Critical Topics: AI Images

"Critical Topics: AI Images was an undergraduate class delivered for Bradley University in Spring 2023. It was an overview of the emerging contexts of AI art making tools that connected media studies and histories of new media art, with data ethics and critical data studies. Through this multidisciplinary lens, we examined current events and debates in AI and generative art, with students thinking critically about these tools as they learned to use them. They were encouraged to make work that reflected the context and longer history of these tools.

As a final project, students collected 500-1000 of their own images, cleaning them to create a unique, personalized dataset. Then, using RunwayML, they extended StyleGAN2’s training data with their datasets to create a custom generative model. Along the way, we discussed the politics of image assembly and archives, the human labor of datasets and content moderation, and more.

The course included interviews with AI artists from a variety of perspectives. Students responded to each with short essays highlighting the diversity of thoughts and opinions about what AI art means, how it is made, and the ethics that surround it.

This website collects all of the asynchronous video lectures, alongside works referenced in the lectures. Guest lectures and artist talks are also archived, with permission." - Eryk Salvaggio

Dublin Core

Title

Critical Topics: AI Images

Date

2023-05-12

Contributor

Language

Date Created

2023

Audience Education Level

Audience

Spatial Coverage

United States [n-us]

Abstract

"Critical Topics: AI Images was an undergraduate class delivered for Bradley University in Spring 2023. It was an overview of the emerging contexts of AI art making tools that connected media studies and histories of new media art, with data ethics and critical data studies. Through this multidisciplinary lens, we examined current events and debates in AI and generative art, with students thinking critically about these tools as they learned to use them. They were encouraged to make work that reflected the context and longer history of these tools.

As a final project, students collected 500-1000 of their own images, cleaning them to create a unique, personalized dataset. Then, using RunwayML, they extended StyleGAN2’s training data with their datasets to create a custom generative model. Along the way, we discussed the politics of image assembly and archives, the human labor of datasets and content moderation, and more.

The course included interviews with AI artists from a variety of perspectives. Students responded to each with short essays highlighting the diversity of thoughts and opinions about what AI art means, how it is made, and the ethics that surround it.

This website collects all of the asynchronous video lectures, alongside works referenced in the lectures. Guest lectures and artist talks are also archived, with permission." - Eryk Salvaggio

Hyperlink Item Type Metadata