The Information reported that Apple is working to integrate AI-powered LLMs into its voice assistant.
Two teams are claimed to be working on image and language models, specifically visual intelligence, 3D scenes, image and video generation, and multimodal AI that can handle videos, images, and text. These improvements are said to be added to Siri, which then unlocks natural language tasks such as the one Google had. Apple has Ajax GPT, which the company believes is better compared to OpenAI’s model.
Apple supposedly set the wheels of generative AI in motion four years ago, but it wasn’t until May that the Cupertino-based company started hiring engineers for the project. Currently, the AI team at Apple is led by Ruoming Pang and consists of 16 members. The project does present several challenges, one of which is adding the software on-device for better performance and privacy. LLMs are large and will be difficult to put into the iPhone, for instance.