August 31, 2026

No, AI Isn’t the End of the Creative Economy

Objections to LLMs — that they were built on work taken without consent, and that they devalue the work of creators — are valid. They’re also the oldest story in art.

By Christine Miao

The Lady Snowblood poster beside the snow-garden duel between Uma Thurman and Lucy Liu in Kill Bill — *Lady Snowblood*, a film about a woman’s violent revenge, heavily influenced Tarantino’s *Kill Bill*. Uma Thurman’s fight scene with Lucy Liu in the snow was a not-so-subtle love letter to the original work.

AI has a problem: creatives hate it, and are opting out.

Their objections are serious: that LLMs were built on work taken without consent or compensation. That they now compete directly with those same creators. That they devalue creative labor with shortcuts and slop.

All of the above are valid, but they rely on a faulty premise of how creation actually works. Consider:

Who made the chairs in your home?
What inspired your favorite movie?
Who designed the font on your computer?

Miranda Priestly’s team assembling looks in The Devil Wears Prada, the scene behind the cerulean monologue — *Can you tell me the origin story of your sweater?*

Nothing is created from zero. Creativity is always derivative because it’s driven by experience. Just by living your life, your perspective is being shaped by untold influences, which were in turn shaped by other untold influences.

When you create from that perspective, a few things are happening all at once. They’re worth separating, because we almost never do:

Transformation: how much did you change the original work? Were you simply influenced by it? Or was it replication?
Fidelity: did you honor the original intent of the work? Or did you hollow it out?
Attribution: did you acknowledge the source, or did you erase it?

Frida Kahlo’s self-portraits alongside Mattel’s Frida Kahlo Barbie doll — Mattel’s representing Frida Kahlo, whose self portraits emphasized and celebrated her conventionally unattractive features, as a literal Barbie doll.

These three separate dials move independently. But often, we flatten the whole thing into one small, dirty word: copying.

Copying in itself is neither good nor bad. Go to any museum and you’ll see artists replicating the works of the masters. Studying and reproducing what came before is a prerequisite to creating skillful, original work.

And it can feel good. Open source software, sharing with creative attribution and aligned intent, tends to generate warm and fuzzy feelings all around - no matter how much the end product is transformed.

Even when the original creator loses, copying can be morally right. In 2001, Cipla reverse-engineered the formulas of Western pharmaceuticals and brought AIDS medication to the Indian and pan-African markets at a massive cost reduction - $12,000 a year down to $350. Western pharmas cried foul, but it saved millions of lives.

Braun radios designed by Dieter Rams lined up beside first- and fourth-generation Apple iPods — Dieter Rams’ pocket radios would eventually inspire Jony Ive’s iPod, which would become an icon in its own right. (Source)

So copying is an essential part of the creative process. And in turn, the end goal of the creative process is creating something worth copying.

When you create distinctive, relevant work - you add to the cultural dialogue. Your work then becomes the inspiration and influence for what’s next. The influenced becomes the influence.

At that point, being copied isn’t a risk: it’s a certainty.

But certainty doesn’t make it feel any less shitty.

Portraits of Janis Joplin, Big Mama Thornton, and Elvis Presley side by side — Janis Joplin and Elvis Presley would both cover songs originally written for and performed by Big Mama Thornton - *Ball and Chain* and *Hound Dog*.

Making a career as a creator in any field takes tremendous strength of heart. Most of the time, you’re building something that no one else cares about. You then need to share the thing (that no one cares about) publicly, all the time, in a bid for relevance. Then, in the face of apathy, criticism, and even ridicule, you need to try again.

You endure this for decades. Finally, something breaks through - your work is recognized, seen, appreciated.

… Only for it to be ripped off, stripped for parts, with your name erased.

It can feel like having your heart torn out with a chainsaw.

An AI image editor turning a photograph of a woman in a flower field into a Studio Ghibli-style illustration — All the Ghibli generators in the world does not a Miyazaki film make.

And now, thanks to AI, you can rip someone’s heart out with the push of a button.

No wonder so many creatives hate it. No wonder they believe that it cheapens and disrespects the artistic process.

But look at what the objection assumes: that AI absorbs other people’s work in a way that human creators don’t. That’s just not true. Humans absorb other people’s work constantly and often without asking. For artists, who have gathered since time immemorial in salons, collectives, and guilds - absorbing other influences is the entire point.

The 1923 Bauhaus Weimar exhibition poster beside the Wagenfeld table lamp

Marcel Breuer’s Wassily chair beside a geometric Bauhaus painting — The Bauhaus was formed of its namesake artists who gathered to influence one another across architecture, design, and art. Source.

There’s a reason these collectives still produce distinct, unique work. No amount of influence (or magic AI buttons!) can replace artistry: the taste, judgement, and accumulated perspective that originated the work.

AKA: You can dress, sing, and dance like Beyoncé… but no one will confuse you with Beyoncé.

The creative process - influences in, transmutation, then re-expression via artistry - hasn’t changed. What AI collapsed is the cost of that loop: access to influence, magnitude of references, and the speed of iteration.

What also hasn’t changed is attribution. Whether work is healthy derivation, homage, or icky theft still comes down to human responsibility. “AI” can’t steal work - but humans and corporations using AI can.

