Look at the new mutant animals that the neural network came up with

The programmer Aldo Cortesi created an algorithm – he draws silhouettes of non-existent animals. Some of which look believable, while others do not look like anything you’ve seen before. This was reported by Futurism.

In order to make a neural network, Cortesi used the PhyloPic database, which collects animal silhouettes. After that, he loaded the entire cache of the neural network with new data, on the basis of which the program was to create new types of animals.

AI animals

As a result, he got a set of believable and frightening images.

A total of 50,000 images were generated. Cortesi sorted them by hand and chose the best.

As a result, he got a huge menagerie with species that range from believable to nightmarish.