Nominee Threshold Award Autonomous Practices
Meet Lode Dijkers! Lode was nominated for the 2021 Threshold Award Autonomous Practices with his graduation project ‘Antidepress Bots’. In this interview, Lode tells us about his work and future plans.
Can you tell us a bit about your graduation project and how it came about?
My graduation work is called ‘Antidepress Bots’, and it is about how social media algorithms amplify depression in already depressed people. Depression is quite an important subject to me personally. Growing up, I had quite a few friends and close family members who were depressed. Then two years ago I also felt unhappy myself for some time, but luckily managed to get out of it just before the lockdown. But a few of my friends were not that lucky. The lockdown and the isolation it brought along obviously did not help with the mental health issues. And what I noticed was due to the quarantine that the only form of regular social interaction went through your phone and specifically through social media. The kind of posts that you see on social media have a strong influence on your mood and wellbeing. Whereas positive content could make your day, negative content might actually break it. Then I saw a documentary by Danny op Straat, a Dutch journalist, about how algorithms can amplify depression. I actually noticed this myself when I was depressed. Social media posts were pushed to me with a much darker tone. I found this very shocking! These algorithms know your darkest, deepest desires and fears, and go as far as to engage in your mental health. I decided to do something with this in my graduation project.
I developed social media bots that can disrupt social media algorithms, restricting unpleasant posts from reaching depressed people. The algorithms that operate on social media are called recommendation algorithms, and they recommend posts based on your online behavior. They track everything that you do, what you view, what you like, and they categorize you with other people that behave like you. But what this means is that people who are depressed are categorized together and are exposed to the same kind of content over and over again. This creates a ‘depressive filter bubble’, so to say. It is very hard to escape this filter bubble.
During my research I discovered that these algorithms depend on behavior of other users, and that they can be disrupted by implanting fake users into the filter bubble. These fake users are grouped together with depressed people, but behave in a very positive way. This causes people who are actually depressed and part of the depressive filter bubble, to receive positive images instead of negative images. I chose images of clear blue skies to be pushed to depressed people, as a universal symbol of happiness. Besides the peaceful imagery, blue light is also used on train stations as a suicide-prevention method.
I developed the Antidepress Bots in collaboration with Robbert Koning, a Master’s student at TU-Delft in the field of machine learning and data science, to make sure the bots could actually affect the algorithms. He confirmed that it is very plausible the use of bots can create a noticeable difference on social media platforms. This gave me a lot of motivation to carry on with the project. Together with much help from Jasper Kamphuis I managed to actually make the bots run. My goal is to distribute them open-source, so that people can download them and run them on their own computers. More running bots will mean a greater effect on the depressive filter bubble. I see this way of distributing the bots as a form of digital activism. It helps us spread the word and raise against toxic algorithms together. For the exhibition of the work, I hosted the bots on a website where their live activity can be viewed.
Which themes or societal concerns are you addressing in your work, and how?
First of all, Antidepress Bots is about depression and the mental health crisis we are currently facing. As I mentioned before, this is a topic that is very close to me. I am fascinated by the way in which the technologies that we surround ourselves with influence our behavior and our state of mind. Often, we are not even aware that they do. Social media, of course, are the places where this happens mostly. I also looked into the motives of social media companies for using these algorithms. The primary motive is to generate more money, because when we are shown content we like, we tend to spend more time on that platform. To me, it seems absurd that money is prioritised over mental health. So in a way, my project also addresses capitalist society.
What will you be working on in the near future? What are your next steps?
I hope to bring these bots into the real world, get other people to run them too, and hopefully infiltrate the social media algorithms. Moving forward I see that these bots can be used to protect more than just depressed people from algorithms. They can be used to protest against all negative filter bubbles created by algorithms. I am also working together with Jasper Kamphuis to create an experimental design studio called Breadcrumb Trails. As digital artists we will focus on the hidden potentials and impacts of machine learning, and will continue to work together with field experts.
To read more about Lode’s work, pay a visit to his Graduation Catalogue Page. Furthermore, his journey can be followed on Instagram and his Website. The winner of the Threshold Award Autonomous Practices will be announced during a festive ceremony as part of the Graduation Show. Keep an eye on our Graduation Show page for more information.
