LOG 12: The creative industry isn't dying. It's sorting.

There are four types of creatives right now. Only one of them is building something that lasts.

This isn't about talent. It isn't about which tool you use, how many followers you have, or whether you jumped on AI early. It's about the relationship between what you make and what that making produces after you stop working.

Every creative discipline is asking the same question right now. Designers, art directors, filmmakers, brand strategists, architects. The ground has shifted under all of them in the last two years. Not slowly. Violently. And the question they're all circling is the same: What am I supposed to do now?

The answer isn't in the tools. It's in how you relate to the work itself.

Four stages. Not a hierarchy of talent. A hierarchy of leverage.

The event that made this click

Last week, a group of coders shipped a

small web app.

185 cinematic camera techniques, organized and optimized for AI video generation. Clean interface, copy-paste ready. These are not filmmakers in the traditional sense. They don't work in the creative industry (afaik). They treat generative video as a weekend hobby obsession.

I reposted it. A well-known blogger did so too it. It went viral.

Now consider: I've spent years in the industry, teaching people how to think cinematically, and recently trying to help them expand this into the AI sphere. How to build visual worlds. How to approach a frame like a director, not a prompter. I published 2 90-page codex on visual architecture. I've given talks, written threads, built a methodology.

And a handful of hobbyist developers, in a single weekend, packaged a tool that made cinematic knowledge instantly actionable for thousands of people who will never read my codex. They didn't explain why a Dutch angle works. They just gave people the angle. Named it. Coded it. Shipped it.

Buckminster Fuller called it decades ago:

"You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete."

And somewhere else, even more directly:

"If you want to teach people a new way of thinking, don't bother trying to teach them. Instead, give them a tool, the use of which will lead to new ways of thinking."

That's the shift. Not "who knows more" but "who built something useful with what they know."

The same dynamic is playing out across every creative discipline. Branding agencies are watching AI swallow the production tail of their business. Design studios are seeing clients show up with Midjourney boards instead of briefs. Architects are watching non-architects generate photorealistic renders in seconds. The execution layer, the part that used to be the entire value proposition of being a creative professional, is compressing to near zero.

That's the sorting. And the four stages map directly onto who will end up on which side of it.

Stage 1: Consumer

You use the tools. You try the new model the week it drops. You generate images, render a few clips, follow tutorials, screenshot other people's results. You're learning the interface, getting a feel for what's possible.

There is nothing wrong with this stage. Everyone starts here. Every single person now doing interesting work with AI began by typing a prompt and being amazed that something came back at all. And for designers and creatives coming from traditional backgrounds, this stage can feel electric. (Unless you are an AI hater gooner) You're suddenly able to prototype an idea in minutes that would have taken a team and a budget before.

But the trap is staying here.

Because consumption, by definition, does not produce anything that outlasts the session. You make an image. You share it. Maybe it gets some likes. Then you make another one. And another. Each one exists in isolation. Nothing connects. Nothing builds.

The Consumer's relationship with creative tools, AI or otherwise, is reactive. A new model comes out, you react to it. A new feature drops, you try it. Someone posts an incredible result, you try to reverse-engineer their process. You are always downstream of someone else's decision.

This is true whether you're an AI artist generating images or a graphic designer trying every new Figma plugin or a brand strategist running prompts through ChatGPT. The medium doesn't matter.

The pattern is the same: consumption masquerading as production.

Here is the uncomfortable part: most creatives in this moment are Consumers. Not because they lack ability, but because consumption is the default mode. The platforms are designed for it. The dopamine loop of "input prompt, get output" is real, and it mimics the feeling of making something without producing the compounding effects of actual creative practice.

The Consumer mistakes access to powerful tools for having a practice.

And when the model changes, which it does every quarter (Sora 2, Veo 3.1, Grok Imagine, Kling, and whatever ships next month), the Consumer resets to zero. Because they never built anything on top of the tool. They only used it.

Stage 2: Curator

Something shifts. You stop generating randomly and start developing taste. You recognize what works and why. You build reference collections that actually mean something. You're the person who spots the good stuff before everyone else does. The designer who curates a reference library that other designers steal from. The creative director who can look at a mood board and immediately identify what's missing.

Curation is underrated, especially now. In an era of infinite generation, the ability to select, to point at something and say "this, not that, and here's why," is a genuine skill. The best creative directors have always been Curators at their core. They shape taste. They set standards. They create the context that makes other people's work legible.

