Computational tools based on Machine Learning / Artificial Intelligence are being utilised in an increasing breadth of applications. Most recently there has been a groundswell of interest in generative visual art, largely driven by advances in the capabilities of Generative Adversarial Networks (GANs), as they require fewer and fewer resources. Such visual productions are primarily achieved through semantic parsing of ‘prompt’ text into inputs for the learning system.
Is it disingenuous for a human to claim ownership or creative agency whilst working in these process-based modes? The trajectory of the system navigating ephemeral intermediary states is not accessible, and the machine operator is not in control of these intermediate phenomena, nor their ‘final forms’. Is the tail also wagging the dog in an aesthetic sense?
“The idea that that AI-generated imagery poses a threat to “authentic” visual expression is interesting. In some instances, we’re just feeding these things poetry. The strange thing is we might see AI as an “emancipated spectator” before we come to see ourselves in the same way.”
Matt Colquhoun, Twitter
Public Arena Board:
Peli Grietzer (interview)
AI And The Limits Of Language
Jacob Browning & Yann LeCun (article)
Algorithmic Realism Salon Reports
0x Salon (articles)
The Trouble With Dall-E
Kevin Buist (article)