Skip to content

How DALL-E 2 turns a sentence into a picture

  • Article
  • Generative AI

Read the announcement on Facebook
Title slide: DALL-E 2, Hierarchical Text-Conditional Image Generation with CLIP Latents, by Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen, on a purple background with a head silhouette and a server room.
1 of 4

Is AI-generated art even art? DALL-E 2 creates images from a text prompt and with surprising accuracy! The AI system uses state-of-the-art Diffusion Models in its image generation and CLIP in finding the link between linguistic and visual information. In this pub, you’ll see an in-depth look at how the AI system is structured and its performance against its predecessor.

Query console

Connected to dssoc. Type \h for the commands this console understands.