SLAI Seminar 39th Session will be discussing the topic on “From Intelligent Assistants to Coding Agents”,  from 2:30pm-4pm, June 4 (Thursday) at Room B401, online participation is welcome.

 (Tencent Meeting ID: 880-246-101)

报告主题： 从智能助手到编程智能体

时间：2026年6月4日（周四）下午2:30-4:00

地点: 深圳河套学院B401教室

线上参与：腾讯会议号   880-246-101

主讲嘉宾：曹亮亮教授   崔英云博士

主持人：欧阳万里教授

讲者简介About the Speaker:

曹亮亮教授现任香港理工大学讲席教授，此前曾任谷歌深度思维（Google DeepMind）首席工程师、部门负责人。他先后在苹果、谷歌、雅虎、IBM 担任研究与工程管理岗位，并联合创立了斯维提科技（Switi Inc.），该公司于 2018 年被收购。他是电气和电子工程师协会会士（IEEE Fellow），同时斩获了国际计算机学会多媒体专委会青年新星奖（ACM SIGMM Rising Star Award）。

Liangliang Cao is an incoming Chair Professor at Hong Kong Polytechnic University and former Principal Engineer/Director at Google DeepMind. He previously held research and engineering leadership roles at Apple, Google, Yahoo!, and IBM, and co-founded Switi Inc., which was acquired in 2018. He is an IEEE Fellow and recipient of the ACM SIGMM Rising Star Award.

讲者简介About the Speaker:

崔英云是资深软件架构师与工程负责人，在应用程序、操作系统及大型软件平台领域拥有丰富从业经验。他曾先后任职于 OPPO、美图、魅族，担任高级技术管理职务，主导打造了多款服务亿万用户的产品与系统。

Yingyun Cui is a veteran software architect and engineering leader with extensive experience spanning applications, operating systems, and large-scale software platforms. He has held senior technical leadership roles at OPPO, Meitu, and Meizu, building products and systems used by millions of users.

报告摘要：

人工智能正经历一场根本性变革：从问答助手，迈向具备感知、推理与行动能力的智能体。过去十五年来，行业从 IBM 沃森系统、语音助手，逐步发展到多模态人工智能系统，以及深度融入实际工作流程的代码智能体。

本次分享将结合大规模 AI 产品研发与一线软件工程的实践视角，剖析代码领域为何成为智能体式人工智能首个实现重大突破的应用方向，分享代码智能体落地生产环境的实战经验，并探讨这类系统将如何重塑软件开发、知识型工作，以及人与机器未来的协作关系。

Abstract:

Artificial intelligence is undergoing a fundamental transition—from assistants that answer questions to agents that can perceive, reason, and act. Over the past fifteen years, the industry has evolved from IBM Watson and voice assistants to multimodal AI systems and coding agents that increasingly participate in real-world workflows.

In this talk, we combine perspectives from large-scale AI product development and practical software engineering. We discuss why coding has become the first major breakthrough application of Agentic AI, share lessons from deploying coding agents in production environments, and explore how these systems may reshape software development, knowledge work, and the future relationship between humans and machines.