But Curation alone has a ceiling.

You're organizing and filtering. You're a signal processor, not a source. The Curator's value depends on the continued existence of things worth curating, which means you are always one step removed from the act of creation itself.

I know many brilliant Curators, both in the AI space and in the broader creative industry. People with extraordinary eyes. People who can look at a brand system and tell you exactly where it breaks, or look at a generated image and identify the precise model and technique that produced it. But when you ask them what they've built, the answer is usually a collection. A gallery. A thread. A Pinterest board with 50,000 saves. Impressive work, important work even, but work that doesn't compound without their continued attention.

There's also a subtler problem. The rise of AI has made Curation seem more valuable than it is, because the volume of generated output is so overwhelming that someone needs to sort through it. But filtering noise is not the same as creating signal. And as the tools get better at generating high-quality output by default, the Curator's role gets squeezed. When everything looks pretty good, the ability to identify what's great becomes less differentiating than the ability to make what's great.

The Curator develops taste. The question is whether taste alone is enough when the tools make execution trivially easy for everyone.

Stage 3: Creator

Now you're producing original work. You have a style. A voice. A recognizable output. People follow you for what you make, not for what you share. You've pushed past the default aesthetic, the generic "AI look," and developed something identifiably yours.

This is where most serious practitioners live, and it's a genuinely good place to be. The Creator has something the Consumer and the Curator don't: a body of work. A portfolio. An identity tied to output rather than consumption.

This is also where most traditional creatives, designers, illustrators, art directors, filmmakers, naturally land when they make the transition into AI-augmented work. They bring craft sensibility. They bring years of understanding composition, color, hierarchy, narrative. And that foundation gives them an immediate advantage over someone who picked up Midjourney last week.

Aristotle made an observation about this level that still holds after 2,400 years: we become builders by building, harp-players by playing the harp. The virtues and the skills are acquired by first performing the acts. There's no shortcut past the doing. The Creator has done the work, and the work has shaped them in return.

But here is what I've learned, both from watching others and from my own experience over two decades: even at this stage, you are still trading time for output. Every piece you make requires you to be in the room. Your best work lives and dies with your direct attention. The moment you stop producing, the value stops flowing.

Think about it this way. A designer creates a stunning brand identity using AI-assisted workflows. It takes two weeks. The client is thrilled. But the next client needs another one. And the one after that, another. Each project exists in isolation, connected to the next only by the fact that the same person did the work. This is the classic freelance trap, and AI doesn't solve it. AI just makes each individual project faster. It doesn't change the fundamental equation.

The Creator's problem is not quality. It's leverage.

The Creator creates value, but the value is linear. One unit of effort produces one unit of output. The question that separates Stage 3 from Stage 4 is: how do you make one unit of effort produce a hundred units of output?

Stage 4: Compounder

This is the shift that changes everything.

The Compounder stops building things and starts building things that build things.

Frameworks. Methodologies. Tools. Products. Systems. Templates that encode your thinking. Processes that carry your taste without requiring your presence. Platforms that turn your knowledge into something other people can apply to their own problems.

The Compounder's work generates value beyond their direct effort. Their knowledge works while they sleep.

Those coders who shipped the cinematic prompting templates? Compounders. They didn't write a treatise about camera theory. They didn't need to explain the history of the Steadicam or the philosophy behind a whip pan. They packaged operational knowledge into a form that is useful without them being present, and it spread exponentially because of that portability.

The

World Building Codex

I spent months creating? That's compounding. Seven hundred plus downloads still generating value months later without me touching it. A single post that gets 10K likes? That's not compounding. That's content. It fades the day after it's posted. Hopefully this library of logs and articles can compound as well.

The same principle applies across the creative industry:

A designer who creates a brand identity is a Creator. A designer who creates a system for generating brand identities, a methodology other designers can license, adapt, or learn from, is a Compounder.

A filmmaker who makes a short film is a Creator. A filmmaker who builds a production framework that lets non-filmmakers produce cinematic content is a Compounder.

A strategist who writes a brilliant brief is a Creator. A strategist who builds a briefing tool that encodes their strategic thinking into a repeatable process is a Compounder.

David Deutsch, the physicist, makes a distinction in The Beginning of Infinity that maps directly onto this. He separates the creation of new knowledge from the reproduction of existing patterns. The Enlightenment didn't happen because people accumulated more information. It happened because they developed a method for generating new explanations. A system that could compound. The knowledge itself became generative.

