About four years ago, when DALL-E was introduced, I found myself wondering why I studied a design degree. EOD (Elements of Design), a subject I had struggled the most in, required me to generate multiple forms from an object of inspiration. Here, I typed in - "an avocado shaped chair" - and within seconds, I had 20 versions of wonky looking green chairs. A couple years ago, it would have taken me a week of forced inspiration against procrastination to make something even close to it.
Come a few weeks later, Midjourney arrived and the creative studio I worked in was abuzz with designers typing in prompts. This was path-breaking. Instead of hiring a conceptual artist who would spend weeks drawing out a single vision, the extensive time taken due to the back and forth of feedback between director and artist, this was a no hassle solution. A genius with ideas could generate their visions within a few seconds. For a moment, I thought that this would change the world for me. I hated the drudgery of sketching out what was in my head, and relying on someone to execute the unseen visuals inside my mind.
A month after the craze had died down, I started working on my first speculative fiction novella. I had a million concepts exploding in my head and raced alongside Midjourney to make them real on screen. My experience was far from satisfactory. I found it difficult to shift these hidden levers that were as invisible as the huge amount of data and energy powering it. Mind you, I had a flair for writing. Language flowed easily, but prompting was hard. Even when I was tasked to generate images for a game prototype, I scorned at the ugly iterations that the image generator regurgitated at me. Maybe my prompting was bad. I didn't have the penchant for AI art. I resigned to my fate and procrastinated on the project. Only once did I try a 'Midjourney prompt generator' with the results being flawlessly fabulous, (with glowing comments on Instagram) but these images somehow limited my imagination. It was not what I saw in my head.
A few months after, the giant - ChatGPT arrived. Asked to design a board game and it could make an awkwardly fair set of rules for a hypothetical board game based on the concept. I tried my hand at working with the LLM again. This time, as a writing companion. It wasn't going to be so I bad, I thought to myself. I brainstormed concepts by asking the bot to make iterations of my ideas. Though they weren't as convincing, I felt eerily comfortable looking at these possibilities rather than idly staring at a blank screen waiting for ideas to arrive through divine grace.
That was my last attempt at trying to work with any of these bots. Since then, I have completely resisted and tried to erase the presence of these technologies in my creative process. Beware the huge elephant in the room that these so called 'intelligences' are literally built on the foundation of data theft, but the scarier part is of artists embracing these technologies as helpful productivity or emotional companions.
There are certain things that how much ever you resist them, intrude into your life by forcing their presence upon you involuntarily(or not). It was college and grant application season - yes, that time of the year. I was manually writing and rewriting essays explaining to institutions about why I was an impressive candidate - something that exhausted me terribly. A friend had suggested I use his Grammarly account (the infamous Grammar checker) and I gave in. I used the checker to better my chances of being chosen amongst other similarly generated pieces but was reluctant of even hovering close to the AI button. No, I would not click it.
Like many of you know, I work in critical tech and design - which means that I make art to critique AI and hence do not use AI. In fact, I believe that we should stop using the term AI ( a successful marketing technique) and make it data theft technology. (DTT doesn't sound bad, but more like a chemical compound or pesticide, like TNT maybe?)
Amidst navigating this landscape of AI hailers and naysayers, you do occasionally meet some with whom you form solidarity. Over the course of my work, I met some artists and technologists critical of such technologies. The project that I fondly remember is by Fred Wordie, a critical designer who created a satirical website called Dear.AI. He had sent me the link and asked for some initial thoughts on it. Dear.AI was a website that used an AI companion to draft and send personalised letters to friends not just by email, but even with an option of sending physical copies using a personalised handwritten font. It was a critical design project designed to gauge reactions from the public about such an absurd service.
At the time, I was finding it hard to express my feelings to a boy whom I had a crush on, and readily thought that something like Dear.AI would make my job much easier. Type in a love letter to a so and so friend whom I have known for so and so years and is in a so and so relationship with me. Choose moods - quirky, fun or compelling. Now, when I look back it, it is frightening. I had also just started therapy, and felt very uncomfortable talking to a stranger about my most personal emotions.
In the feedback session with Fred, I argued that this would strangely be helpful for people like me. I would prefer an AI therapist to a real one. He was shocked, to say the least. Having received similar thoughts from many, his design fiction project had begun to take shape of a service people might have needed. I even remember him arguing that emotional conversations are supposed to be tough and awkward, and having them is what make life what it is and enhances the experience of being human. I did not fully understand what he meant at that point. Fred and I had initially contacted each other through reading each other's blog and newsletters, and he emphasised that humans writing to each other is precisely what forms interpersonal relationships.
A week went by, and Fred's newsletter arrived in my inbox. Dear.AI had been (ridiculously) included in a list of "Most useful AI tools". There were apparently a couple of other companies who were actually doing this as a paid service. Fred's made-up email had received a request from an animal care centre who wanted to use this (joke) of a service to send letters about pet deaths to their owners. Wow.
A couple years later, shedding all my contact with such supposed 'intelligences' or large language models, and after reading a lot of fiction, I started writing again. While I was re-reading fiction that was sometimes a century old or a decade old, I found myself tearing up and experiencing those exact emotions. It was like time travelling to historical periods and living the lives of other through literature. I even wrote the short story project that I had procrastinated upon during the time of introduction of these LLMs. I gathered up the courage and confessed my feelings to my crush. I expressed my emotions healthily to my friends and family. It was awkward and tough, sometimes with words choking in my throat, but I tried my best.
Yet, again, like how intruders barge in a house uninvited, AI entered the Whatsapp chat. On an especially bad day, after a year of eroded connection with my best friend, I typed in the first letter of her name in my chat. A. But AI showed up. This time, with prompt options. I was weak and couldn't resist. "How do I re......" - Predictive text thought faster than me. How do I reconnect with a lost friendship? Message sent. Beep. "I understand that losing contact with a friend can be tough. Here are some ways: ......." I was angry at myself. Deleting the chat immediately, I looked her name up. With great difficulty, I messaged how much I missed her and wanted to talk. Unexpected, we ended up having a genuine conversation after a whole year and I felt relieved. We don't talk as much now, but at least I've made peace with it. Those simple words about how she missed me as well, were enough for my heart. I was proud of myself that I did it. Without a chatbot.
I am currently working on an intensively graphic design heavy project. I hope to sign my name under it saying, "No AI was used in the making of this project."
I also do not judge anyone who uses these tools for making a living or to help their daily work faster. However, as artists I believe we have an immense responsibility to humanity as a whole, on fostering genuine human connection and centering our work on embodied experiences that evoke emotions through the deep spirituality embedded within the depths of our hearts. Not to specify whether it be human or non-human intelligence but an all encompassing intelligence that flows through matter in this universe.
----Swarna
Postscript - My most recent argument against AI art has been that of process driven gratification. For any artist, if one asks the question, "Why art?", many would answer that it is the enormous joy that they create for themselves in the process of creating something. It is not about the fidelity of the output, how grand it is or how many people have been impacted by it (these maybe important external parameters), but in the creative gratification gained by making art.
In recent times, I have seen the rise of artists putting in more tedious hours of work creating a 3d environment on Blender (that may take 8 hours to create and more to render) just for the sake of doing it. I see these attempt as even resisting the rise of automation in the creative industry.
Comments