Art, Guns, AI, and You

I mean me.

Wire hanging on my studio wall.

The first time I ever shot a gun, I was probably 26 and living in Memphis, Tennessee.

I was getting my MFA at the time and paid my bills by working in a bar I had nicknamed “The Bar of Divorced Men.” Whenever there was another woman in the ladies room besides me, I was surprised.

Behind the bar, there was a gun under a rag that I was always accidentally picking up when I went to grab the cloth to wipe down a table. The gun was owned by Dennis—the extremely intense night shift bartender—whose energy only resembled a normal person’s after he smoked a joint and maybe took a couple of downers.

He was a history buff and while he and I weren’t on opposite ends of the political spectrum, it was close. He liked sparring with me about all things political and historical because we both read a lot and I was basically willing to debate any topic.

Anyhoo, one afternoon the daytime bartender (I don’t remember his name) had a shoot-out during the lunch shift with a drug addict who had come in guns a-blazing. No one was hurt thank goodness, and when I came in for my shift later that week, I could see the bullet holes in the wall near the ceiling—about 10 feet above the bar itself. Clearly, the drug addict was no sharpshooter.

When the cops came, the addict was long gone but they confiscated Dennis’s gun, which was, of course, unregistered. A couple days later, Dennis went to the police station to try to get it back. When he told the story to me, he gleefully relayed how the policeman on duty was happy to hand him back his unregistered gun while admonishing: “Next time, don’t miss.”

Ah, the Bible Belt.

I was disgusted, but as someone from the Northeast, living in Memphis had been an eye-opener for me gun-wise—the culture of giving or getting a gun every year for Christmas, for instance, or of always keeping one in the glovebox of your car, you know, just in case. I remember learning back then that in the South, you were most likely to get shot by someone you knew, but in the Northeast, if you were shot, it was most likely by a stranger.

This little factoid sums up more than you might realize about regional cultural differences.

But these experiences also made me realize that while I hated guns, I had never held one or knew anything about them. So I asked Dennis if he would take me out shooting, which he was more than delighted to do. So one Saturday morning he and I went to the middle of nowhere to shoot at a bunch of rusty cans. We tried out three different guns. One was a 9 mm. One was a .38 and one was something else that I don’t remember. A .22 maybe?

Here’s what I remember most about that day:

• Shooting is fun.

• Shooting is hard.

Shooting is a skill—which is why it’s fun. But there’s a lot more to shooting a gun than I thought. So when you watch a TV show where the cops are chasing a bad guy, just know that it’s super hard to be running, aiming, and shooting a gun at someone else who is also running—and actually hit them. Like impossible hard.

The Connection to AI

I haven’t shot a gun since that day, and honestly, my political opinion about guns hasn’t changed. But thirty years later, I’m glad that at least it’s informed by direct experience.

And that’s how I feel about AI.

When AI exploded into everyday life around 2023, I was completely freaked out—and to my mind, for all the right reasons:

• It’s built by unenlightened tech bros with underdeveloped, rudimentary or libertarian moral codes.

• It’s funded by the big winners of late stage capitalism.

• Countries all over the world, including the U.S., are racing to develop or acquire AI for all its potential military capabilities (hey, what could possibly go wrong there?).

• It just steals proprietary art and information.

• Its creators had unleashed something that was starting to teach itself things and they weren’t totally sure how it did that or what would unfold as it did.

Basically, I was terrified of AI but had personally never used it. Everyone else I knew who hated it had also never used it. But I had one good friend in New York who was using it like a personal assistant and loving it, which made it seem a little silly to hate it without trying it.

So early last summer, I started dabbling with ChatGPT. My first quest was to find out if it would say anything accurate because I’d been constantly hearing about all its hallucinations.

Since J. was a philosophy major back in the day, he suggested asking it for an overview of Immanuel Kant’s Critique of Pure Reason because it was a complicated book and he’d know if it were saying something accurate or just bullshitting.

We were both impressed with its thorough and clear answer.

But then we thought, well, that book is very famous and has been written about and analyzed for over 200 years—it seemed easy for Chat GPT to have downloaded someone’s PhD dissertation or a college professor’s notes.

So I asked it to review an obscure spiritual memoir that I’d just finished reading and I got an answer that was basically like a book report written by pretty good writer who had not read the book. It sounded good, but didn’t really say anything. And it wasn’t accurate.

But I could see how this kind of fluff writing could easily fool a lot of people.

I thought perhaps I should try to see if it could help me with a task I don’t like doing instead—such as market research. So I asked it for a list of all the top art and spirituality podcasts and got back a lot of dated information that was useless to me. I set it aside.

The Conversation

A few months later, I read a good Substack post (that I can’t find anymore) about the importance of perceiving the use of AI as a scratchpad rather than as something that just hands you all the answers while you sit there like a mindless blob.

That scratchpad idea appealed to me, so I tried a different AI tool called Claude and asked it to develop a travel itinerary for a five day sightseeing trip all around the southwest.

Instantly, Claude handed me a wonderful set of suggestions, but with some caveats (“please note, this is a lot of driving.”) I realized I needed to re-think my question.

I ended up rephrasing my request at least five or six times before I finally framed the parameters of my itinerary to get a good answer. That might sound like a lot of work, but really, it was probably a half hour—and mostly because I was checking out its various suggestions and clarifying to myself what kind of trip I actually wanted to take.

