The MGIE tool, however, still has limitations similar to other AI image generators, with results often being bizarre and not as requested. Despite this, it has shown impressive results in some cases, particularly with familiar subjects. Apple has also published an open-source version of the software on GitHub, and while it may never get an official release, it represents a significant investment in AI technology by the tech giant.
Key takeaways:
- Apple has collaborated with researchers at the University of California at Santa Barbara to develop a new AI image editor called Multimodal Large Language Model Guided Image Editing (MGIE).
- The tool allows users to edit photos and images using text-based instructions, with the AI making these instructions more explicit and descriptive.
- While the tool is not officially released, a demo version is available for users to try, and an open-source version of the software has been published on GitHub.
- Despite some limitations, the tool has shown impressive results in some cases, particularly with familiar subjects.