[Apple in Talks with AI Startup to Enable Large Models to Run Directly on iPhone] According to a report by Golden Finance on July 15, citing CNBC, Apple is engaging in preliminary discussions with AI startup PrismML and evaluating its AI model compression technology. The goal is to allow large AI models to run directly on iPhones without relying on cloud computing. PrismML claims its technology can reduce model memory usage by 10 to 15 times, enabling models with 27 billion parameters to operate on iPhone 15 and newer devices, while improving response speed by 6 to 8 times and reducing energy consumption by 3 to 6 times. If the technology matures, Apple could potentially shift more Siri and generative AI functionalities to local processing, further enhancing response speed, privacy protection, and offline usability, while lowering cloud computing costs. However, industry experts have pointed out that the technology still requires large-scale practical application validation, and its impact on battery life, stability, and long-term performance remains to be observed.