That's what the Compounder does. They don't just create work. They create the conditions under which better work becomes possible for others.

Why the jump from Creator to Compounder is so hard

Most creatives get stuck between Stage 3 and Stage 4. Not because they lack talent or ambition, but because the jump requires a fundamentally different way of thinking about value.

The Creator asks: *What can I make?*The Compounder asks: What can I make that makes other people better at making things?

That second question is uncomfortable because it requires you to let go of something: the idea that your value lives in the execution. For twenty years, the model was clear. You have a problem, you hire me. My hands, my eye, my taste, applied to your brief. The value lived in the doing.

Every creative discipline was built on this exchange. Graphic design. Architecture. Advertising. Film. Brand consulting. The entire service economy of the creative industry is structured around the premise that the professional's hands are worth paying for.

But that model is inverting. When AI makes it possible for a marketing manager to generate a brand concept deck in an afternoon, or for a startup founder to produce a product video without hiring a production company, or for a client to walk into a meeting with their own AI-generated moodboard, the value of the execution layer drops. Not to zero. But enough to force a reckoning.

The value isn't disappearing. It's migrating. It moves from "I can do this for you" to "I can show you how to think about this." From execution to framework. From service to system.

The old model: I can do this for you.

The new model: I built this for me. Here's what I learned. Now you can use it too.

That's the difference between a freelancer and a brand. Between a service and a product. Between offering your time and offering your knowledge.

The model treadmill, and why fundamentals are the only escape

Here's something that applies to every creative, whether you work in AI, traditional design, or anything in between.

The tools keep changing. If you're in the AI space, this is obvious: every quarter there's a new model, a new interface, a new set of controls. But it's equally true in traditional creative work. The software updates. The platforms shift. The client expectations evolve. The trends cycle. Every year there's a new "must-learn" tool, a new workflow, a new best practice.

The temptation, the very natural temptation, is to chase mastery of each new tool. To become the person who knows the current thing better than anyone else. And that works, for about six months, until the current thing becomes the previous thing.

Mastery of tools is perishable. Always has been.

So what survives?

The fundamentals.

Composition doesn't change when the tool changes. Neither does narrative structure. Or color theory. Or the psychology of visual hierarchy. Or understanding how rhythm and pacing create emotional response. Or knowing when a design needs restraint instead of spectacle. Or understanding how to make disparate elements feel like one coherent thing instead of a collection of separate parts.

I keep experiencing this firsthand. A few weeks ago I was building an AI agent that needed to navigate a long-form video experience, making it feel like one unbroken journey instead of a series of clips stitched together. The technical problem was entirely new. I'd never built anything like it. But the creative problem, how do you create continuity across elements that were produced separately, that's a problem I solved a hundred times in motion design, fifteen years ago. The language was different, but the underlying structure, "how do you make this feel like one thing instead of many things," was identical.

Skills from 2008, solving problems in 2026. Not because the software carried over, but because the thinking did.

This is precisely what makes the Compounder's position so powerful. They build on bedrock. When the model changes, or the software updates, or the trend cycles, the Consumer panics. The Curator scrambles. Even the Creator has to relearn the interface. But the Compounder's frameworks still work, because those frameworks were never built on the specific behaviors of a specific tool. They were built on principles that predate the tool and will outlast it.

This applies whether you're a motion designer, a brand strategist, an architect, or an AI artist. The fundamentals of your discipline are the only asset that compounds across every tool, every platform, every era. Everything else is perishable.

The sorting has already begun

Every few months someone publishes a piece declaring that AI has killed the creative industry. That designers are obsolete. That the entire profession is over. And every few months they're wrong, but not for the reasons most people think.

The creative industry isn't dying. It's sorting.

It's separating the people who consume from the people who compound. And the distance between those two groups is widening fast, because compounding is exponential and consumption is linear.

The people who will thrive are not the ones who master the tool of the moment. They're the ones who turn what they know into something others can build on. Knowledge generators, not content generators.

The tools will keep changing. The landscape will keep shifting. New models will drop and old workflows will break. That is not a crisis. That is the environment. The question is whether you're building on sand or bedrock.

I'll be honest: writing these LOGs has been my own attempt to move from Creator to Compounder. To stop just making things and start building a system of thought that works without me being present for every output. Some of these ideas are still rough. The framework is still being tested against reality. But I keep coming back to the same conviction:

The question is not which tool to learn next.

The question is: what are you building that still works after the next tool drops?

— Ivan