➀ RISC-V's flexibility and scalability make it an ideal choice for AI chip design, allowing customization of AI accelerators. ➁ Two main models of RISC-V AI chips are identified: Integrated mode for low power and Attached mode for high computational power. ➂ Challenges in the RISC-V+AI ecosystem include fragmentation and insufficient resources, addressed through international standards and open-source software. ➃ Focus on edge computing and smart terminals to build a competitive software ecosystem against NVIDIA's CUDA. ➄ International collaboration and open-source community development are crucial for RISC-V's global market positioning.