What has been true for a long time, because you have been able to manipulate images with Photoshop for some time, becomes even more obvious with AI: Don't trust an image that you haven't faked yourself. That's why it should be said right here at the beginning: All images in this news report were generated using artificial intelligence.

AI (Artificial intelligence) or in German KI (Artificial Intelligence) is on everyone's lips. I was seized by curiosity and I brought something with me Mid-journey experiments, one of the leading tools for converting an image description into a photo. You enter as detailed an image description as possible in English, and the system uses it to generate four variants of an image. The source for this is apparently five billion photos circulating on the Internet. The whole thing is currently still in a legal vacuum because, with the exception of one provider, the tools use images for which they have no right to use. This isn't a problem for experiments (yet), but what happens if a company or brand uses AI-generated images in which one or more of the original images are still clearly visible? It is only a matter of time before courts have to address the question of whether an image generated by AI is a completely new work or whether a license fee must be paid to the authors of the photos used.

Depending on the subject, the images or at least parts of them still look very artificial. Sometimes everything isn't quite right - for example, take a closer look at the hands in the last picture (click on the picture for an enlarged and complete view). I'm pretty sure that this will improve in the future, and my limited knowledge of AI image generation probably also has an impact on the image quality.

Why the Pope and the Dalai Lama?

I have taken the liberty - and may I be forgiven - of using the Pope and the Dalai Lama as a starting point for this article. I wanted to create an image that could exist like that. This earned me criticism from a dear friend who I respect very much. Because he rightly says that this image could spread further on the Internet, and at some point the image is taken so out of context that people no longer know that it is an AI-generated image. He suggested labeling the image as an AI photo, but unfortunately the AI is very capable of eliminating any annotations in an image. The latest versions of Adobe Photoshop show how frightening and fascinating the image editing programs have already become in this area. Another acquaintance thought the idea with the two spiritual leaders was very good and suggested generating a similar image with Putin and Zelensky - as a contribution to the peace process... This feedback shows the great danger of this technology. Nobody knows anymore what is real and what is not. At the moment, the AI-generated images still look a bit artificial, and here and there - as mentioned above - they have small errors. Therefore, upon closer inspection, you can often determine whether the image is a real photograph or an AI image. But in my opinion it is only a matter of time before the AI-generated images can no longer be distinguished from an original photograph.

What does this do to us as a society?

When I talk to friends about AI-generated images or show them how to use AI to generate photos, I feel two emotions: fascination and frustration. Fascinating because it's exciting that you can describe a scene in words and the AI generates an image from it. The sky (or your own imagination) is the limit - a trite saying that also proves true here. We take the wings of technology to fly higher and higher. But after a little thought, the Icarus effect often occurs with my conversation partners: they crash. They then formulate sentences such as “Now I can no longer trust a picture”. Image editing programs have already greatly reduced the belief in the photo as a reflection of reality. And it seems to me that the AI completely destroys what faith was left. In a society that is already unsettled by the events of the last few years and the tense world situation, this new technology further increases the uncertainty. Due to the abuse of extremist groups, absolute values are also frowned upon, but now that everything is becoming relative, we feel a loss because there no longer seems to be anything to hold on to. Maybe that's a bit dramatic, and AI-based images are perhaps just one piece of the mosaic in this whole web of uncertainty, but it seems to me that this artificial generation of images suddenly makes the uncertainty more tangible and real. As a society, we are facing major challenges, and technology – which also brings a lot of positive things, which should be explicitly mentioned here – certainly cannot be stopped. We must not lose authenticity as people and as a society. A lot has been and is being written about this, and it challenges each and every one of us to confront the question of authenticity. Interestingly, the photography industry saw this development coming and launched an exciting initiative:

Content credentials

Content credentials is an organization that, as I understand it, provides the RAW file in the camera with a code that allows the original file to be recognized as an unedited original. This means you can prove at any time whether an image has been manipulated or not. In the future, photographers will have the option of having the content credentials code added to their images when taking photos. The first company to incorporate content credentials into a camera is Leica in the Leica M11P. How the whole thing works you can find out here.

The results of my experiment

As a curious and open person who works in advertising and photography, this technology is definitely important enough for me to pursue. The art lies in formulating the so-called “prompt”. “Prompt” is the name of the image description that you enter. My first attempts with AI-based image generation showed me that even small changes in the text can have a big impact on the image. “Prompting” is a kind of programming language in which you describe the images. The crux of the matter is that sometimes you have one image in your head and then the AI generates something completely different. You can upload your own images as a style template to Midjourney, but my first test with an uploaded image showed that the style of the image was adopted to some extent, but otherwise I didn't recognize my photo in the AI-generated version (which is for sure the copyright issue is very good for users...).

Here are the first results of my experiments with Midjourney: