A new study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) demonstrates a novel method for training AI models to understand and generate images from textual descriptions. The approach, called ‘MAGE’, treats image creation as a unified process of masked token generation, blending elements of generative and discriminative modeling. This allows the system to perform multiple tasks, such as image generation, editing, and classification, using a single framework. Researchers report that MAGE achieves state-of-the-art performance on several benchmarks, generating high-fidelity images more efficiently than previous methods. The work suggests a promising direction for developing more versatile and capable multimodal AI systems. Read the full article at https://technologyreview.com/2023/06/05/1073377/mage-ai-image-generation.