Adapting to Gen AI

How I feel about Generative AI doesn’t matter. I may as well hurl toothpicks at elephants for all the good it will do. The market has spoken.

Over the past 3 years, Generative AI has gradually upended the market for creative skills. Now illustration, films, writing, code and music can all be generated by AI in minutes, so it should come as no surprise that people from these industries don’t look kindly at Generative AI There has been considerable backlash against AI technologies by artists.

What I did not expect was that people who use Generative AI to create art, would hit out at artists in retaliation. I suppose this blind spot was to be expected. I am an artist, so my own view would be to look at AI as the disruptive force, but the vitriol that AI prompt-artists direct at non-AI artists took me by surprise.

Artists have good reason to distrust and fear the phenomenon that threatens careers they have developed over a lifetime. That they hate AI art for it is not a mystery. But what possible reason could be driving the hatred of AI prompt-artists for traditional artists?

So I joined a few subreddits that support AI art to try and understand the point of view of AI prompt-artists. Here is a screenshot that illustrates the kind of discourse that one may find in these subreddits. This one is from r/DefendingAiArt.

From what I understand, it is the lack of respect for AI art that feeds the vitriol. AI prompt-artists feel that the art that they create is seen as less worthy, low skill and low effort by the rest of the world, and they blame traditional artists for this. If traditional artists did not disparage AI art, they believe that everyone else would treat it on the same footing as other non-AI art.

As justification, prompt-artists compare earlier inventions such as photography, digital art and so on to justify that AI art requires just as much talent as creating art using traditional or digital tools.

I don’t think so. These earlier disruptions still required the use of drawing tools. The difference between wielding a real paint brush and a Wacom stylus imitating a real brush was a matter of convenience—specifically, avoiding the messiness of traditional paints—though an understanding of colour theory, composition, brush work and basic drawing skill are all still prerequisites to creating art.

With Generative AI, one prompts the software, using lines of text to describe the expected outcome. The artist has no direct control over the artwork, and has to settle on the closest match to their expectations. However, this is not as easy as it sounds, either. Prompt-artists sometimes spend hours describing and fine-tuning their prompts over and over again until they achieve an outcome they are satisfied with. So, to be fair, there is effort involved, but of a different kind.

I created this image around 16 years ago. Today it makes me cringe a little, but back then I was proud of it. It took me a week, using dozens of layers in Photoshop.

This image took about 3 minutes to generate using ChatGPT. Notice the detailing, especially the reflection in the water, and the lighting. How can artists expect to compete with tech like this?

Still, the difference in approach is jarring. The idea of creating art using words is hard for me to digest, but I think the days of making pictures for a living are behind us. I don’t think the vast majority of artists are in a position to compete with the capabilities that Gen AI provides. Picture making is no longer even a commodity. It will soon be just another utility, like water or electricity. Something that is expected in a technologically driven, modern society.

People who run tech firms know how to use technology to deregulate markets. A decade ago, the Karnataka Government had regulations in place that did not allow autorickshaw drivers to overcharge. Meters were compulsory in autorickshaws to ensure that regulations were followed by the drivers. Uber and Ola used technology to deregulate this, bypassing legislations. Now, the concept of metered charges for autorickshaws don’t apply in any place these tech firms operate in. There were never any regulations in place to protect artists, and even if there were, tech would find a workaround. I know it’s better to simply adapt. I’ve done it before.

Around 1991, I got a job as an illustrator at MAA Bozell, a large advertising agency. My job was to illustrate storyboards for TV commercials. It was around that time that MAA invested in a DTP studio to ‘computerise’ their layouts. Print advertising layouts at the time were all handcrafted using type set on bromide, which would later be sent out to the printing press. I was in my late teens and also among the earliest to pick up on computers. From that vantage point, I saw the slow decline of expert studio illustrators who created pen and ink and watercolour illustrations, losing their jobs to a younger generation that used tools like Photoshop and Fractal Painter.

Toward the end of the 1990s, visualisers from advertising agencies were leaving their jobs in droves to the better paying web development industry. I made the jump too, joining a web development firm as a creative director, and then to another one to head their information architecture division. Once again, I had to learn a new set of skills as technology upended the market.

All this is to say that I’m a survivor. Each time technology played hooky with my career, I made it a point to leap over to wherever the grass was greener and now, it’s happening again. This time, the scale of disruption is unprecedented. As uncomfortable as it is, Gen. AI is here to stay and picture making has become just another technologically driven utility available to everyone.

I used it to generate the pretty picture at the top of this post.