Generative Art Timeline · Chapter 7: Net Era (1990s)

1995

Machines Dream in Pictures

Publication · North America

Geoffrey Hinton in Stockholm, 2024.
Geoffrey Hinton in Stockholm, 2024. Photo: Arthur Petron, CC BY-SA 4.0 (cropped), via Wikimedia Commons.

Generative AI? It begins here. In 1985 Hinton and Sejnowski’s Boltzmann machine could dream but not in pictures. A decade later, Hinton’s wake-sleep networks did. Awake, they studied handwritten digits; asleep, they made up their own: among the earliest images a neural network generated. The pairing of a network that recognised with one that imagined returned in 2014’s variational autoencoder.

Text by Peter Bauman. Generative Art Timeline, Le Random.

Sources

Movements

artificial intelligence

Filed under

Ideas
ai history
Technologies
artificial intelligence, generative AI, neural networks
Organisations
University of Toronto
Other subjects
Brendan Frey, Geoffrey Hinton, Peter Dayan, Radford Neal

Related moments

Le Random editorials

Explore all Le Random editorials and artist interviews

Le Random podcast