Transcript:
↳ Babylonian Vision

Nora Al-Badri on “Babylonian Vision”

With my work I am expanding on a speculative archaeology and a decolonial machine learning based museum practice by generating technoheritage. You can see a pre-trained neural network based on GAN technology. That means general adversarial networks, that was trained with 10.000k images from five different museum collections with the largest collections of Mesopotamian, Neo-Sumerian an Assyrian artefacts, GAN technologies also known as for its deep fake image generation.

So, in my case in the work here the images were in the majority collected through web crawling and scraping without the institutions approval. But the two of them had an open API where I could just use the data sets for training, the neural network.

Subsequently which you can see here are new synthetic images evolved as a living memory of the images. So the machine learns basically the pattern and the form and the „Bildsprache“(imagery) from the Mesopotamian area and then generates new objects and images of objects of that same area.

So the generated image is at the same time the artefact itself. And materiality is very important even though you can see a digital version only, because the input images are images of material objects of our past. And if machine learning is seen as a technology performing in processing our collective memory it absolutely to me makes sense to apply it to our, what we could call it big culture data of the past and to generate new images as traces and circulating image worlds.

And applying machine learning to culture big data supplies other more speculative and abstract insights into the search of a visual language form in pattern of an area within a specific spatial context: Babylonian. So the input images of this data basis in this case actually from Iraq all of them carry time and memory themselves because the input was also patina broken pieces most of them mid to low resolution.

Subsequently the new synthetic images evolve as a living memory of the images. What was important to me is also the question of decolonising databases and the, you can call it southern data basis, because yeah and to contribute to a few charity of a country like Iraq that is usually very much hunted to buy distraction and to contribute a few charity that is rather influenced by creation of new heritage and techno heritage.