Building bridges with AI (Part 1) May 2019
I've long been fascinated by the work of Jared Tarbell having first been captivated by his original open source, beautifully composed physics experiments in Action Script. More recently, when trying to find ways of visualising some geodata on a side project at work, I became aware of Casey Reas, one of the founders of Processing.
In spite of my admiration for these artists and love of their work (and even though Processing comes in Javascript flavour) I knew that generative art was not going to be a medium I was going to flourish in.
AI text to image generation became news early in 2018. In 2019 Runway made this more accessible by running the models in a desktop UI and I saw the possibility of taking my printmaking in a new direction.
I'd talked earlier in the year with a local arts group about hosting a spoken word event as part of the local arts festival, themed 'Bridges'. Mytholmroyd and Hebden Bridge are famous for the flooding of the River Calder and, each being at the bottom of the Calder Valley, with two sides separated by the river, bridges are an essential part of our everyday lives.
As well as offering to facilitate the spoken word event, I decided I wanted to make a piece for the festival open exhibition with a hope I could help make it part of our bridge mythology.
There aren't as many 'bridge' idioms or song titles as you might think, which is where I wanted to go with the text to image modelling. I'd originally wanted the AI to do all the work but, I'd also had that the idea that I wanted to present the images as prints. However, just outputting them on an inkjet printer and putting them in a frame didn't make for a complete story.