Rodin’s The Kiss, a photograph of Auguste Rodin with Camille Claudel, and Claudel’s The Waltz — Camille Claudel worked at Auguste Rodin’s workshop, and later became his muse, lover, and collaborator. It’s impossible to untangle the full extent of their mutual influence, though Rodin signed many of her works as his own. Left: Rodin’s *The Kiss*, Right: Claudel’s *The Waltz*

This is not the first time a technology made copying exponentially cheaper and everyone concluded that creation was over.

In 1906, John Philip Sousa published an essay called “The Menace of Mechanical Music” and testified before Congress that recorded sound would destroy musicianship and hollow out American culture. The phonograph, he argued, sings and plays for us “in substitute for human skill, intelligence, and soul.”

Two Strothmann illustrations from Sousa’s essay: a couple in a canoe listening to recorded music, and a gramophone rigged over a crib — Sousa’s essay also included illustrations of terrifying future scenarios: like kissing under the moonlight to recorded music, or using a gramophone to put a baby to sleep.

In many ways, Sousa wasn’t wrong. Home musicianship dramatically changed. Piano sales fell off a cliff, and the parlor tradition of making your own music largely died out.

But Sousa couldn’t see the industry that would grow around recorded music. Instrument sales fell… but you no longer needed a piano to enjoy music at home! He was right about disruption, but couldn’t see the full story because he was in the middle of it.

Today, we’re all in the middle of the AI story. We don’t know the precise ending, but there’s enough historical precedent to make out the arc.

Finneas O’Connell and Billie Eilish laughing together in a bedroom studio — Finneas O’Connell produced Billie Eilish’s Grammy-winning album on his laptop in their childhood home in suburban Los Angeles.

One theme: when the cost floor of creation drops, more people create.

The other theme: creators panic. With each wave - the printing press, VCRs, desktop publishing, even cookbooks - creators assumed that more creative activity meant devalued creation. But… that’s not true. For example:

Apple: when Steve Jobs put beautiful UX in every pocket via the iPhone, he didn’t cheapen design. Instead, it birthed UX and UI as distinct design practices. The rise of the profession then proliferated “UI kits” of reusable components - which made good design even more attainable…. and expanded the field further.

Craft beer: In 1978, the US had just 89 breweries. Then Sam Adams made better beer mainstream, raising American palates. Better palates meant demand for better beer, which led to an explosion of small, local breweries. Today, there are 9,800+ breweries in the US.

This pattern has a name: Jevons paradox. In 1865, economist William Stanley Jevons noticed something counterintuitive. As steam engines got more efficient, total coal consumption went up, not down. Efficiency lowered costs, and cheaper power meant everyone found more uses for it.

Abundance doesn’t saturate markets, but expands them. We make more, want more, and expect more. And the people best positioned to ride this expansion? Existing creators who already have the hard-won artistry, skill, and strength of heart to make things.

These are the same creatives opting out of the AI wave right now.

It’s not just a lost opportunity, there’s a real cost:

Sousa’s testimony on the “menace of mechanical music” would later help inform the 1909 Copyright Act. He didn’t have the whole story, but his active engagement helped shape timely legislation that would help his fellow creators get paid out over the next century.

Musical sampling took the other path. As synthesizers, DJs, and hip-hop proliferated, it was courts - not artists - that wrote the rules. Gregory Coleman’s 1969 drum break became one of the most sampled recordings in history. His band never saw royalties, and he died homeless in 2006.

Rules written with creators in the room get built for creators; rules written without them rarely do.

Some creators are opting out because of sheer despair. They’re feeling the apathy of today’s average consumer - happily, unwittingly buying a knock-off without a second thought.

But as the AI wave mints a new crop of creators, I think more people will understand why attribution matters. You don’t really appreciate what it means to steal until you have something worth stealing.

And this more educated consumer will live in a market where attribution has never been easier. Giving credit used to be expensive: you had to know the field, chase the lineage, or be old enough to remember.

Now you can just ask the LLM of your choice.

Who made the chairs in my home?
What inspired my favorite movie?
Who designed the font on my computer?

That’s the twist of AI, creation, and attribution. AI makes copying faster and cheaper. That’s not necessarily bad, because copying is the essence of creation. Nothing substitutes the hardest part: artistry.

Stealing is something humans have always done, and will continue to do. AI makes it faster to steal, but it will make it easier than ever to give credit back.

All you have to do is care enough to ask.

I use Claude in every post I write, including this piece.

I always ask for better references and ideas. It found the Sousa “mechanical menace” example - an anecdote I never would have come up with myself.

When I get stuck, I ask for multiple drafts of specific sections. It saves me hours of toiling through different versions.

I found this piece especially hard to structure. I asked it to create a spine - which helped guide the final format. It came up with the idea of this footnote.

As much as it makes things easier, it also doesn’t. Every piece takes me months to concept and write. I need the judgement to sort through the thousands of ideas and suggestions that AI generates. I need to be able to say “not perfect, but good enough to publish”.

Then… I need to post it on LinkedIn, which will never not make my soul cringe.

AI has made it easier than ever to create. Creation is still really, really hard.

Still, I wish it on everyone.

With or without AI, I hope you make something you love.

A huge thank you to Steven Machado and Sofia Hou for reading and proofing versions of this.