- Apple is developing its own large language model (LLM) to bring AI features to iPhones.
- Apple's AI features will run on-device and not cloud services like the existing AI services.
- On-device models offer faster responses than sending requests through a cloud service. Also, they are less capable when it comes to complex tasks.
Apple has been reportedly working on several artificial intelligence features for the upcoming iOS 18 update. Besides investing heavily in all sorts of AI technologies, the Cupertino tech giant was reported to be in talks with some major players like Google, Baidu, and OpenAI to build their AI engine into upcoming iPhones. Now, the latest newsletter by Mark Gurman suggests that Apple is developing its own large language model (LLM) to bring AI features to iPhones. This means the AI features on an iPhone would be powered by Apple’s own LLM that runs on-device to prioritize privacy and speed.
Mark Gurman said “All indications” apparently suggest that it will be entirely on-device. Therefore, the AI features will run on-device and not cloud services like the existing AI services. That means, the technology is tuned to the Apple processors inside its devices. When it comes to on-device models, they offer faster responses than sending requests through a cloud service. Additionally, they offer the advantage of functioning offline in areas with limited or no connectivity.
Despite some advantages, we cannot ignore the fact that Apple’s on-device AI tools may be less capable or less powerful in doing complex tasks than its direct cloud-based rivals. However, Gurman suggested that Apple could “fill in the gaps” by leveraging technology from Google and other existing AI service providers.
Last month, there were reports that Apple is in discussions with Google to build a Gemini Artificial Intelligence engine into an iPhone.
The on-device AI features will provide users with quicker response times and complete privacy, as compared to cloud-based solutions. This AI technology strategy seems to be based on how it can help users in their daily lives, rather than complex capabilities and power. For instance, an on-device LLM could help users generate auto-replies to Messages or deliver better responses for many common Siri requests.
We’ll get all the details about Apple AI goals at its annual Worldwide Developers Conferences which is scheduled for June 10 to June 14. During the event, the giant is expected to give a first look at iOS 18 and AI features which will give us a clear picture of what to expect from the iPhone 16 series.
