Apple Enhances AI Capabilities on iOS 18 with Integrated Language Model and New Chatbot: Exclusive Insights

Apple is enhancing its AI capabilities on iOS 18 by developing its own large language model (LLM) to be integrated into iPhones. This model will allow for AI processing to be done on the device itself, prioritizing privacy by eliminating the need for cloud access. The company is focusing on introducing new AI-based features ahead of the launch of the next major iOS operating system update and has been negotiating with Google to implement the generative AI engine Gemini on iPhones through software updates later this year.

Apple’s goal is to create AI functions that can be used on its devices without relying on the cloud. This includes features like image creation and essay writing based on simple instructions from the device. By developing an integrated LLM model that processes functions directly on the device, AI tasks can be carried out privately and securely, without leaving the device or depending on external connections.

In addition to the LLM model, Apple is working on a chatbot called Ajax to challenge other companies in the AI technology space. The company’s AI unit dedicated to foundational models is focusing on conversational tools, including visual intelligence and the Ajax chatbot. Despite having fewer parameters than other models like GPT-4, Apple’s RealLM model offers similar performance in understanding ambiguous references in conversations, enhancing the naturalness of interactions between AI and human users.

Although LLM integrated into devices may have some limitations compared to cloud-based models, Apple is investing in new tools and technologies to implement AI functions directly on iPhones. The company will announce further developments in AI and language models at the upcoming Worldwide Developers Conference (WWDC) 2024, scheduled for June 10 to 14.