It’s no secret that 2024 will be a very important year for Apple when it comes to artificial intelligence. A week ago, Tim Cook announced that we would have news in this area “at the end of this year”, and there were rumors that iOS 18 would be one of the most important updates in the history of Apple’s operating system.

Thanks to a new paper published by Chinese Apple researchers, we already have an idea of ​​where some of the AI ​​functions that will soon enter its ecosystem might be heading. Apple tells us about MGIE (MLLM Guided Image Editing), tools that allow manipulation of the image using multi-modal language models.

Talking about AI is talking about photo editing

We saw this in the magic editor of the Google Pixel 8, and it was repeated on the Samsung Galaxy S24. Talking about artificial intelligence, we are talking about phones that can replace components of our images, eliminate them, change their size and, in the case of the Pixel, even replace them with others (like the sky).

The document Apple submitted shows a slightly different tool: it refers to a type of tool. language model guided editing. This is the MGIE model, a collaboration between Apple and the University of California. Can crop, resize and replace elements of a photo using text prompts.

The real benefit of the editor lies in these instructions; Just type “add more contrast” to improve the image or “make this photo healthier” to add some green to the pizza photo. Just like Midjourney or DALL-E 3 did, the AI ​​has the ability to make changes to the image, in this case found in our gallery.

At this time, Apple hasn’t announced what the plans are beyond the investigation, and there’s no guarantee that iOS 18 will implement this tool. However, it is an important clue about Apple’s research in the field of artificial intelligence.

Image | Xataka

in Xataka | From Bing to Copilot and Bard to Gemini: These are Microsoft and Google’s strategies for winning the AI ​​war