Centre for Internet & Society

Gene Kogan will give a talk on "A.I. hype cycles and artistic subversions" on Friday, January 22, 2016 at the Centre for Internet and Society office, 6 pm - 8 pm.


Gene Kogan - Style Transfer - Mona Lisa

Mona Lisa restyled by Egyptian hieroglyphs, the Crab Nebula, and Google Maps. Style Transfer. Gene Kogan.


Recent years have seen a resurgence of popular interest in machine learning and artificial intelligence, as emerging methods have set new scientific benchmarks and introduced classes of neural networks capable of imitating human behavior, among other impressive feats. More importantly, the study of these algorithms is rapidly crossing over into mainstream culture and industry as AI applications begin to inhabit more of our daily lives. Numerous initiatives have appeared, attempting to demystify and make these previously obscure research tracks more accessible to the public. Open source software like Torch, Theano, and TensorFlow have equipped amateurs with the same software which is achieving state-of-the-art results in industry and academia.

This talk will examine the most recent wave of artistic projects applying these methods in various cultural contexts, producing troves of machine-hallucinated text, images, sounds, and videos, demonstrating a previously unseen capacity for imitating human style and sensibility. These experimental works attempt to show the capacity of these machines for producing aesthetically meaningful media, yet challenging and subverting them to illuminate their most obscure and counterintuitive properties.

A recent article by the speaker about this: From Pixels to Paragraphs: How artistic experiments with deep learning guard us from hype.

Relevant projects by the speaker that will be presented include: Style Transfer, A Book from the Sky 天书, Learning to Generate Text and Audio, and Deepdream Prototypes.

Gene Kogan

Gene Kogan is an artist and programmer who is interested in generative systems and applications of emerging technology in artistic and expressive contexts. He writes code for live music, performance, and visual art. He contributes to numerous open-source software projects and frequently gives workshops and demonstrations on topics related to code and art.

He is a contributor to openFrameworks, Processing, and p5.js, an adjunct professor at Bennington College and NYU, a former resident at Eyebeam Art & Technology Center, and a former Fulbright scholar in Bangalore, India, 2012-2013.