Artificial intelligence as a creative partner?
Updated: Mar 18
In the spring of 2022, I already dealt intensively with the AI generation of images. However, I was more than disappointed with the results at that time and could not find any fun in the long run. In the meantime, however, the tools have improved - some of them are even frighteningly brilliant.
Creative freedom or mere imitation? The topic is discussed controversially everywhere.
As an artist, it was always the path to the finished image that gave me the most pleasure. The visualization of my mental image and the creation of a perfect picture have always been a special pleasure for me. However, when I first tried to do this using AI technology, I found that while it can be addictive, it doesn't give the same results as Photoshop. Besides, I just enjoyed composing my own images more than just typing in text (prompts*).
That's why I paused for a while. Now the AI models are getting better and better and I'm figuring out how best to use them for my art and otherwise. It's opening up unimaginable possibilities in new genres that I personally really like, but that people haven't known me to do before. For example, abstract, 3D design, illustrations for children, interior design, architecture, still life and macro photography.
*A prompt is a short instruction or question given to a computer program to cause it to behave in a certain way or to perform a certain task.
However, it is important to emphasize that AI is not yet able to generate what a human eye sees. Anyone familiar with photography, graphic design, images in general, will still be able to recognize an AI-generated image just fine. AI-generated images can look textured because many AI models are based on processing pixels and textures. There is a difficulty in correctly identifying and rendering complex shapes and objects. However, I am convinced that it will not be long before this problem can be solved very well in the medium term.
However, there are also concerns about copyright rights in AI-generated images. Who owns the copyright to an image generated by an AI? This question is not so easy to answer, as it depends on various factors as to how the image was generated and who might own the copyright. The AI model used is clearly at the center of this issue.
However, there are also cases where AI-generated images have no copyright issues. This is the case when the AI model has been trained from publicly available data and information that is not subject to copyright protection.
It is important to obtain information and, if necessary, legal protection before using or publishing AI-generated images.
I'm wondering how I can best use this technology for my current style as an artist. That's exactly what I'm still figuring out, but maybe something completely new will develop in the area of my creations or completely different new exciting projects.
In the last few years, AI technology in image generation has developed rapidly, and it is quite possible that in the near future we will be able to use AI image generators just as we use text in ChatGPT. The technology has already made considerable progress, and there are already a variety of applications that allow us to use AI-based image generators.
At this point I would like to mention that I already use ChatGPT on a daily basis as a useful tool (and with a watchful eye) when creating image titles for my creations, as well as a support for writing marketing texts or blog posts as inspiration. Also for this text I used ChatGPT to support for formulations, but not as a replacement for my personal writing style.
Currently, however, most AI image generators are still specialized for specific use cases and require a certain amount of data to work well. For example, an AI model can be trained to generate portraits of faces to produce realistic portraits. However, the same model may struggle when it comes to generating images of nature or landscapes.
Understanding the subject matter and knowing how to apply it is key to success in AI image generation.
Without the knowledge and experience in prompting, the generated output may be incomplete or incorrect. Creating an accurate prompt* is key to the successful use of AI models.
I have invested a considerable amount of time to bring my AI image generation skills to the current level. Nevertheless, considering the complexity and constant developments in this field, I still feel like I'm at the beginning of my journey.
The way wonderful creations are made will also change as AI technology becomes more prevalent. There will be fewer photographers, digital artists, etc. Am I excited? No. Can I change it? No. So, better to deal with it now than later. You should not demonize it, even if you are not enthusiastic. You can't stop the development and we shouldn't be afraid of it. Instead, we should get to grips with it and learn how to make the best use of AI technology for ourselves.
When I feel an AI-generated image is sufficient, I use my image editing skills to perfect it further.
A good "prompter", a prompt engineer, who knows photography, image editing, montage, etc. will certainly be in great demand in the future. For hobby artists, AI offers many new opportunities. Also, over the decades, there have been many innovations that initially caused fear, but ultimately helped to improve the world. Therefore, we should rather deal with the topic than demonize it. We can't stop development, but we can learn to deal with it and make it work for us.