Ideas worth spreading

How computers are learning to be creative

2,156,109 views |

Blaise Agüera y Arcas |

TED@BCG Paris

• May 2016

We're on the edge of a new frontier in art and creativity -- and it's not human. Blaise Agüera y Arcas, principal scientist at Google, works with deep neural networks for machine perception and distributed learning. In this captivating demo, he shows how neural nets trained to recognize images can be run in reverse, to generate them. The results: spectacular, hallucinatory collages (and poems!) that defy categorization. "Perception and creativity are very intimately connected," Agüera y Arcas says. "Any creature, any being that is able to do perceptual acts is also able to create."

We're on the edge of a new frontier in art and creativity -- and it's not human. Blaise Agüera y Arcas, principal scientist at Google, works with deep neural networks for machine perception and distributed learning. In this captivating demo, he shows how neural nets trained to recognize images can be run in reverse, to generate them. The results: spectacular, hallucinatory collages (and poems!) that defy categorization. "Perception and creativity are very intimately connected," Agüera y Arcas says. "Any creature, any being that is able to do perceptual acts is also able to create."

About the speaker

Blaise Agüera y Arcas

Blaise Agüera y Arcas

Software architect

See speaker profile

Blaise Agüera y Arcas works on machine learning at Google. Previously a Distinguished Engineer at Microsoft, he has worked on augmented reality, mapping, wearable computing and natural user interfaces.

| Explore

Christopher Olah's blog

Christopher Olah is a machine learning researcher who offers easy to understand explanations of some of the less intuitive aspects of neural networks and machine learning.

| Explore

Andrej Karpathy's blog

Andrej Kaparthy is a PhD Student at Stanford working on Deep Learning. His blog provides a comprehensive overview of the fundamentals of machine learning and neural networks, citing the most relevant papers and interactive learning tools for people interested in using these techniques.

Leon A. Gatys, Alexander S. Ecker and Matthias Bethge | Article

"A Neural Algorithm of Artistic Style"

This journal article is the basis for the 'style transfer technique,' which played a major role in establishing deep learning as an artistic tool. The researchers introduce a deep neural network based system that creates artistic images of high perceptual quality, and propose a mathematical model of how humans create and perceive artistic style.

Blaise Aguera y Arcas | Article

Art in the age of machine intelligence

Machine intelligence is inspiring new approaches to art making. Inspired by Walter Benjamin's writings early in the 20th century about the crisis in art brought on by mass media, this essay contextualizes MI art in the broader landscape of art and technology.

Memo Akten | Article

Review of machine / deep learning in an artistic context

Memo is an artist and PhD Student at Goldsmiths University in London and shares the ways he uses deep learning and reinforcement learning in his own work. This article includes text from his dissertation literature review and helps frame the importance of technology in his creative process.

Eric Kandel | Random House, 2012 | Book

The Age of Insight

A brilliant book by Nobel Prize winner Eric R. Kandel, The Age of Insight takes us to Vienna in 1900, where leaders in science, medicine and art began a revolution that changed how we think about the human mind — our conscious and unconscious thoughts and emotions — and how mind and brain relate to art.

Learn more

About TED Institute

Every year, TED works with a group of select companies and foundations to identify internal ideators, inventors, connectors, and creators. Drawing on the same rigorous regimen that has prepared speakers for the TED main stage, TED Institute works closely with each partner, overseeing curation and providing intensive one-on-one talk development to sharpen and fine tune ideas. The culmination is an event produced, recorded, and hosted by TED, generating a growing library of valuable TED Talks that can spur innovation and transform organizations.

Read more about the TED Institute