Skip to content ↓
MIT blogger Rona W. '21

how i feel about ai-generated art by Rona W. '23

from an artist's perspective

photo of ai-generated hands

why is ai so bad at making hands?

Recently, I read something by a fiction writer that went something like, it’s so upsetting how people can just enter these trivial prompts and then pass this work off on their own! I’ve spent so much time learning how to write well and then this random guy in his mom’s basement will generate a Pulitzer-Prize novel with a few keystrokes.

As somebody who has spent many years learning about the craft of writing, I certainly understand the frustration—if my goal was to produce well-written sentences, then perhaps the advent of ChatGPT would mean that all my efforts had been wasted. But AI-generated writing doesn’t bother me, and if somebody were to create a story generated by AI “written in Rona’s style” or trained on my publicly available work, I wouldn’t mind—I’d be amused, more than anything else. Of course, this is only my reaction, and certainly other creatives might not feel the same (and a quick glimpse at Twitter would prove that they don’t feel the same).

Artists’ ire is two-pronged. One, artificial intelligence poses a professional threat. I’m privileged in that I never expected to rely on writing as a steady source of income, and if everything goes as planned, by June, I will hold bachelor’s degrees in mathematics and computer science from MIT. I’m uncertain how I would feel if this weren’t the case, so it’s possible that, under the stress of economic insecurity, my opinion would be vastly different, but my current belief on this topic is: AI is absolutely a professional hazard to artists, as it is a hazard to software engineers, data analysts, executive assistants, and a menagerie of other professions. It is a hazard in the same sense ride-sharing was a hazard to taxi drivers and photography was a hazard to painters. All this is to say, technological innovation reshapes our processes and systems, and it inevitably threatens current jobs. This does not mean we shouldn’t move forward.

But even if you can agree with that notion on an intellectual, abstract level, that doesn’t mean it isn’t incredibly stressful to see your career prospects shrivel up. One of my friends predicts that AI will lead to widespread unemployment, which will bring about civil unrest—and while I hope this won’t be the case, I can certainly understand the underlying rage that comes with losing a future you worked so hard to achieve.

Two, artificial intelligence is creatively humiliating. It can hurt to see a machine churn out a derivative of your work, to know that a soulless language model trained on snippets of your soul to produce said derivative work.

I have limited knowledge of copyright law and how data is used within these models, so I can’t comment on the legal ramifications here (and I suspect that the government will move to enact new legislature in this direction). I also don’t want to undermine other people’s emotional responses, so I only speak for myself here. I don’t quite see my work as an extension of myself, or as a deep excavation of my inner life, or perhaps the more accurate statement here is that I view the process of writing as distinctly different from the product of writing. Translating the confusing, limitless galaxies within me—that process is intimate, a conversation between the page and me, and that journey won’t disappear simply because a computer can generate pretty paragraphs. The joy of reaching for an apt metaphor, of conjuring a visceral image, of understanding myself more fully—that remains regardless. Once those words are written, once they exist as separate entities beyond myself, they are for the reader. And if the reader is a robot that wants to feed upon my dangling participles only to spit out an imitation, that’s okay too.

Cross-posted on Substack here.