Jessica Fuentes and William Sarradet discuss the wider implications of machine learning in the creation of digital images.
“I worry that we’ll get caught up in discussing how many rules need to be placed on these technologies, when sometimes you can describe the issue simply by saying ‘it’s a machine — it can only solve specific problems.'”
To play the podcast, click on the orange play button below. You can also find Glasstire on Apple Podcasts and on Spotify.
If you enjoy Glasstire and would like to support our work, please consider donating. As a nonprofit, all of the money we receive goes back into our coverage of Texas art. You can make a one-time donation or become a sustaining, monthly donor here.
Related Reading:
—NY Times: Meet DALL-E, the A.I. That Draws Anything at Your Command
—Machine Learning at Berkeley: Understanding VQ-VAE (DALL-E Explained Pt. 1)
—Machine Learning at Berkeley: How is it so good ? (DALL-E Explained Pt. 2)
—DALL·E: Creating: Images from Text
—The Verge: OpenAI’s DALL-E AI image generator can now edit pictures, too
—TechTalks: DALL-E 2, the future of AI research, and OpenAI’s business model
—DALL·E mini: Generate images from text
