Artificial Intelligence (KI) in art creation is a fascinating and multifaceted topic. From mere support to becoming a dominant force in the creative process, the dynamics between human and machine are evolving. In this article, I explore the current developments and possibilities that arise from the use of AI in art.
In today’s era, AI models such as Stable Diffusion, Midjourney, and Dall-E are at the forefront of the creative process. These models utilize vast datasets to generate context-based associations that serve as the basis for creating artworks. However, the range of possibilities depends on the trained data models and the parameters used.
More control over Artificial Intelligence
We are witnessing an upsurge of tools that grant artists more control over AI. An outstanding example is ComfyUI, a user interface that allows controlling the outputs of AI not only through text prompts but also through image prompts and videos. Through the so-called Image Prompt Adapters (IP-Adapters: https://github.com/tencent-ailab/IP-Adapter), artists can deliberately influence and adjust the visual style of their works. Furthermore, the ControlNETs plugin (https://github.com/lllyasviel/ControlNet) enables precise control of composition. It gives the artist the ability to define various elements through arrangement, overlay, posture, and facial expression.
How much space should be given to the machine?
The challenge lies in finding the right mix of text prompts, image prompts, and ControlNETs so that the desired creative results can be achieved. The question of the AI’s freedom plays a crucial role in this. How much space should be given to the machine to act autonomously? And how much should it be controlled by the artist?
In my opinion, this development leads to a new role for the artist: that of the Creative Director. Their task is to control the AI systems. Similar to managing an experienced team of employees, the artist can direct and adjust the AI’s capabilities. And this through an intuitive user interface. In other words, they can contribute more or less emphasis to certain aspects in the image or video. This is akin to the iterative design process within a team, where desired results are achieved through meetings and subsequent adjustments.
What makes this development particularly exciting is the fact that images and videos can now be created with far less time and effort than previously required. The ability to create complex artworks with relatively minimal effort opens up new creative horizons and allows artists to express themselves in novel ways.
It is exciting to engage with these tools and to go beyond traditional text prompts. The experiments with the various possibilities that AI opens up in art introduce new dimensions and offer a plethora of creative opportunities for the future.
Comments are closed, but trackbacks and pingbacks are open.