Support the channel by watching this video ad-free on Nebula: [ Ссылка ]
Artificial Intelligence is enabling a new generation of artists to create bold, fascinating art. But the companies powering it don’t seem to care.
______________________________________
CORRECTION: CLIP, the massive training set used for lots of AI-generated art (including mine), doesn't comprise photos with descriptive captions. Instead, it pulled photos from web pages, and used the text from those web pages (which is usually related, but not always a plain description of the photo). My bad!
______________________________________
Patreon: [ Ссылка ]
Twitter: [ Ссылка ]
______________________________________
Help a trans refugee come to Canada: [ Ссылка ]
______________________________________
Voices by:
Akira: [ Ссылка ]
Bobby: [ Ссылка ]
Vic: [ Ссылка ]
Artists:
Felix Gonzalez-Torres: [ Ссылка ]
Memo Atken: [ Ссылка ]
Katherine Crowson: [ Ссылка ]
HeavensLastAngel: [ Ссылка ]
Colab notebook I used: [ Ссылка ]
Further viewing:
Luis Serrano's intro to GANs: [ Ссылка ]
Simple tutorial on creating GAN art: [ Ссылка ]
______________________________________
Table of contents:
0:00 Introduction & groundwork
3:33 1. The skyline that wasn't
13:08 2. Authorless intent
21:04 3. Art as artifact
28:26 Closing thoughts
______________________________________
tags for the algorithm lol:
artificial intelligence, ai generated art, VQGAN, VQGAN+CLIP, machine learning, neural network, deepdream, video essay
![](https://i.ytimg.com/vi/Bi4sJEE8wCs/mqdefault.jpg)