A Brief history of AI Generated Art
AI Generated art is not a new concept, in fact one of the first AI programs capable of producing art (AARON) was created by Harold Cohen around 1972. The original version of AARON was only able to create black and white line drawings while later versions were also able to use color. Today AI programs are able to create complex works in a variety of styles with quality that is often indistinguishable from art created by a human artist.
They do this by utilizing a technology developed in 2014 called Generative Adversarial Networks (GANs). GANs at their most simple are made of two components. First, a generator AI is used to create images and then a Discriminator AI is used to judge the output of the generator by comparing it to a set or sets of real images. This allows the AI to learn and work with significantly less human interaction.
Further developments in the use of GANs have resulted in the release of a new wave of AI software. These AIs have quickly become popular in mainstream media due to their ease of use and ability to create complex works in a variety of styles. Possibly the most well-known of these AIs is “DALL-E-2". DALL-E-2 works on a subscription model charging the user a monthly fee in exchange for tokens. These tokens can then be spent to generate images based on a text prompt for the AI to interpret. According to Openai's (maker of DALL-E-2) terms of service, once generated the user is given ownership of the image and prompt to use as they see fit. It is important to note however that the U.S Copyright Office has stated that AI generated works “lack the required human authorship necessary to sustain a claim to copyright” and there has not yet been a case of someone enforcing copyright on an AI generated image.
This drastic increase in the use and development of AI programs has not gone unnoticed by the art community. In 2018, a painting generated by an AI algorithm sold at an art auction for $432,500. Then, in late 2022 an artist named Kris Kashtanova registered a trademark for the graphic novel Zarya of The Dawn. This novel was created with the help of the AI art generation program Midjourny. This registration is currently under review by the U.S Copyright Office however because there is doubt over whether there is “substantial human involvement in the creation of this graphic novel”. Both of these events have thrown fuel on the fast-growing debate over the ownership of AI Art and brought backlash from artists who feel like their work is being stolen.
