News

The experiences I've had with AI image generation over the last few weeks and months have been really exciting and I feel like the pace is getting faster every week.

For me, this journey started last year, late summer 2021 to be precise, when I was looking into style transfer techniques derived from Google's Deep Dream algorithm to realize an NFT series, more specifically I was looking for a tool to quickly and effectively edit my photos in a certain direction. Actually, at that time I had more of an idea to use some kind of Photoshop filter ;) and indeed the workflow was more or less similar to a conventional image retouching.

Here is an example of this workflow, my "Mystic Island" series with 1/1 NFTs on foundation, which is a homage to the beautiful island of Rügen

Another series with this workflow I published on opensea.io. In this "Trippy Graveyard" series I worked on my photographs of graveyard statue, which I took over the years on old cemeteries of Berlin.

This was to change in April 2022 when I became aware that there was more ;) I and started experimenting with google colab and disco diffusion. I was immediately fascinated by the possibilities and was "intoxicated" by the results and started to switch the process of my NFT production to AI art.

Then came Dall-E 2 and Midjourney and I was allowed to be a beta tester relatively early, the results fascinated me at the beginning, but the chaotic-artistic moment, the unpredictability of disco diffusion was missing to me, that was of course my very subjective impression ;). When finally Stable diffusion popped up and I was allowed to experiment with it again early in the beta stage I was completely hooked, the approach of making an AI tool freely available and the prospect of being able to work with it independently with Google colab notebooks didn't let me sleep anymore.

Meanwhile I have a colab Pro+ account ;) and love to experiment with different notebooks with prompts and settings and tweak around until I get a result that satisfies me. But there is one subject in particular that I love, as a photographer I love landscape photography and - although far too rarely - the feeling of being out in the field looking for the right place, the right angle and the right lighting conditions. But I also love the direct feedback that working with AI tools gives me: Instead of looking for the right place, I now look for the right prompt ;) and enjoy the series that come out of that search.

A lot of these images could be found on exchange.art

The answer to whether it is art or not is up to each person to answer for themselves, but isn´t that true for every creative work? For me, AI art is like the beginnings of digital photography: just because you can reproduce things doesn't make you an artist. To turn an AI image into a work of art, you need imagination, a vision, an idea, and a lot of skill with other tools.