Apple continues to invest in research on multimodal large language models (MLLMs), focusing on their use for image generation, understanding, and searching. The company’s latest studies explore how these models can be used to generate images, understand their content, and perform web searches with clipped images.
See also: AI: Emerging models enhance and threaten cybersecurity
With the release ofiOS 18, Apple made it possible to create images on an iPhone using local AI models. TheImage Playground allows you to create cartoon-style photos of almost anything, without the need for a Wi-Fi connection.
Apple continues its efforts in the imaging field through research examining how MLLMs can be used to understand, create, and search for images.
See also: Data theft with AI: is Automated data poisoning the solution?
MLLMs technology is one of the most exciting areas of artificial intelligence, as it combines language understanding with image processing. This combined approach allows models to better understand the content of images and provide more accurate and useful information to users. Apple, with its research, is trying to improve user interaction with its devices, offering more sophisticated image recognition and creation capabilities.
The ability to create images in iOS 18 using local AI models is a significant step in this direction. The Image Playground app allows users to create images without needing an internet connection, which offers greater privacy and security. Additionally, Apple is exploring how multimodal models can be used to perform web searches with images, offering new possibilities for content discovery and exploration.
See also: Generative AI does not directly threaten lawyers' jobs
Apple’s ongoing research into artificial intelligence and multimodal models demonstrates the company’s commitment to delivering innovative solutions to its users. As technology evolves, we expect to see even more applications of these models on Apple devices, improving the user experience and offering new possibilities for understanding and creating content.
