AI ART Hackathon Finalists


Lilia Li-Mi-Yan and Katherina Sadovsky

Link to more art

The artists explore the theme of the possibility of human interaction and connection with other forms of existence. What would happen if we had a new body created through interaction with new technologies, materials, and bacteria? Will we be eternal, and will we remain the same humans? We are concerned with the question, what will happen to the emotions of the new person, the posthuman, the cyborg...? Will we be able to refuse to reproduce ourselves?


Lerie Pemanagpo

Link to more art

Link 2


I started my way in AI visual art in the beginning of 2021, when my friend troduced me to the first GPT-based Russian-made text to image model "Malevich" (nicknamed RuDall-E, as it was based on the same principles as the Open AI's first version of Dall-E). Even though it's dataset was relatively small (only 600GB) and representative power rather awkward, I immediately fell in love with the poetry of its prompting language: it was purely intuitive, irrational, and thus demanded purely artistic approach. I think that the experience of creating with this particular model affected the way I approach AI art. At that time a single 512x512 generation took about 30 to 40 minutes, so there was no way to "try 100 generations and pick the best". In order to achieve favourable results, I had to switch off my logic and turn on my 6th sense. I also noticed that I only had those results when I was in a very particular state of mind, so my prompts needed to defy the algorithms of the model's training.

Abhay Singh