Emerging Faces (2017)

Two AI Agents cooperated on a simple task:

1: The first tried to imagine faces and paint them on a canvas.

2: The second stopped the painting as soon as it recognized a face.

The algorithm was inspired by conversations artist Pindar Van Arman had with Harold Cohen. Cohen, a pioneer in artificial creativity, held the belief that all claims of AI “creating” imagery were false. Instead, he insisted that AI merely acted as a filter of preexisting imagery, no more special than a Photoshop filter.

At the time this may have been true, but Van Arman had begun working with a new kind of neural network called GANs (Generative Adversarial Networks) and noticed that they were “imagining” things very similar to how humans imagined. He shared this view with Cohen and tried to convince him that maybe machines could finally be creative.

Cohen passed in 2016 before they could finish their conversation, but his critiques ultimately sparked a series where Van Arman's robots began to truly be creative. The synthetically creative systems that emerged are an attempt to see the world through the eyes of AI.

Emerging Faces (2017) consists of 9 unique canvases from this groundbreaking series. Historically, this series represents the first examples of paintings autonomously created from neural networks. This and related works have been shown around the world, awarded first place in RobotArt 2018, and been accepted into LACMA’s permanent collection.

Emerging Faces (2017)
Pindar Van Arman
Acrylic on Canvas
37x46 inches

Executed in 2017