Apple developing on-device LLM for better AI features: Analyst

Apple on-device AI: Apple is currently focusing on working on large language models (LLMs) to create an on-device solution prioritising speed and privacy, according to a recent report by Mark Gurman. 

Gurman's insights, shared via his "Power On" newsletter, indicate that Apple's upcoming generative AI features will be powered by this proprietary LLM. The model will operate entirely on-device, diverging from the prevalent cloud-based approach adopted by most AI services.

While this on-device approach may imply some limitations compared to cloud-based counterparts, Gurman speculates that Apple could augment its capabilities by potentially licensing technology from industry players like Google. According to media reports, discussions have surfaced regarding the integration of Google's Gemini AI engine into the iPhone as part of iOS 18.

The primary benefits of on-device processing are expected to include swifter response times and better privacy measures, positioning Apple's AI tools as a compelling alternative in the market.

Apple's upcoming AI endeavours are expected to pivot towards showcasing practical utility in users' daily routines rather than focusing on raw computational power. A comprehensive release of Apple's AI strategy is expected alongside previews of significant software updates at the upcoming WWDC event in June.