The artists created all sorts of surreal compositions using the OpenAI DALL-E artificial intelligence tool, which creates images from a few words of descriptive text.. But some AI enthusiasts have used it for a different purpose: to “extend” classic art with a machine learning model.. First of all, to show art lovers what the picture would look like if the artist undertook to expand the mise-en-scene around. The new invention is reported by Vice.

The experiment began as an attempt to replicate a work in progress with George Washington on a dollar bill.. Development initiator and director of AI neural. love Denis Shiryaev, using a Telegram extension that gives access to the closed DALL-E test collection, began to create models of "

[see_also ids\u003d"

The results mimic a camera zooming in on the subject, filling in parts of the background that the artist never rendered.. Now you can already see the result of \? "

Experiments with Rembrandt's "

This is not the first experiment using machine intelligence to play with cultural heritage, in particular famous works of art.. Last year, the Rijksmuseum Amsterdam used AI to digitally update Rembrandt's 1642 painting The Night Watch, which was cut in 1715 to fit in a doorway.. Other researchers have found three paintings by Gustav Klimt thought to have been lost in World War II during a Nazi heist..

DALL-E has also been used to completely reimagine classic paintings when the program played it in different styles..

[see_also ids\u003d"

“In terms of art, I think that these results do not give us much,” Shiryaev said.. “But for me personally, it was a pleasant journey to feel the same, familiar art, but from unexpected sides.”.

Recall that a painting by Ukrainian artist Maria Prymachenko, sold for $500,000, was handed over to the National Art Museum.