The photographers’ gallery in London is hosting an interesting experiment in machine vision, artificial intelligence and photography, it is part of Sebation Schmieg’s idea and work and runs until early next year. The artist’s statement has me intrigued and I spent a considerable amount of time on his website input data on the images presented.
They say about their work.
Berlin-based artist Sebastian Schmieg takes a closer look at how machine vision datasets are created: developed on the website of The Photographers’ Gallery, the new commission invites visitors to assign all the images available on the website to one of four categories: Future, Past, Problem and Solution. Through the assignment process visitors are teaching the system how to read and understand images within this existing set of parameters. In a seconds step, after feeding all images and classifications into a machine learning system, Schmieg will synthesise a series of new images that is based on the accumulated labour of all visitors and photographers involved. Each new image will represent abstract visions of Future, Past, Problem and Solution to varying degrees, drawing on the colours, shapes, and concepts of the original visuals found on the website. Through this work Schmieg raises questions around current discussions concerning photography, big-data, tracking, and the hidden manual labour behind algorithmic systems
Not sure what the outcome will really be I guess I could just go to London, or revisit the website at a later date to see what has unfolded. You have four choices to categorise an image, problem, solution, past future. The images are drawn from the photographers’ gallery website.