About
AI in graphic design / art can be tricky. Using Leonardo.ai to generate images is not a big challenge, but writing the right prompts to generate a highly specific concept, scene and elements can be tricky. AI is also very bad at generating authentic images of humans or animals, who often end up having extra limbs, anime-like features or smudgy faces.

The base for these images have been created in Leonardo.AI using specific prompts and reference images (also generated by AI). I’ve used Photoshop for generative expand on all images, to make them the desired dimensions and I’ve edited areas of the scenes that I wanted different eg. resculpting rocks, removing blemishes or adding a waterfall. I’ve generated elements such as the statues in both Leonardo.AI and Adobe Firefly, which have then been manually isolated from their backgrounds. These elements have then been resculpted and editet to blend in with their new background, and the surroundings have been edited to embrase the new objects.