“The next major model release from Microsoft's MAI division will be an omni-modal model combining their work in audio, image, and text.”