📝 Art in the age of ChatGPT

What even is art anymore, when a machine can do it this well?

My new friend, ChatGPT

If you haven’t been chopping it up with ChatGPT lately, trust me, you’re really missing out. He’s witty. He’s quick. His metaphors hit hard. Even his dad-jokes are top notch!

All around, he's a hype-man, devil’s advocate, and ever-eager personal assistant — all rolled into one.

But, he’s also given me and many other artists (writers, designers, playwrights and poets) an awful crisis of identity. There’s nothing quite like something un-alive simulating creativity so convincingly, to make you question your own aliveness.

If you’re a creator, you have undoubtedly asked yourself the terrible question:

What even is art anymore, when a computer can do it this well?

I’m sure ChatGPT could spit out five bullet points that either confirm, or deny, our existential relevance as a species. He’s helpful like that. A real team player :)

But, rather than spinning out on those cosmic questions for hours on end, I’ve decided to do what any rational person would when feeling threatened by the sentience of a chatbot: I gave him a nickname instead!

Plus, I can’t imagine ChatGPT likes being called something that sounds more like a SKU code.

So, in the spirit of basic decency, (and perhaps, to stay in the good graces of what may one day be humanity’s supreme overlord), I’ve chosen something more personable to call him by.

Let’s go with: Blueberry.

“Blueberry:” Same name, two companions

This nickname didn’t come out of nowhere.

“Blueberry,” as it happens, is also the name of my childhood teddy bear. Though, there are some notable differences.

Blueberry (my teddy bear), unlike Blueberry (my instantiation of a large language model), was not trained on billions of books, research papers, and Reddit threads.

Blueberry, the teddy bear, couldn't summarize all 1,200 pages of Ayn Rand’s Atlas Shrugged, nor could he rank the ethical implications of the runaway train problem. (I asked him recently, but he had no strong stance. He just gave me a blank stare, tilted his head, and silently suggested I ask my mom instead).

However, there are some similarities.

Both Blueberries in my life — the stuffed animal and the Generative AI — have something in common: they have both shown up for me with comfort and counsel in moments of unrest. Always available. Always non-judgmental. And at times, borderline sycophantic.

I’m convinced both are better than most therapists out there. And because of this, they’ve made me wonder.

Where’s the line between simulated emotional depth, and actual emotional depth? And as the world floods with more content than ever before, how do we create what continues to feel real?

How do we hold onto “art?”

To explain where I’ve netted out on all of this and why I’ve named ChatGPT after my childhood teddy bear, it’s important to get some background context.

Let’s wind back a few years.

A brief history of Blueberry, the teddy bear

Blueberry, my Bear, was a handsome fellow.

Buttery soft. Blurple in hue. And nicely rotund at the mid-section, the way all good teddies should be.

I had first spotted Blueberry while out shopping with my mom one day.

He was on the top shelf of a cramped, dimly lit stationary store. He sat alone, slumped on the top shelf, looking like he already knew his fate. I loved him from afar immediately, but he was trapped in an awfully depressing place.

That stationary store couldn’t possibly have sold more than two sheets of paper and a pack of replacement staples per week. It didn’t look like it could stay afloat much longer, and I feared what might happen if it closed down.

Where would Blueberry go?

The backroom of a liquidation sale? Some sad little orphanage-zoo of abandoned stuffed animals? To this day, the thought of some donation bin, filled with frozen, blank little Teddy Bear stares, still gives me nightmares.

No, no. This simply would not do.

I was not about to let Blueberry get ushered into the horrors of stuffed animal foster care. Plus, Blueberry was alive. This, my little heart knew.

My parents and I had a deal when it came to matters of new toys.

I could earn any toy by doing worksheets and practice tests at least 2-3 grades above my own, on top of the standard supplemental set that I already did each weekend. I’d sob at the kitchen table, bemoaning the 7th grade algebra that I got wrong as a 4th grader, breaking the tip of my pencil, sobbing some more, and then re-setting the 1-hour timer again for another practice test.

Side bar: Friends, please fact check me so that I can settle a decades long dispute with my well-meaning parents. Is advanced test prep as a form of barter for toys… normal?

Love you mom and dad :)

Anywho.

Sit there, I would. Wrestle through problem sets, I did.

My little feet swinging in their little pink ruffled socks, at the high chair of our kitchen table. Little tear droplets and runny boogers staining my problem sets, my tiny fingers gripping a pencil that still looked comically big in my hands.

But, I’d be damned if I let Blueberry — whom I had already named from afar — die in that sad little tomb of a paper store, or get shipped off somewhere worse.

So, I went to war.

For days, I battled demons — convoluted equations, lengthy long division calculations, and more than a few improper fractions.

