For this presentation artist and software developer, Sarah Friend explores the subject of human and computer relationships by focusing on the digital interface through the eyes of a neural net algorithm trained by machine learning technology, or artificial intelligence (AI).

Perverse Affordances (2018) takes the form of a machine learning algorithm that has been trained with 10,000 screenshots of popular global social media interfaces to recognise and re-create its own versions of these software architectures. At its core is an adversarial neural net, which is a model in machine learning that can generate new images based on dataset give to learn. Through this development Friend creates a narrative of online habits which are largely shaped by the interfaces used by large platforms, such as Facebook, Twitter and Instagram, and recontextualises this seeming hidden everyday element through the alien eyes of a machine.


In history the interface has been touched upon by several artists over the century. Writer and philosopher Alexander Galloway wrote about this artist effect on this visual technic, expressing that visual experimentation on art’s boundaries played a central focus for modernists, and is a tradition that continues on today.

Read more about Sarah’s work on the Digital U project page here.

If you would like to know more about the Digital U exhibition project and get involved go here.


© Art Gene
Tel: +44 (0)1229 825085

Art Gene

Bath Street,


Cumbria, LA14 5TY England, UK

Donate here

Privacy Policy | Terms and Conditions

Registered Charity: 1135536

Registered in England: 4359454

All images, artwork and text on this website are protected by copyright and may not be downloaded or otherwise reproduced or distributed without the permission of Art Gene. © Art Gene associates, artists and partners, 2022.