AI Art

Art that is created based on the input of attributes by an artificial intelligence. The results are for the most part simply stunning. The speed with which the images are generated is enormous. In a single day, you can create an entire portfolio of breathtaking artwork in a variety of styles. It's addictive to describe different combinations of words, descriptions, color spaces and scenes to see what the AI does with them. Now anyone, even without any artistic talent or at least artistic skills, can create an enormous output of "unique" artworks.

But is this really "my" art. I am torn about this. Certainly the basic concept, the idea / the intellectual property behind the generated image comes from me. However, I am nowhere near as good with drawing, and certainly not at the speed of the AI. That means the idea is mine. But the visualization is synthetic. Will this kind of art generation rob many artists of their jobs in the future? In any case, abstract artists and abstract art have got big competition. So have concept artists for films, computer games & co. Technical progress always ensures that certain activities are automated. The quality of the generated art is sometimes very good but often very rough especially when describing fantastic scenes. My gothic viking ships with green torches are very cool as a thumbnail, but when you enlarge them you notice missing details.

The flying guinea pig
My flying guinea pig is kind of cool - but on closer inspection the winged guinea pig is a bit "strange".

Medieval fortress on mars
The medieval fortress on Mars is good in terms of composition and colors - but details are missing.

The crystal snails
My snails with crystal snail shells are just great, but the AI doesn't get the snail body right. The snail shell is drawn super.

Slimey :)
The black slime monster with a red eye is great.

What to do with it

If you think of AI art and the programs that create it as tools, they can do various, yet rather tedious tasks. The concept artist who has to draw 6-8 concept graphics a day to find a style that pleases clients can save a lot of time. Rough style, an interesting image composition and good color palettes are generated really well by the AI. Here you can quickly generate mass to present to the customer as a basis. If the client likes a certain direction, you can fully focus your creativity on it. This can save a lot of time and nerves.

Personally, I like to create 3D models in my spare time, but I can't draw very well. With AI, I sort of have a personal concept artist who gives me inspiration for new projects. There's something about saying I get a spaceship with certain attributes in the style of a Leonardo da Vinci drawing as a template. The details of the spaceships are usually quite diffuse and leave a lot of room to develop one's own creativity.

Da Vinci style pictures

Imagine Human
Simple experiments are also interesting. How does the AI see the "human being"? If you tell it "imagine a human" . Here are the suggestions and an optimized result.

Imagine earth
With "imagine the earth" the last picture is a kind of view from another planet (Mars?) to a burning earth. That is something to think about.

Update as of October 15, 2022

AI Art is Fascinating. Increasingly, 3D objects and animations can be created using AI. However, the way the AI is trained and the fact that through the AI the style of artists whose artwork has been used to train the AI is almost perfectly imitated is in my eyes a very negative aspect of the whole thing. I have also tried queries with "in the style of Leonardo Da Vinci" or "in the style of Tadao Ando". That gives very fancy results.

But may I tell the AI to produce a kind of "bootleg art" so to speak? These artists, architects, designers have worked on their style for a long time and now it is simply taken over by an AI. Morally questionable. One could still discuss whether it is not as important for an artist who has been dead for centuries as it is for an artist who is still alive and has to make a living from his art. In the end, you can never claim that you are the "AI artist" who created this work of art or has the copyright to it. Personally, I will continue to "play" with AI art generators because I find it really interesting what they can produce. However, I will not use names of artists or other creators in my queries, that is just wrong.

My conclusion
AI art or image generation is here to stay. It is getting increasingly better and more effective. Currently, I don't see it as a replacement for artistic professions and activities, rather as a supplement. Art direction will continue to exist in any case, and when it comes to working out details, even the concept artist, the 3D artist and the motion designer need not fear for their jobs yet. Presumably, though, the expected output will be increased - although it might actually be better to focus on one particular line.

Whether I really consider AI art as "my art" now - debatable. I'll think about that for a while and add to this article.

In any case, the images that can be generated are fantastic. Therefore I will also keep them as a section in my portfolio.

Technical Detail
The pictures in this article are create with Midjourney by parameters provided by me