AI got too good at producing within the existing frames. What's left isn't making within creation, but creating creation itself — inventing the way of making. Today we walk that path.
A few months ago a student said this to me: "Midjourney made something in 30 seconds that blew away the illustration I worked hard on. What am I even supposed to do now?" I didn't have an answer for a long time. It stung. Her effort felt wasted, and I felt the deflation too. That night I went home and worked out an answer. I'll share it today. You should be inventing the rules of creation itself. Let me go slowly.
This essay covers three things: why being skilled within the old frame is no longer a weapon, what to do instead, and concretely how to start.
Let me set up what's happening now. For the last 100 years, creators all played the same game. Get better within the frame.
Novelists polished sentences within the novel's frame. Painters refined brushwork within painting's frame. Designers found balance within graphic design's grammar. Directors cut scenes within film's form. The winner of this game was "mastery." Long practice wins.
Since 2023 something shifted. Midjourney started making images indistinguishable from 10-year illustrators. ChatGPT started writing grad-level essays. Suno started composing. Mastery is no longer uniquely human.
This isn't a sad story. It's a story of the game changing. Don't be #2 in the old game. Be #1 in a new game.
At every such turning point in history, those who changed the rules themselves remain. One example.
Yi Sang (李箱). A building engineer at the Governor-General's office in 1930s Korea. He stood inside the most rigid and hierarchical system of his era. And he did something strange. He wrote poems, but they weren't poems as people knew them. Numbers and symbols mixed in, sentences inverted, radical whitespace. People at the time said "this is not poetry." Because it was outside the frame.
Today Yi Sang is a starting point of modern Korean poetry. What he did wasn't writing poetry better. He redefined what poetry is. Not the #1 within the rules — the creator of new rules.
Mastery gets copied. The invention of rules does not.
That's the core of this essay. AI copies up to the top of the existing frame. So you move the frame itself.
Here's an easy way to see it. Think about soccer.
For the last 100 years people trained to run faster, kick more accurately, defend better within the soccer field. World-class athletes are #1 of that field. Now imagine robots start playing soccer. Robots don't tire across 90 minutes, shoot with 95% accuracy, defend perfectly from data. Humans can't win in that field.
So what do you do? You redraw the field. Its size, number of players, goal position, rules — redefine them. The robot is back at the starting line of the field you just made. That's what creation needs to do.
Abstract talk isn't enough. Let me show what I'm experimenting with.
I'm building a new creative rule called "seed-based multi-format publishing." The old game was "one book = written cover to cover in one go." I threw that rule out. Instead, I made this one.
This rule let me stack 800 seeds in the past year. From them, three books came out, and 40 video scripts. Impossible productivity under the old rule. Because this rule is built around a personal time axis that AI cannot replicate. A single seed takes AI 3 seconds. But the network of 800 seeds lived through a year of one person's life — AI can't produce that.
That's a new field. I stopped fighting AI on its field. I drew a new one.
You can do this too. Ask yourself three questions.
One — "What is the basic unit of my work?" For novelists, "one novel." Redefine it. What if it's "one paragraph"? "One sentence"? "One word"? Change the unit and the game changes.
Two — "What order do I take for granted?" Writing is usually intro→body→conclusion. Drawing is sketch→color. Flip or remix the order. That alone becomes a new rule.
Three — "What material am I not using?" In your field there are materials that are collectively ignored. Use them. Math symbols in poetry (Yi Sang). Images in novels. Silence in lectures. What was considered "inappropriate" until now is the start of a new rule.
Honestly, inventing rules is scary. Inside the frame the evaluation is clear — good or bad. Outside the frame there's no standard. People look and say "what is this?" Yi Sang was mocked in his time too.
So the first three months are the hardest. When I started seed-based publishing, I got "why not just write one book at a time" comments non-stop. Nobody says that anymore. The results speak.
One belief carries you through. In an era when the rules themselves are changing, the rule-makers win, not the rule-followers. That belief gets you through the three months.
One reframe. Treating AI as an enemy keeps anxiety around forever. "What if AI does this better than me?" I was stuck there too.
Flip it — use AI as a benchmark. Every time you invent a new rule, ask: "Can AI do this? If not, why not?" A rule with a clear reason AI can't do it is a good rule. My seed-based publishing is uncopyable by AI because it requires a time axis. I designed that reason in on purpose.
With AI as the benchmark, the uniqueness of your work starts becoming visible. Not an enemy. A mirror.
One more thing — don't start big. I started with just 10 seeds. One a day for 10 days. That window was for checking whether the rule worked before scaling.
When inventing a new rule, run a 10-day experiment. Drop the old method, try the new rule for 10 days only. At the end, look at the result. If the answer is "there's something here," extend to a month. If the answer is "no," adjust the rule. This short cycle cuts the risk. Rule invention isn't one-shot. It's iterating through versions until you find what fits you.
I've tested 7 rules so far. Three survive. The other four failed the 10-day test and got dropped. The failure was cheap — only 10 days.
AI has taken mastery within the old frame. The era of winning inside the frame is over. One road remains — invent the frame itself.
Try the three questions tonight. What's the basic unit of my work? What order do I take for granted? What material am I not using? Change any of these answers and you've written the first line of your new rule.
Don't create within creation. Create creation itself.
Remember — unit, order, material.