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

Apple is in early discussions with PrismML, a startup that compresses large AI models to run on devices like the iPhone, potentially enabling more on-device AI processing for Siri. PrismML's Bonsai 27B model shrinks Alibaba's Qwen model from 54GB to as little as 3.9GB, allowing it to fit on an iPhone 17 Pro with some performance trade-offs. This technology could reduce latency, cloud costs, and enhance privacy by processing AI tasks locally on Apple devices.