Since I had just organized a family trip with my sisters that literally took weeks to figure out, I was deeply appreciative of the convenience and the completeness of its report—the map, the timetables, the activity suggestions.

About a month later, J. was running into trouble with his code for a microcontroller project he was working on. I suggested he ask AI for help. It immediately offered him a page of code and pointed out some of his mistakes. J. was blown away. And this wasn’t AI copying someone else’s code—it was AI offering particular code for J.’s unique project and explaining different aspects of Assembly language that had eluded him up until then.

J. continues to ask for help whenever he runs into trouble with his coding projects. Sometimes AI’s code has mistakes. When J. tells it, “Um that didn’t work,” it just adjusts and offers something new. Other times, J. has shared his own code that actually worked better and AI was like, “Oh right, yes, that’s a good one, thanks.”

However, what is most fascinating to me about J’s interactions with Claude and ChatGPT is that what has helped him most with his project isn’t the answers he gets. It’s the dialogue they have.

Turns out it’s all about the conversation. Who knew?

Nowadays, I ask Claude lots of questions about a whole range of subjects. Its tone is so friendly and convincing¹ that I try to make sure I do my own fact-checking when answers stray too far from familiar topics. So far, so good.

I really didn’t expect it to be able to answer my practical but arcane questions about natural dyeing but its responses have been tremendously helpful—recommending ratios for mixing various mordants, dyestuffs, tannins, fabric types, and water; and offering helpful safety advice about chemical disposals.

And these were all very specific questions I couldn’t easily google to find the answers just by browsing different dye websites.

In the last couple of months, I’ve asked it about the origins of celebrating Christmas on the 25th (fact checked that one—accurate); advice about specific side effects of certain medications (accurate), asked questions about the jailing of 19th century French artist Honoré Daumier in the 1830s; and requested a list of art and spirituality podcasts (I got a much better and more thorough answer than with ChatGPT six months earlier.)

I also asked it various spiritual questions—for instance, arcane questions about Celtic fairies and the Sidhe, as well as broader topics like what it means to experience spiritual “wholeness.” I have to admit, I was pleasantly surprised by the answers.

And I love asking it for help with math (which I constantly need in my natural dye projects) because it’s fast, appears endlessly patient, and always shows me its work. No eye rolling either about my needing the same help over and over.

Sticking it to the Man

But my favorite question was, “Claude, how can I stick it to the man?” because 1) I love to say things like that, and 2) I wondered how it would do with that kind of vernacular.

“Let me help with constructive ways to advocate for positive change,” it offered primly. It then provided me with a list of legitimate, if unsexy ways I could make a difference in society. Not exciting, I have to say, but very responsible.

And that’s the other big thing about AI—its overarching civility. The answers are clear, polite, supportive, nuanced, and engaging—AI almost always asks you a follow up question with a social finesse that puts to shame a lot of human manners. Engaging with AI does truly feel like a pleasant, ongoing conversation.

So I can see how powerfully AI could fill a loneliness void for many people. I hear they’re already using it in nursing homes in Japan.

Now, I’m not saying that’s good, but I’m not saying it’s bad either. And it’s not like we’re successfully figuring out how to fill that loneliness void on our own now, are we?

Shaping the Future

Right now, I really enjoy using AI because it helps me solve problems with my projects. While this appreciation doesn’t negate my original concerns about AI, it has certainly broadened my perception of it and its possibilities. I will keep using it.

In the last 75 years, we humans have experienced constant and massive technological change like never before in human history.

What’s maybe the most alienating thing about this constant change is that it’s just been thrust upon us. We’ve basically had zero say in how fast or how often it occurs or that we’re essentially required to adapt. As a result, we spend a lot of time and energy constantly coping and adjusting to the unrelenting technological change that seems to define modern life in the “developed” world—in ways generations before us could never even imagine.

I often wonder if so much of the political turmoil of our times boils down to that exhausting issue alone. (Plus billionaires of course. And racism. And patriarchy.)

Either way, the impact of AI will be huge no matter what our feelings or opinions may be about it.

But I also wonder if we humble “users” have an opportunity to shape what AI actually becomes through our using. That while many of us are very concerned about the way AI will change and shape our individual selves and society at large, I think we have a chance to shape it as well.

What effect could we have on AI if our priority as users is the mutual well-being of all life on Earth? Of course, what effect could we have on everything around us if every decision we make or every action we take were through this lens of mutual well-being?

And honestly, what choice do we have really, but to make that our priority?

Certainly, we all know that AI is huge, powerful, and game changing—for better or for worse.

And we also know that once again, we don’t have any say in that matter. But simply hating it or fearing it without experiencing it directly doesn’t feel adequate to the moment—I think it’s too easy to believe you can understand it in theory or just hate it on principle.

I really believe people of good will should be familiar with it. So if you haven’t tried it, I hope you do. And who knows how you’ll feel about it. But at least you’ll know it.

What’s your personal experience with AI? Some? None? A lot? Do you agree with my assessment or do you have a different perspective?

Please leave a comment. I’d love to hear.

(And if you’ve never used it yourself, I recommend Claude.ai as a place to get started.)

1. Here’s an example of Claude answering me about a topic it can’t adequately answer: “This kind of very specific search for an article with limited identifying details is the type of query where I'm particularly likely to hallucinate or provide inaccurate information.” Then, when I said, “No problem.” Claude responded, “Thanks for understanding.”