第一财经·2025-11-26 04:16Core Viewpoint - Baidu has established two new departments for foundational and application model research, indicating a significant upgrade in its large model research system and a strong commitment to AI development [1][4]. Group 1: Organizational Changes - The new foundational model research department will be led by Wu Tian, focusing on high-intelligence, scalable general AI models, while the application model research department will be headed by Jia Lei, concentrating on fine-tuning specialized models for business scenarios [1]. - Both Wu Tian and Jia Lei are internal talents with extensive experience in model research, algorithm optimization, and team management, showcasing Baidu's commitment to nurturing talent from within [1][2]. Group 2: Model Development - The foundational model serves as the core carrier for technological breakthroughs, with Baidu's Wenxin large model 5.0 featuring 2.4 trillion parameters and utilizing native multimodal unified modeling technology [2]. - Specialized models act as precise tools for implementing foundational capabilities in various industries, leveraging knowledge distillation technology to achieve lightweight and efficient inference [2]. Group 3: Systemic Support - The synergy between the two types of models is supported by Baidu's four-layer AI architecture, which includes Kunlun chips for efficient computing power, the PaddlePaddle deep learning framework, and a feedback loop from industry applications to optimize model performance [3]. - This architecture allows for complex task execution and enhances the training precision of foundational models through knowledge sharing from specialized models [3]. Group 4: Investment and Competitive Advantage - Baidu has invested over 100 billion yuan in AI since the launch of Wenxin Yiyan in 2023, with plans for continued investment in the field [4]. - The combination of a systematic four-layer architecture and the dual-department collaboration enhances resource focus, efficiency, and technological implementation, strengthening Baidu's competitive edge in the global AI landscape [4].