Terence Broad is an artist/researcher based in London, undertaking a PhD at Goldsmiths and is a researcher at CCI. He is researching new methods for the manipulation of generative models. His art practice treats models as materials to be experimented with in unconventional ways.
Been making AIArt since 2015. Doing a PhD in AI, researching new tools and techniques using generative AI, with an emphasis on exploring ways of using AI to generate results that don’t resemble things in the training dataset and can’t be generated using other techniques.
Work with AI
My most well-known work is 'Blade Runner - Autoencoded' (2016) which was exhibited all over the world (The Whitney, The Barbican etc) and is now in Geneva's permanent art collection. My series '(un)stable equilibrium' where I trained GANs without data won the 2019 Computer Vision Art grand prize. Two of my recent series of works 'Disemboided gaze' and 'Teratome' are both made using my Network Bending techniques from my PhD research, where I have been analysing GANs and inserting filters inside them manipulate the generation process.
It is a very exciting time to be an artist working with these tools, and I believe there is still a huge amount of untapped potential for really interesting experimentation with these tools. However, in the wider context of AI, I think we are seeing it's limitations and dangers.