Jen Southern

Seeding Things No.3 (August 2021) Vimeo (3.30mins) commissioned by Arthouses for 'LOCALWifi' with mountains made and cared for by participants from Whitley Bay, UK.

Seeding Things No.3 REVISIT (30-31 July, 2022) commissioned by Arthouses for 'ARTHOUSES: REVISIT' with mountains made and cared for by participants from Whitley Bay, UK.

Seeding Things No.3 is a video about growth, care and collaboration made with Whitley Bay residents, ‘Machine Learning’ software, and a pile of seeds.

In June 2021 fifty people made small clay mountains and planted them with grass seeds. They cared for and photographed them for eight weeks as they grew. Machine Learning software 'Runway ML' was trained on these images and learned to make new versions of the grassy mountains, specific to Whitley Bay. The resulting video animates some of the infinite versions created by three generative systems working together: machine learning, the growth of grass, and the communities of Whitley Bay. The video was shown within the Arthouses LOCALWIFI app, streamed live to people's phones at a hot spot in the 'Station Masters Garden' and later archived on the arthouses website.

Thank you to everyone who grew mountains.

In July 2022 the LOCAL:WIFI projects were revisited with new versions of the work, enabled by relaxed COVID precautions that allowed people to physically visit and meet the works. We printed all 1000 images that had been used to train the software, so that participants and new audiences could browse through them, alongside the video playing on an ipad mini. This gave access to the sets of images produced by individual participants, and enabled comparisons between the image set and what the machine learning had generated.

Both these works extend the original 'Seeding Things' project, to make it particpatory, and further explore ideas about care. Through this project I started to develop the idea of 'particpatory more-than-human machine learning'. This development of the original work extends authorship from the original combination of artist seed growth, and machine learning to include other human participants. New features started to emerge, for instance two participants who took a lot of photographs had an orange wall or carpet behind their mountain, which became very evident in the model. The 2021 participants saw how the mountain they had made grew and responded to the amount of watering and care they could give it, and then how their own images fed into the machine learning video. One of the aims of the work was to offer ways into understanding machine learning, through tangible practical actions.