For the past several months, people in and around the tech industry have been marveling at a tool for creating automated images. Called Dall-E (derived from the names of the surrealist artist Salvador Dalí and the cartoon robot Wall-E), the software produces original pictures from text prompts of as many as 400 characters or images that users upload. Someone might ask for a portrait of Shrek in the style of the Mona Lisa, or upload a file of the painting Girl With a Pearl Earring and ask Dall-E to imagine it as a behind-the-scenes glimpse at a fashion shoot starring its subject.
Like many successful products that come from Silicon Valley startups, Dall-E became a phenomenon during a testing period when it was available to only a relatively small group. The hype built with online chatter from early adopters, who documented the highlights on Twitter and Reddit, giving the broader world a taste of what was to come.