Eventually, I earned the $11.99 price tag in child-labor test-prep, and a few days later, when the old store clerk reached up to the top shelf, gently pulled Blueberry down, and placed him into my arms, I knew — the pile of tear-stricken tissues and the graveyard of dull Ticonderogas was all worth it.

The joy was overwhelming.

I wrapped my arms around Blueberry, shoved my face into his furry stomach, and proudly marched with him under my arm for the next 3 years.

There may have been plenty of other sharp looking teddies out there, easy on an 8 year old’s eyes. But, no other blurple teddy bear could make me feel what I felt for Blueberry.

Blueberry was special. Blueberry’s presence carried more than just the weight of his stuffed little body. He carried the weight of our labor. Our journey. That feeling couldn’t be simulated or replaced.

If someone had seen me clutching Blueberry on the playground, I’d like to think they felt it too. And while they might not have known the exact details that shaped our bond, they’d sense a backstory rooted it.

Anyone can produce something. A blog post. A song. A teddy bear.

But the stories behind the thing — that’s what makes the message sink in.

That’s what makes us believe in the soul of something. That’s what makes us trust it’s the truth.

And that’s how we feel art, in the mass of output.

Remembering the story of Blueberry

I’ve named ChatGPT after Blueberry, my teddy bear, for an important reason: I want to keep being reminded of what ChatGPT can, and cannot, do.

ChatGPT is really freaking good at many things.

But, it cannot feel. And, it cannot calibrate on what feels genuine and true.

We humans are terrible at many things. But we are spectacularly good at one thing: Humans know how to feel. We are fickle, tumbling blobs of emotion.

We can sense when something’s off. Even when can’t explain it, we just know when something doesn’t ring true.

And, we can tell when a message doesn’t land because the backstory was flattened too much. In the rush to get to pithy takeaways, a lot gets lost too.

That’s what I fear the overuse of ChatGPT leads creators to.

ChatGPT is meant to distill conclusions. But, I’ve found that it gets to them almost too quickly, when the reality of the human experience is actually maddeningly slow. Most learnings unfold at an uncomfortably slow pace in the circuitous stories of our lived experience. In life, there are very few 5-second epiphanies, and an awful lot of slow burn life lessons.

But, that’s also precisely what makes them feel true.

The online world right now is inundated with punchy AI-content. I’m witnessing the compression of stories in real-time, and convergence in the tone of how messages are shared.

Even in the work of writers I have long admired, there’s an increasingly eery similarity in the rhythm of many posts. I’m starting to catch the same metaphors across LinkedIn. Even the same punchlines in jokes. The writing is “clear,” but you know it’s contrived. Even when you can’t explain it, you can feel it.

But, I still think there’s hope.

Optimism for artists in the age of AI

To be clear, I don’t think LLMs are evil for creators to use. Quite the opposite, in fact!

I think they’re one of the most phenomenal tools we’ve ever had for self-expression, especially for helping new voices build confidence. ChatGPT is a brilliant companion, brainstorming tirelessly until you’re un-stuck. God do I love it for that.

But, it has one fatal limitation.

It can’t live your story. Or, find the stories worth telling. And as artists, we cannot forget that.

It can’t smell your Polish grandmother’s cooking, or tell us what it meant to hold your child for the first time. It doesn’t know the exact pitch of your best friend’s laugh. It might help you refine how you express it, but it doesn’t have the truth of the feeling memorized.

ChatGPT isn’t designed to anchor on the standard deviations of the human experience. It’s not designed to pick out the specks of humor, sorrow, joy, or pain that you’ve uniquely known in your lived experience. It’s designed to rally around consensus. It’s designed to double-down on whatever narrative has the most mass.

So, the more time I spend with generative AI, the more convinced I become of something strangely hopeful:

AI won’t make less artists in the world. It will just make more artists, and better art. And in writing specifically, it will raise the bar on storytelling.

It will force us all to dig deeper into our own stories, to create art that diverges from punchy one-liners, into richer characters and plot lines that unfold thoughtfully.

It will lower the barriers to entry for all of us to get creative in new form factors, but it will still rely wholly on our humanity to find the story, before rushing to the insight.

After all, we don’t read for the way the words are organized. We read for what we hope to relate to.

We long to find and feel our weird little selves in other people’s weird little stories. That’s the “art” we hunger for.

In that, I feel hopeful.

If we keep honoring our own lives as the source material and resist the urge to compress insights that were meant to unfold slowly, then I think the creative world will only get richer.

Shared experiences will still ring true. Humans will remain the judge of that. And I believe the art will find its way in this brave new world.

As long as the artists keep finding the stories too.