unthinking
photography

The idea of abundance in indigenous cultures is often related to ideas of contentment and the realisation that we already have what we need. This contrasts particularly with ideas of aspiration and eternal growth facilitated globally through the use of new technologies.

In 2021 an article was published about an AI experiment in which Democratic congress woman Alexandria Ocasio Cortez’ portrait was fed into an algorithm along with a male counterpart. The resulting machine-generated portrait portrayed her in a bikini while her counterpart was wearing a suit. During the summer of 2021 Bente de Bruin and myself assembled a dataset called ‘Women Not Wearing Bikinis’ consisting of a wide variety of Latin femmes, teachers, firefighters, scientists, mothers, artists, doctors, designers…We wanted to find out what image-solutions a GAN (Generative Adversarial Network) would feed back to us. We started with 250 images.  Initial results were disappointing distortions and, to our surprise, after feeding it more photos, the output didn’t change much. Most of the images have no eyes for example, just eye sockets. By the end we fed almost 5,000 photographs into the GAN simulation.  A varied group of strong distinctive women never appeared on our computer screen.

Results were assembled in a film entitled ‘I am not data’ which also evolved into a live performance with code. These are part of a larger project called ‘Second Nature’, in which I explore bias embedded in algorithms used in web search engines.  Our notes and impressions can be found on the project website: http://www.iamnotdata.site/ 

In the case of this experiment, more was not better. Is there a way of being content with less while remaining inquisitive? Is stewardship of natural resources possible without a digital cloud that contains every single second of our lives?  How is life on Earth being improved by an excess of data?