I create this piece by exposing the machine to the hardware of the human body, by curating a dataset of about 60,000 images of human surgical dissections, and letting a machine creates its own representations of it. The surgical dissections are of various kinds and parts of the human body, to create the overall aesthetic of the 20 piece panel. The title of the work (The Anatomy Lesson of Dr. Algorithm) is a reference to one of Rembrandt’s earliest, and most well known paintings, ‘The Anatomy Lesson of Dr. Nicholas Tulp’. That painting was done at a time of troubled fascination with medical technology, and today we are in a parallel time of troubled fascination with Artificial Intelligence, debating about topics like how much of human(ity) should a machine be exposed to. In such a time, I create this piece showing a version of the machine’s representation of humans, which may not be anything more than its hardware of flesh and blood, rather than a glorified version of mind and soul.

The panel of artworks are created using a Generative Adversarial Network (GAN) with a custom curated dataset of various kinds of human surgical dissections, and by sampling the network after various intervals of training.

http://harshitagrawal.com/