Google add image generation and long drafts to Search AI

Google is adding new features in SGE, Search Generative Experience, its version of AI in Google Search. Now you can generate images and write your first draft right inside Google Search.

What's going on here?

Google is adding image and drafting capabilities to its Search AI.

What does this mean?

You can create images when you search for terms similar to image generation prompts, like “photorealisitic image of a cute bear”. You can refine the generated images by editing the text prompt in a chat mode after the first generation. The image tab in search results will also occasionally have a “create something new” card when you’re looking for image inspiration.

Google will also launch an “About this image” tool soon that will help in identifying AI-created images (based on assigned labels).

Additionally, for longer research sessions, the AI can provide written drafts to capture ideas and save time. For instance, it can generate a note to a contractor based on a garage renovation project research trail. You can export these drafts to Google Docs and Gmail.

Why should I care?

These features are only available for US users in English. Though terrible name (SGE), I’ve found it is actually helpful for Googling stuff: when I need to find stuff that I think will be deep inside some web result. Not so much, when I want to imagine something new.

Based on that previous experience, the first drafts could be a good addition. For the images, I think comparing web images and your generation side by side could be an interesting aspect.