Artist statement I am a research-based media artist with a background in film and photography but I am particularly interested in the intersection of art and technology. By conducting playful research, I hope to get a better understanding of emerging technologies and how I can use these technologies in my work as an artist. I want to get to know the characteristics of algorithms and believe I only can by interacting with and creating them. I tend to have an iterative workflow which means I start with a question, that leads to experiments with code, photography or another medium which leads me to another question. Having this workflow means I often create multiple works in one process, one could argue that the process eventually ís the project. Currently, the main focus of my artistic research is on machine vision. An important question in my work is what it would mean if machines were to see more humanlike. I am interested in new technologies but above all, I am interested in what it means to be human in this ever changing technological world. Project statement In my project An algorithms’ introduction in the white cube I did several attempts to unite the subjective art world with the so-called objective world of big data and algorithms. It consists of three different experiments that correlate with each other. Below I will explain these experiments one by one.
I (The Initiator) My initial question was if I could write an algorithm that produces new successful works of art. To achieve this, I used a lot of public data from the Museum of Modern Art New-York and Tate Modern. The algorithm I wrote works similar to an algorithm used by Amazon. It is called a recommendation algorithm. Instead of it recommending books to buy, it suggests that you should use both a boat and the color gray in your work for example. The thing I was initially interested in was how value is created in a world as subjective as the art world. I wondered if there would exist a sort of objective ‘formula’ to create a successful work of art. My algorithm looked for patterns and frequent combinations in the datasets. It analyzed the colors and objects used in the artworks and outputted suggestions for new works of art. I named this algorithm 'The Initiator', since it can be seen as an initiator for a new work of art. I printed some of the suggestions on museum signs. These are successful works waiting to be made.
II (The Curator) The second part of my project is a multi-screen video installation. For this part of my project, I formulated a new question that came to mind while writing my first algorithm. While examining the artwork data, I noticed how subjective this data was. The data was full of terms like ‘banality’, ‘humorous’ or ‘loving’ to describe a work of art. This sparked my interest because these words mean something different to every single person. This is when I formulated a new question: ‘What would it mean if algorithms become subjective?’ To give an answer to this question I started training an object detection algorithm. Such an algorithm normally sees cars, dogs, people, bikes, cars, etc. But I am not interested in dogs and cars, I am interested in beauty, art, melancholy. Therefore, I started training an object detection algorithm to recognize these subjective things. On the videos in my installation, we see three artists making a work of art. I showed these videos to my algorithm which I named 'The Curator'. On the videos, you can see what The Curator thought of it.
III (The Educator) The last part of my work is an educational, interactive app. I think we’re all familiar with the interface. It is a CAPTCHA test (completely automated public Turing test to tell computers and humans apart). As the name suggests, such a test is used to determine whether the user is human or not. At the same time, the user input is used by algorithms in self-driving cars to learn from. Normally one is asked to select all traffic lights, or parts of an image with a crosswalk. In my version of this test, the user is presented a more subjective questioning. With this app, I want to give the user insight into how an algorithm learns: it learns from human beings. Human beings generate data autonomous algorithms learn from. Therefore, it is full of biases, mistakes and preferences. This app is still a prototype, but eventually, I want to collect the data from this app to better train 'The Curator'. It currently only learns from me which is very one-sided. Letting other people use this app, will enable me to provide 'The Curator' with a more open “mind”. My work is an ongoing extensive research into the subjectivity of algorithms. With the art world as my battleground, my aim is to show that algorithms are as subjective as our own human gut feeling.