I believe it’s for a few reasons. First, it is jaw dropping incredible for most people in tech who have at least a hint of how most ML works. Second, the AI image generation field is racing ahead, in academics and new trained models, so there’s lots of new news. Thirdly some really great models like Dall-e have been opened for wider access and lots of everyday users are discovering its capabilities and doing blog write-up’s which are not news, but are surely interesting to most.