Photo: Arun Nura
Recently, a curious,unique set of images popped up on the Instagram profile of Arun R, a Malayali animator and Artificial Intelligence (AI) artist, displaying DC/Marvel superheroes set against the backdrop of the day-to-day life of Kerala. In a series of images posted on his social media handle, Arun presented Batman holding an umbrella and standing in a paddy field in Palakkad, Wonder Woman washing clothes in Kozhikode, and Spiderman sweating it out after getting caught stealing coconuts in Kollam.
Arun's innovative works evoke a sense of two worlds colliding with each other, with never seen, never thought out ideas coming to fruition. However, many netizens were largely unaware of the techniques used by Arun to bring Western superheroes into the familiar topography of Kerala. Many mistook the imagery as choreographed outdoors, with models posing as superheroes and photographers skillfully weaving utopian ideas into reality.
Myth vs reality!!!
The truth is that thee images are 'artificial-intelligence-generated imagery', a rapidly emerging technology that has opened up opportunities to create any work of art using a set of "words." Instead of an artist, an AI model interprets user needs and creates random images using the "keywords" they type into a "prompt." With one or two clicks, AI delivers unique art.
The Oxford English Dictionary defines Artificial Intelligence (AI) as " the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages."
Here, the artificial intelligence text-to-image generators are used similarly to interpret a sequence of words typed in by a user and create an original, sometimes accurate, work-of-art without an artist spending hours doing the same. Thus, the need for an artist has been severely compromised -- a scenario that could render many artists jobless in the near future.
Worrying aspect; time to set boundaries: Animator Suresh Eriyat
"There are several apps and softwares that allow a user to use a few keywords and create images for them. However, one must realise that the final output you receive here is compiled by superimposing images from a pool of millions of art and photographs created by artists and photographers around the world. As a result, even an amateur could use AI-generated images to create art using our works without attributing or worrying about copyright issues," says Suresh Eriyat, a senior animator, director and founder of his animation studio, Studio Eeksaurus.
Suresh also believes that young and upcoming artists would struggle to compete with AI-generated images. In fact, the more worrying aspect of these models -- which create images based on a text prompt -- is that they offer new possibilities for users – including the right to use these final outputs for commercial purposes. Following the developments in recent months and the increase in the number of AI-generated imageries storming social media platforms, members of the online community of worldwide artists called ArtStation began protesting against AI-generated artwork by placing "No AI Art" images in their portfolios. The members, including some prominent artists, demanded the removal of AI-generated images on ArtStation and called for measures to stop AI companies from using their artworks found online.
Interestingly, Art Station refused to pull down AI-generated images but assured launching tags to let artists choose whether to "allow or disallow the use of their art for training non-commercial AI research, and training commercial AI." However, the protest continues around the globe, with artists, especially those active on social media platforms, calling for governments and concerned authorities to act on behalf of creators to bring forth rules to regulate the use of AI to generate images.
Meanwhile, noted visual artist Vimal Chandran predicts that AI-generated images could revolutionise art if concerns over the subject matter are resolved amicably.
"Programs like Midjourney -- a proprietary artificial intelligence program -- create an image from textual descriptions, similar to OpenAI's DALL-E and Stable Diffusion. These programs have trained AI models that reference millions of photos, illustrations, art and others available on public platforms to create the imagery that suits the vision of the user. This is where the problem lies. There is no proper set of rules to ensure that original content creators receive any royalties or rights over their images used by an AI model. Once we resolve this issue, we could really utilise the potential of these AI programs," Vimal Chandran said.
Interestingly, OpenAi, a renowned company conducting research in the field of AI, and developers of DALL-E and DALL-E 2, revealed that more than 1.5 million users are actively creating over 2 million images per day using their deep learning AI models. Artists and photographers are worried that "simplifying the process" would hamper creativity and adversely affect the field.
However, many who have ventured into the uncharted territories of AI claim the field unlocks areas which were nearly impossible to achieve in real life. Arun R says he created AI-generated images out of curiosity i "These images were generated using a programme called mid-journey. I conceived the idea of superheroes in our native surroundings years earlier. However, the latest version of the mid-journey offers more features that open up wider possibilities for using AI to conceptualise ideas, nearly impossible in the past. I started with the image of a Batman, which turned out to be much better than I expected. Netizens were able to connect with the helplessness of superheroes in a familiar landscape relatable to Malayalis," he said.
Though artists differ on the threat posed by AI-generated images, many suggest bringing in checks and balances to regulate the new technology. But all do agree that Artificial Intelligence is a new technology in progress -- which will redefine our future.