You have probably had this moment. You type a few words into a music or video tool, and something comes back in seconds. It sounds clean. It looks polished. And somehow it feels like nobody is home.
That feeling has a name, and one of the most inventive filmmakers working today gave it to us. Chris Miller is one half of the team behind Spider-Man: Into the Spider-Verse and Across the Spider-Verse, along with his longtime creative partner Phil Lord. At a press conference in November 2023, Miller called AI “a copying machine” and said plainly: “It’s an imitator, not an innovator.”
In a 2026 interview with The Hollywood Reporter, he went further: “AI can only regurgitate the average of things that have come before it.”
Read that again, because it is the whole problem in one sentence. The average of everything that came before.
What “the average” sounds like
Think about what an average actually is. Take a thousand love songs, blend them together, and you get the most typical love song possible. The chord changes everyone expects. The rhyme you saw coming. The chorus that lands exactly where a chorus usually lands.
There is nothing wrong with any single piece of that. The trouble is that nobody ever fell in love with a song because it was typical. You fell in love with the weird lyric about a specific street. The voice that cracked on one word. The drum pattern that should not have worked and did.
Those are choices. A person made them, usually because of something that happened to that person and nobody else.
The tool is not the enemy
Here is the part that gets lost in a lot of loud arguments online. Miller did not say “never touch a computer.” In the same press conference, he said his hope is that, in so far as AI gets used, “it’s used as a tool to simplify some drudgery.”
Drudgery. That is the line. Not creativity. Drudgery.
And it is worth knowing that the Spider-Verse films themselves used machine learning in exactly that spirit. According to TheWrap, Sony Pictures Imageworks built a tool trained on the animators’ own work on the project to help place comic-style ink lines from frame to frame, a task that would have eaten weeks or months. The tool did the tedious part. The artists kept making the choices.
That is the distinction we care about at PLAIGROUND: tool versus replacement.
Where artists actually lose their time
If you are an independent artist, you already know where the hours go. It is rarely the songwriting. It is the forms. The metadata. The split conversations that drag on for weeks. The copyright filing you keep meaning to do. Checking whether your release is actually ready for the stores. Figuring out where the money went.
None of that is art. All of it is necessary. And all of it is exactly the kind of work a machine is good at: repeatable, rule-based, the same way every time.
So when people ask us whether AI is coming for artists, our honest answer is this. The copying machine is very good at copying. Let it copy the paperwork. Let it fill the forms, track the deadlines, and keep the records straight.
What it cannot copy is you, because there is nothing to copy from until you make it.
Who this series is for
PLAIGROUND works with human artists and AI-assisted artists alike. Plenty of talented people use AI somewhere in their process, and that is not the question we are interested in. The question is who is making the choices. When the artist is the one deciding what the song is about, what it should feel like, and what makes it theirs, the tools are tools.
Over the next four posts, we will look at why the most original work comes from a single, specific point of view, what Spider-Verse can teach you about your next song, why genre fusion is invention rather than averaging, what fans actually pay for, and what you should be handing off to bots so you have more time for the parts only you can do.
Your move: if you want the busywork handled so you can spend more time on the music, take a look at what Plai does at wannaplai.com. It runs around the clock, the same way every time, and you keep 100% of your royalties.
