Core Insights - The fifth Intelligent Computing Innovation Forum was held in Hangzhou, focusing on the theme "Computing Relies on Intelligence, Computing for Intelligence," attracting international experts to discuss advancements in AI technologies and their applications across various scientific fields [1] Group 1: AI Model Development - Scientists are exploring the potential of AI in solving scientific problems, emphasizing that current large language models have not yet reached human-level reasoning capabilities [2] - The development of scientific foundational models requires collaboration with scientists to effectively tokenize and train diverse scientific data, addressing complex interdisciplinary issues [2] - The learning paradigm of foundational models is evolving through imitation learning, reinforcement learning, and autonomous learning, with a shift towards task processing applications [2] Group 2: Efficiency and Resource Consumption - The efficiency of foundational models is critical for large-scale AI application deployment, with a noted exponential increase in token consumption correlating with model capability improvements [3] - The cost of generating tokens decreases with higher reasoning efficiency, necessitating collaborative optimization across the industry to enhance model performance [3] Group 3: Practical Applications and Collaboration - The application of intelligent systems in dynamic environments is gaining attention, highlighting the importance of responsive robotics [4] - China is recognized for its leading capabilities in intelligent manufacturing, serving as an excellent testing ground for new technology applications [4] - There is a call for scientists worldwide to establish collaborative networks to enhance research outcomes and create new possibilities through cooperation [4]

Core Insights - The efficiency of model inference is identified as a key factor for the large-scale application of AI, requiring optimization across the industry supply chain and system model co-design [1] - The founder of Jiyue Xingchen, Jiang Daxin, highlighted that the application of large models has transitioned from a chat-based era to a task-oriented era since the beginning of this year [1] - Future intelligent terminals will be characterized by their ability to perform tasks, be omnipresent, have memory, and evolve, with a focus on intelligence rather than hardware [1] Industry Developments - Jiyue Xingchen collaborates with multiple domestic computing chip manufacturers, achieving a maximum context inference efficiency of its Step3 domestic card that is three times higher than that of DeepSeek R1 [1] - The effective working time of agents in the digital world is expected to increase, with a focus on enhancing model inference capabilities and generalization [1] - Agents are anticipated to transition from the digital realm to the physical world, evolving from experience-based learning to self-evolution, as human data is limited and biased [1]