Today, most generative image models basically fall into two main categories: diffusion models, like Stable Diffusion, or autoregressive models, like OpenAI’s GPT-4o. But Apple just released two ...