Generative AI is still topic number one. There is hardly an event that does not have AI as an unofficial main topic in the off-stage areas. In Düsseldorf, too, everyone is interested in the image generating AIs like DallE, Stable Diffusion and Midjourney, but those, in turn, are not interested in Düsseldorf at all, as our columnist Markus Sekulla had to find out ...
It started many years ago on social media. In social media lingo, Düsseldorf doesn't get much love. Not from the users, but from the platforms themselves. Example: as a Düsseldorf resident, I was located in the "Greater Cologne Area" on LinkedIn for a long time. Lol, seriously? Admittedly, in the USA, where most of the current AI services and social media platforms are based, the old feud between Düsseldorf and Cologne is of zero interest - except perhaps with Dr. King Schultz.
While experimenting with generative AI imaging, I noticed very early on that virtually all platforms have problems depicting Düsseldorf's skyline or landmarks like the Gehry buildings in the harbor. Above all, the services have a hard time with the Rhine Tower. Sure, maybe I'm writing the wrong prompts, but I've very often found the Berlin, Cologne or Toronto TV Towers in my images – right in the heart of Dusseldorf. Especially Midjourney, but also Stable Diffusion seem to have a problem with this. Feel free to try it yourself and write me your experiences!
There were similar observations on LinkedIn the other day, when Lorenz Gräf from STARTPLATZ Düsseldorf stated that AI doesn't really know Düsseldorf yet. A text-rich discussion with many screenshots broke out.
Text: Markus Sekulla