If you’ve been on social media lately, chances are you’ve seen some AI art, even if you didn’t realize it. AI art programs such as DALL-E 2 and Midjourney have recently gained popularity, but how do they work?
Continue reading to discover all you need to know about AI art, the programs individuals are utilizing, and why some artists are concerned…
What exactly is AI art?
The phrase “AI art” broadly refers to any work of art generated utilizing artificial intelligence, or AI.
There are several approaches to creating AI art, but one frequent approach is to employ a software that uses generative adversarial networks (GANs) to generate a picture.
GANs are made up of two models that were trained concurrently using big datasets, which often include work by notable historical and contemporary artists. The first model (the generative model) generates a picture, while the second model (the discriminative model) calculates the probability that the created image originated from the training dataset.
The produced picture is sent back and forth between the two models until it can no longer be distinguished from the training dataset by the discriminative model.
DALL-E 2 and other common image generators have been taught to grasp the link between visuals and text descriptions. They then utilize a technique known as diffusion to create a picture based on a given suggestion.
It starts as a random dot pattern and eventually changes into a picture when the network recognizes certain features of the prompt.
As you can see in these tweets from the Twitter account @weirddalle, this has resulted in some bizarre, spooky, and downright horrifying consequences.
What tools are people utilizing to create AI art?
Some of the most popular programs for creating art include DALL-E 2, Stable Diffusion, and Midjourney.
Lensa, an Android and iOS app, recently went viral, with many users sharing their own “magic avatars” on social media, whereas Google’s Deep Dream began as a way to help scientists and engineers see images through the eyes of a deep neural network, but later saw a new application as an abstract AI art generator.