Technology
PUBLISHED
1
There has been a lot of excitement recently about image-generating algorithms, particularly Imagen by Google Research and Open AI's DALL-E 2.
The rest of this article is behind a paywall. Please sign in or subscribe to access the full content.Both systems use a blend of natural language processing artificial intelligence (AI) with massive image training sets to produce extremely impressive results, whether you want to see a painting of a fox sitting in a field at sunrise in the style of Claude Monet or a majestic oil painting of a raccoon Queen wearing red French royal gown.
However, with most people unable to access these two systems, there has been a resurgence in people using DALL-E 2's predecessor DALL-E mini. Though not as impressive or realistic as the latest software, it's accessible and comes up with some pretty pleasing and, depending on the input text, surreal or messed up results.
It turns out, the limit is pretty much your imagination. Below are a few of our favorites, as well as several that will make you wish people didn't have imaginations.
Check out more on the Weird DALL-E subreddit. Or, let's face it, every other post on Twitter.
