Midjourney has introduced version 5.2 of its AI imaging model, which includes a new unique Zoom out feature. It allows you to save the central created image, and automatically build a larger “scene” around it, simulating reduction by the camera lens. This is reported by Ars Technica.

The feature can take an existing AI-generated image and expand its boundaries while keeping the original object in the center of the new image. However, users cannot choose their own image for the extension.

On the Midjourney Discord server, users can generate and scale any v5.2 image. After that, special Zoom out buttons will appear below the result, with different options to zoom out – 1.5x, 2x, or a custom value between 1 and 2. Another button, called Make Square, generates material around the existing image so that create a square with an aspect ratio of 1:1.

Despite the excitement of the new possibilities, image generation remains highly controversial among some artists. All because artificial intelligence systems learn by using millions of images from the Internet without proper permissions.

We remind you that in the spring Midjourney stopped free trial access due to “extraordinary demand and abuse”. The abuses allegedly involved people creating fake accounts to get free images, which put a strain on the service for paid users.