We’ve generated the worst visions of geek horror and video games with AI from the E mini


Unless you live in a cave, you’ve probably seen some fun AI-generated images of DALL · E mini lately.

The phenomenon is so popular that the The New Yorker even refers to it in its June 13 cartoon. Even our friends at Bag of Chips have joined the web trend with some very strange things. Difficult to escape.

For those scratching their heads: DALL E mini is an artificial intelligence model designed by Boris Dayma to generate images from any text. The model learns by looking at millions of images on the web with associated captions. Some concepts are learned from memory, but the model can also learn to create unique images that don’t exist by combining multiple ideas.

He gave us a great idea: why not write some crap related to video games and geek culture to see what happens? For science, yes.

Here are the creations of Pèse sur Start, made with DALL · E mini.

Bruce Willis on a Chocobo

Han Solo takes a bath with Chewbacca

Crash Bandicoot by bike

Lady Dimitrescu at Costco

Kirby with a shotgun

Lara Croft eats cheese

Kratos holding a butterfly

grogu smoking weed

Diddy Kong inside gta V

Ganondorf drinks a beer

Link taking the train

Lickitung on the roller coaster

Geralt in a flower field

Battle between Cloud Strife and Tom Nook

Hulk does the shopping

Dragon Ball but at Coachella

Red Dead Redemption but in Quebec

If Vin Diesel was in Ring of the Elden

Waluigi driving a school bus

Tom Nook smokes a cigar

Pikachu smokes a bong

Terminator takes a shower

Super Mario under the influence

Wario, President of the United States

Sonic in the bathroom

Snorlax + bread

Ninja turtles eat sushi

If you want to try DALL · E mini, visit his site and write something funny. You have to be patient, however, as the tool’s popularity sometimes causes web traffic jams.




Leave a Comment