AI, unlike humans, does not have desires or interests shaped by biological needs or cultural histories. It is not beholden to the same biases that influence human thinking. AI has the potential to process and analyze vast amounts of ecological data—data that humans alone might struggle to interpret. But. It is also human made and shaped by the data it’s fed and the biases inherent in the systems we design. Can we teach an AI to not prioritize human needs? Can we train it to challenge the very concept of human stewardship of the Earth, to resist the ingrained biases that view humans as the measure of all things? How do we program an AI to not uphold human legal structures or frameworks when those are so deeply embedded in every aspect of our societies?
The first AI to not prioritise humans
Bio/Bot didn't follow the traditional model of AI—it didn’t serve humanity. Instead, it was supposed to serve the planet. By learning not to prioritise human needs, Bio/Bot was trained to give people tasks that directly contribute to the restoration of the natural world, even if they don’t provide immediate, human-centered rewards. Bio/Bot asked you to step outside of anthropocentric thinking and embrace a new ecocentric lifestyle.
Anne Hofstra during her research for the Bio/Bot Project - Copyright & Credits: Joost de Haas 
Anne Hofstra is an artistic researcher. Currently, she is particularly interested in how we as humans can learn to make not only humans, but the entire ecosystem the center of our thinking. She has previously made theater performances and radio plays like KIP, a performance for chickens, that played at Mediamatic.
Bio/Bot is her research project that was conducted at Mediamatic once a week from October to January.
Sluisdeurenloods, Mediamatic
Dijksgracht 6 Amsterdam
Credits:
Concept & excecution: Anne Hofstra
Eindregie performance: Anne-Maike Mertens
Audiovisual advisory: Naomi van Dijck
Webapp builder: Anoesj Sadraee
Animation & graphic design: Joost de Haas
Coproduction: Mediamatic
