Khaberni - Microsoft announced its first AI image generation model developed entirely by its internal teams, a move that enhances its inevitable split from "Open AI" as reported by the technology site "The Verge".
The new model, named "MAI-image-1," clearly indicates that it is the first generation of AI dedicated to image generation by the company.
Microsoft says it has collaborated with several imaging and image generation experts to evaluate the model and distance it from repetition or similarity with other image generation models. These experts believe the model excels at generating realistic images and simulating lighting conditions and natural scenes.
The model is capable of handling and editing images much faster than larger models, which are slower, according to a Microsoft statement.
Despite being launched just hours ago, the model has climbed into the top 10 list on "LmArena," a site dedicated to AI model evaluation and comparison.
The launch of this model underscores Microsoft's steps towards complete separation from "Open AI" after the relationship between them became complicated following the latter's steady growth and introduction of several models, despite Microsoft being one of the first investors in it.
Microsoft recently integrated "Anthropic" artificial intelligence services directly into its "Office 365" products and is currently focusing on enhancing its investments in AI model training instead of relying on external companies to develop AI models.
Microsoft's model is the latest addition to the sector of AI generative models specialized in images, following Google's introduction of the "Nano Banana" model, which amazed the world with its image generation and editing capabilities.
For now, the model remains confined to community trials on the "LmArena" site without an official release, and it is expected that the company will improve it based on user feedback on the site before making it available to the public.
