在测试新的年龄验证工具，以遵守数字服务法案（DSA）等法规。这些工具包括通过政府 ID 验证、由第三方专家进行身份验证，以及通过家庭共享功能进行家长验证。GitHub 则通过其 Copilot Workspace 引入了智能体工作流，旨在将自然语言指令转化为可执行的开发任务，实现从需求到代码的自动化。这些智能体能够理解需求、规划解决方案、生成代码、执行测试并进行迭代，显著提升开发效率。这些进展预示着 AI 在商业变现、用户身
📄 English Summary
AI News Roundup: Ads in ChatGPT, Discord age checks, and GitHub agentic workflows
Recent advancements in the AI landscape are increasingly focused on practical applications that directly impact developer realities, rather than just benchmark improvements. OpenAI has initiated an advertising test for logged-in adult users in the US on its Free and Go tiers of ChatGPT. These advertisements are explicitly separated from model answers, labeled as sponsored content, and OpenAI asserts they do not influence the model's responses. However, ad selection can leverage conversation topics, past chat history, and prior ad interactions. Advertisers do not gain direct access to user data but can receive attribution based on ad engagements. Meanwhile, Discord is implementing new age verification tools to comply with regulations like the Digital Services Act (DSA). These tools include government ID verification, third-party expert identity verification, and parental verification via family sharing features. GitHub is also revolutionizing development workflows with its Copilot Workspace, introducing agentic capabilities that transform natural language instructions into executable development tasks. These agents are designed to understand requirements, plan solutions, generate code, execute tests, and iterate, significantly automating the journey from idea to code. These developments highlight a growing trend of AI integration into monetization strategies, identity management, and software development automation, signaling profound shifts in how these technologies are applied and experienced.