Imagine a chef who just bought a high-end food processor. It doesn’t suddenly make them a Michelin-star cook, but it does mean they aren’t spending three hours dicing shallots by hand. The tool removes the friction of the grunt work, allowing the chef to focus on the actual flavor profile and the plating. For a while, the industry treated AI as the processor—a way to handle the boring bits of asset generation or the initial drafting of a scene. But the conversation is shifting toward something more ambitious, and slightly more delusional.
The latest piece from MIT Tech Review argues that storytelling is baked into our DNA and that AI is simply the next evolution of the medium, following a line from cave paintings to the printing press. The premise is that we are now “scaling creativity.” It’s a nice sentiment. It suggests that by removing the technical barriers to entry, we are expanding the human capacity to tell stories. It frames the LLM as a universal brush, allowing anyone with a prompt to paint a masterpiece.
Here is the problem: “scaling” is a term for server farms, database queries, and GPU clusters, not for art. When we talk about scaling creativity, we aren’t talking about making people more creative in a meaningful way. We are talking about increasing the volume of output. (I suspect most of us are already feeling the fatigue of this volume). The result isn’t a world with more great stories; it’s a world where the cost of producing a “pretty good” story has dropped to zero. We have replaced the scarcity of skill with an abundance of adequacy.
This is where the argument for scaling falls apart. If everyone has a tool that can generate a structurally sound, emotionally resonant narrative arc in six seconds, the value of that arc vanishes. We’ve seen this before with the democratization of photography—everyone has a 4K camera in their pocket now, yet the world is not suddenly filled with more great photographers. It’s just filled with more photos of brunch. The technical ability to capture an image is no longer the bottleneck; the bottleneck is having something worth photographing.
Do we actually want a world where every story is mathematically optimized for engagement? Probably not. The real friction in creativity isn’t the “grunt work” of execution; it’s the struggle to find a perspective that isn’t a statistical average of everything already on the internet. AI, by definition, is a machine for finding the average. It is the ultimate “middle of the road” generator. It doesn’t innovate; it interpolates.
The real danger isn’t that AI replaces the artist, but that it convinces the amateur that they are an artist because they can produce a polished result. It’s like the difference between a hand-drawn storyboard and a generic stock video—one has intent, the other has a budget. (Or in this case, a token limit). We are seeing a collapse of the “craft” phase of creativity, where the struggle to execute actually forces the creator to refine the idea. When you can iterate a thousand times in a minute, you stop thinking deeply about any single iteration.
The tools are great, but the taste is missing.
We are heading toward a saturation point where the “polished” look of AI-generated storytelling becomes a signal for boredom. When the floor is raised, the ceiling becomes the only thing that matters. The developers and creators who survive this won’t be the ones who “scale” their output, but the ones who lean into the weird, the inefficient, and the specifically human errors that a model would try to “fix.” This is the same cycle we saw with the rise of digital synthesizers in the 80s—eventually, people started craving the hiss and pop of analog vinyl again because it felt real.
By Q4, we will see a significant market shift toward “human-certified” content labels as a premium brand signal to escape the AI noise. It won’t be about the quality of the prose, but about the provenance of the thought. We’ll pay a premium for the knowledge that a human actually suffered through the writing process, dealt with the writer’s block, and didn’t just burn through five dollars of API credits to find a plausible ending.
The obsession with scaling is a distraction. You can’t scale a soul, and you certainly can’t scale the kind of insight that comes from living a life that isn’t summarized in a training set.
