George Orwell, AI, and the Soul of Storytelling

Why creative AI tools don’t replace the writer

It’s fascinating that George Orwell, even in 1946, was speculating about the potential for machines to automate storytelling—a notion that feels oddly prophetic today, especially in the age of AI and large language models. Orwell’s idea of machinery-driven fiction, where human creativity is minimized, has interesting parallels to today’s creative tools and raises questions about the role of the writer.

“perhaps some kind of low-grade sensational fiction will survive, produced by a sort of conveyor-belt process that reduces human initiative to the minimum. It would probably not be beyond human ingenuity to write books by machinery...”

Orwell’s “conveyor-belt” fiction, where books become products churned out by machines, does bring to mind how some people view AI-generated content today. We’ve all seen examples where AI creates quick, surface-level content that might feel disposable or formulaic—think of autogenerated news summaries or “clickbait” content. Yet, what Orwell anticipated with his versificator and “novel-writing machines” feels more like a dystopian take on AI, focusing on efficiency over artistry.

In the context of Subtxt, however, this vision takes a different spin. Instead of a mechanical, soulless generation of content, Subtxt helps writers engage in an interactive, thoughtful process. Unlike Orwell’s fictional machines, which produced finished “rubbishy” products without much human input, Subtxt invites the writer to think through their story rather than replacing them in the storytelling. It’s not about pushing a button to get a finished story, but more like a structured dialogue with your narrative to uncover layers and deepen meaning.

There’s also a key difference in how Subtxt respects the writer’s vision. Orwell’s machines remove the “initiative” and personal imprint of the writer. In contrast, Subtxt emphasizes a writer’s agency, using AI and narrative theory to enhance the creative process without taking it over. Where Orwell saw machinery as replacing human effort, Subtxt acts more as a guide, helping authors shape and refine their work in a way that feels organic.

As for large language models (LLMs) like ChatGPT, they certainly don’t produce “jam or bootlaces”—at least not if guided with the right prompts. LLMs aren’t about blindly spitting out content; they can be as sophisticated as the guidance they receive. However, they still lack the genuine creative spark that Orwell’s vision feared losing—the one that comes from a writer’s individual thought process and values.

Ultimately, Orwell’s predictions serve as a cautionary reminder: the real magic of storytelling lies in that unique human ability to infuse a narrative with intent, nuance, and emotional truth. Tools like Subtxt are less about shortcutting this process and more about enhancing it, giving writers a way to “think their way” into their story rather than just passively generating content. Orwell’s machinery-driven writing is a warning of what could be lost, but when approached thoughtfully, AI can actually serve to help writers deepen and clarify their unique visions.