Today, Alibaba DAMO Academy announced the "Ali Star" top talent recruitment and training program for the 2027 university graduates, and opened up 15 cutting-edge research topics covering directions such as AI chips, new CPU architectures, medical multimodal agents, and AGI decision-making, aiming to attract outstanding talents from universities to participate in the exploration of next-generation artificial intelligence technologies.

In the field of chips, DAMO Academy has laid out multiple research directions around AI computing infrastructure, including the design of AI chip programming models, development of AI compilers, research on AI SoC architecture for large model training and inference, and optimization of AI chip inference systems. These studies will involve key technologies such as self-developed AI chip programming languages, compiler toolchains, hardware-software collaborative optimization, advanced packaging, and system architecture design.

In addition, DAMO Academy is also focusing on the evolution of computing architectures in the Agent era, launching the AI Agent CPU architecture exploration topic, researching the next-generation RISC-V high-performance CPU design direction, analyzing the characteristics of agent workloads, and exploring new micro-architectures suitable for Agentic AI. At the same time, the team will also carry out AI-driven RISC-V compilation optimization research, using AI models to enhance the intelligence level of compilation systems, and participate in related international standard exploration.

In the AI application direction, DAMO Academy focuses on medical AI, multimodal agents, and general AI technologies. The relevant topics include a basic large model for medical imaging, a medical multimodal agent, a medical vision large model, and AI Agent development, with the goal of solving practical clinical problems such as disease diagnosis and medical data analysis through large models, reinforcement learning, and agent technologies.

At the same time, DAMO Academy is also conducting AGI decision-making technology research for the industrial field, exploring the application of long-term memory, self-cognition modeling, continuous learning, and autonomous agent capabilities in scenarios such as energy, aviation, high-speed rail, and ports, and promoting the development of AI video cutting-edge algorithm standards and productization implementation.

This "Ali Star" program covers the complete technical chain from the underlying chip architecture to industry intelligent applications, reflecting that the competition in the AI industry is expanding from model capabilities to full-stack technological system construction, including chips, software ecosystems, and intelligent agent applications.