Generating ideas with remix, prompts, and AI models
Let's generate some new ideas from this imagery.
Select two or more images and hit Remix.
And you'll see that it has generated a few good options for us.
[A prompt bar opens at the bottom displaying tags like White Flower and Black Background.]
Now if we click Load on a remixed image, it opens up a Prompt bar, and by selecting Sample from canvas and clicking on another image, [The cursor highlights options from a dropdown menu, selecting Use as style reference before clicking Generate.] you can use the whole image as a style or composition reference, or use one of these prompts.
For example, I'm going to select style reference here, hit Generate, and you'll see we have some new options to pick from.
[The cursor opens a dropdown menu under Model to select a different generation tool.]
You may have also noticed that there are also different models to choose from.
For example, the ChatGPT model allows for instructional editing, meaning you can suggest changes to your imagery.
[The text "Change to a mint green color scheme and make it more glowy" is typed into the prompt bar.]
So, if I select this image as my reference, and let's say I want to 'Change to a mint green color scheme and make it more glowy', it will create that version for me.
And the Veo model here actually allows you to create imagery with motion.
Let's test that out.
[The model changes to Veo 2, generating a moving video of a glowing green flower dress concept.]
Super cool.
[The canvas zooms out to show an overview of all the generated images and sketches.]
Let's take a high-level look at my board.