Scott Griffith, 9 November 2024
Introduction
As I have solidified into my full ‘adult’ form, I have found myself deeply interested in process-oriented hobbies. I have a long history of tinkering with Illustrator designing t-shirts; some that my formally-trained-and-actual-artist partner has said are “pretty good.” Sitting down at my computer to translate a concept in my head into a digital form of a design is life-giving, but the real magic happens when I descend into my basement where emulsion, squeegees and tiny threads come together as tools of screen printing.
The first step in getting a design onto a shirt is the emulsion development process. The end-goal of this introductory step is to have a screen with a hardened stencil, ready to be put on the press. You start with a bare screen, nothing but potential, you then have to apply a thin and even layer of this thick paint-like substance. After this undeveloped emulsion dries, you can then start the process of burning your design. After getting your digital design onto a transparency (for people older than 30: think of the old school transparency projectors from chalk-filled elementary school rooms) you then arrange your soon-to-be masterpiece between a powerful UV light source and your virgin emulsified screen. The next step is where you do the least work, but apply the most stress. You turn on the UV light and burn the design into the emulsion. The unmasked portions solidify, becoming hard and opaque. The masked bits will eventually be washed out and form the ink well that will be pressed onto your shirt.
Before you can start slinging ink around, you have to let the newly-formed stencil dry. So, at this point you take a break to run some errands. As you are walking down the aisle at Target looking for some new underwear you walk past a cleverly designed and perfectly printed shirt of your favorite pop culture property, something that your self-trained design skills can never match, and of a quality your barely functioning hobby print shop will never approach.
At no point have I ever stopped to consider “Wait, why am I doing this if I can buy a better, cheaper product for a big box retailer?” The same is true for the amount of time I have spent in my woodshop, haphazardly building things that IKEA can do faster, cheaper and better. The same is true for my truly sub-par pizzas that I bake in my home oven, when I live right down the street from one of the best artisanal pizza places in Spokane (Perry St. Pizza, represent!). The same is true for my arduous offset smoker that requires a full day of hourly tending to make a passable brisket. I do all of these things not for the outcome, I do them because I love the process.
Outcomes vs. Process
In discussions around the use of Large Language Models / Generative AI, I am always drawn to application: What will I do with this neat thing? Well, I could use ChatGPT to aid in developing questions for my coding classes. I could use image generating tools to make just the perfect, and yet still witty, illustration to augment a presentation. I could use AI systems to automatically grade assignments. I could use a well-trained chat-bot to replace advising conversations. Heck, I can just train a chat-bot on my voice and have a computer talk to my students when they come into my office looking for a conversation.
I have yet to be tempted to do anything with generative AI, both professionally and personally, because I love what I do. I deeply enjoy talking with students. Does it take time? Yes! Could I be more efficient if I leverage an AI agent? Probably, but at what cost and for what outcome? The vocational enjoyment of my craft is what gets me up in the morning and motivates everything I do. It informs my professional identity, and it pushes me to hone my abilities to be better. I have spent time and intention to increase the amount of time doing things in my life that give me fulfillment and purpose, while I have also become increasingly disciplined in avoiding things that don’t give life.
The lack of generative AI use in my life is not a rejection of technological progress (some of my research involves genetic algorithms, which fits into the broad category of artificial intelligence and machine learning). The absence of generative AI is because it does not offer what my effort otherwise does: meaning and fulfillment …or at least I am too stubborn to cede that ground so that I can still stand in my self-righteousness.
Value of Process
Ok, fine… it is not very compelling for me to sit here on a proverbial porch yelling at the new-AI-kids playing on my lawn. However, it is becoming pretty well understood that people may like using AI, but people don’t like receiving something generated by AI, in particular, when it is in-place of the work of a real person.
When my colleague Pete Tucker talks about the practical value of junior developers over AI, there is a subcurrent which relates to value. I would imagine that if I were to tell my students that all of my lectures were developed by AI, all of their homework assignments were written by AI, all of their assignment feedback was filled in by AI, all of the emails I have been sending them were written by AI, they would reasonably ask “Then what am I paying tuition for?” While I agree with Pete (I think I am more of an expert in my field than ChatGPT), there is also an emergent intellectual uncanny valley: my students don’t want to be mentored and taught by a computer because mentorship and teaching are human activities. The human-ness is what is valuable about the interaction, especially in a context like Whitworth, where we value and implement relational education. I am different than a book in a similar way, there is more going on here than simply a transferal of information.
Imagine you go to an art gallery. There are three rooms. All three rooms have visual, two-dimensional, art pieces. You walk through all three rooms. Like all art, some of the pieces resonate with you, some you can’t connect with, but you would judge all of similar quality. You eventually decide that you want to purchase one piece from each room. When you go up to start this transaction, the gallerist informs you that one room is all pieces developed by local painters using traditional techniques, one room is all pieces that are high-quality reproductions of original work, and one room is all pieces generated by AI. Would you be willing to pay the same amount for all three pieces of art? Of course not! My guess is that a ‘yes’ to that answer would only come from the most die-hard AI evangelist. Graffiti is whitewashed, while a Banksy piece is protected, insured and sold for more than the value of the building it was painted on. We care about the source of a thing.
Role of Privilege
At this point it is imperative to point out the massive pile of privilege my perspective sits on, and possibly some long-term irresponsibility. I am in a position in time and space that affords me the rare ability to make these kinds of choices. I have the freedom and resources to pursue inefficient hobbies, I have latitude in my professional work operate in whatever way I want. I respect that is not true for everyone. This high-minded (or stuffy) call to ‘commit to the craft’ is easy for someone whose station does not involve manual labor or minimum wages. In this world that is increasingly on fire due to overconsumption, I question if my inefficient print shop is helping or hurting the situation.
This context illuminates that my privilege masks the incoming inequities of an overly reliant AI future. Instead of thinking deeply on how my fellow human may be impacted, I instead retreat to principled stands on erudite feelings of meaning. I wonder if scribes in the 15th century were given the space to consider their craft, or if the story of John Henry is about his love of hammering or a rejection of the cruelty of machine.
Charge
So, what should you do about this? I implore you, the next time you write out a prompt, to think about the task you are pursuing. Is the AI agent you are interacting with helping or hurting your experience? If your thought is “Well, this is a menial task… it doesn’t matter.” My response would be to inspect the original goal for the outcome. If the outcome is menial, then why are you doing it in the first place? If your heart was not in the task to begin with, is it really worth doing? Will filling that meaning-gap with AI be valuable to anyone?
I can assure you: the grass is still green on this side of the AI fence.