Apple is in talks with the startup shrinking a 27B AI model onto an iPhone

Apple is in preliminary discussions with PrismML, a startup that compresses large AI models to run efficiently on devices like iPhones. PrismML recently released a compressed version of Alibaba's Qwen model, significantly reducing its size to fit within an iPhone 17 Pro's memory while maintaining most of its performance. This technology could enable Apple to run more AI processing locally on devices, enhancing Siri's capabilities, reducing latency, lowering cloud costs, and improving user data's