Creating campaigns with images generated by artificial intelligence has something of a democratizing force. It is said that in the drawers of agencies and studios around the world lie a lot of ideas that were never realised due to lack of budget or time. These ideas, which used to involve a great deployment of resources, can now slide from the drawer to the table and reclaim their place with a good sequence of prompts. Something that we experience here with a mixture of amazement and excitement.

These are some of the seasonal fashion campaigns we have created for illa Carlemany. Vivaldi would be happy.

Summergeix-te (submerge yourself), with two emes, from summer, is the summer campaign. First came the idea, the word, then the image in the head: fashion elements immersed in a swimming pool.

A campaign that not only was liked by us and illa Carlemany, but also by the Lüzer Archive.

Fast forward to the autumn campaign. After a tough negotiation of prompts with the AI, we got what we had in mind: sweaters that evoke the image of autumn landscapes. A touch-up phase and the result is there for all to see.

Jersei or landscape, AI or AI, fear or enthusiasm, in the end, it’s all a matter of perspective.

There is no other way to get these images than by ripping them out of Midjourney through a trial-and-error bombardment of prompts. As with any creative process, we go through a phase of chaos before we start to get the results we are looking for. Dealing with artificial intelligence and being effective with text to image means interpreting the process not as a succession of commands but as a dialogue. The prompts are the sentences you throw out, the images are the AI’s response.