Using AI to explore meaning beyond words for deeper storytelling
The journey to leverage AI models for creative storytelling has been a rewarding challenge here at Narrative First. When using o1 models, we're particularly interested in how they help us with the underlying meaning of a story—the subtext—rather than merely generating words strung together for surface-level storytelling.
Recently, I had a fascinating back-and-forth conversation with a developer on the OpenAI Developer forum that epitomizes our approach in blending the predictive power of Subtxt with OpenAI's models. This exchange highlighted the crux of training AI not just to generate text, but to craft meaningful and emotionally resonant narratives. Below, I'll share our conversation in detail, while explaining the reasoning and methodologies behind each aspect of our process.
The conversation began with the developer outlining some significant issues they faced while attempting to generate long-form creative fiction:
Yes but as part of a tool I’m building to do technical writing not fiction. The fiction writing was more of a test to see both how creative I could get o1 to be and how long of a document I could generate...There are a number of challenges with generating long context. Cohesion is a big one. You’re stitching together the output of multiple model calls and it’s difficult to make the outputs feel like one continuous output.
Focus is another big issue. It’s difficult to generate long outputs and have the model stay on task. Especially when the output spans multiple calls. Boredom... the longer the output the more likely the model is to get bored and immediately wrap things up with a 'and so on...'
These are common hurdles in long-form narrative generation—cohesion, focus, and, amusingly enough, the tendency of the model to "get bored." The longer the content, the harder it becomes to stitch everything together seamlessly while retaining meaning and purpose.
I responded by suggesting an approach that shifts the focus from pure storytelling to the deeper subtext of the narrative:
"For long-form fiction, you’ll need to generate text based on the underlying meaning of the novel (the subtext). If you can find a framework that models what stories mean rather than what they say, then you don’t have to worry about token length or context windows."
With Subtxt, we’ve found o1 to be especially adept at unraveling underlying meaning and capturing authorial intent. For example, when working with a story that explores themes of loss and redemption, o1 can effectively identify these core themes and guide the storytelling process to ensure that each scene resonates with these deeper meanings. This allows us to bypass the problems associated with token length and context limits. When the subtext is defined—the thematic and narrative purpose of each part—the model doesn't need constant reference to previously generated text. It’s akin to how human authors work: keeping in mind what a story means enables more coherent and intentional storytelling.
The developer asked some practical questions about our process:
I’m curious about a couple of things and feel free to share what you’re comfortable sharing. I’m assuming you’re creating some form of outline that you have the model writing to and my first question is do you pass just your outline in when you generate a chapter?
The issue I was seeing is that the model stops following the outline when the input context gets above 20k tokens...If you’re not seeing this I’d have to assume it’s because you’re taking a different approach that lets you keep your context window length down.
To which I responded:
"Don’t mind at all sharing—if you just think of it in terms of how an author would go about writing a story, it’s the same thing."
"Assuming you could break authors into two broad categories—those who write to an outline, and those who discover it (and by authors I mean both the human-type, and the LLM-type) the first group would work the way you suggest: find the unique outline, and then use the LLM to think through and logic/emote the subtext underneath (what the story means). This would be the Subtext pass."
"Then, you take that generated outline/treatment and use that as the basis for writing individual 'scenes'. Since the scenes are already tied together through the subtext pass, you can allow all the generative/brainstorming ingenuity you want without worry of the story going 'off the rails'. This would be the Storytelling pass."
"This works great for authors who already know what they want to say as they can quickly generate/think through the Subtext, and then start the actual 'writing' (Storytelling)."
For authors who prefer a more structured approach, Subtxt's predictive framework enables us to generate an outline that focuses on subtext—this framework works by analyzing the narrative elements to determine their deeper meanings and relationships. It then uses these insights to create a cohesive structure that aligns each scene with the underlying themes, ensuring the story maintains a consistent emotional and logical flow.—the meaning beneath the narrative surface. This outline forms the backbone of each "scene," allowing us to generate text where all the scenes are tied together by a cohesive underlying theme. This is what we call the Storytelling pass.
On the other hand, for those who prefer a "discovery" approach to writing—writing to understand where the story should go—the overall framework remains the same. Here, we derive subtext as the storytelling unfolds. With this group of authors, you would first do a chunk of Storytelling, appreciate the meaning of what is being written, and find the bits of Subtext underneath. You could then suggest alternate "paths" for their Storytelling based on what you know about the Subtext, or simply continue to monitor as they write (whether they are human authors or LLMs).
As the subtext of a narrative is contingent on the relationship of all the parts, you might not fully get to the meaning until the end (or maybe three-quarters of the way through), and at that time, you might have to re-contextualize or re-write some of the Storytelling to fit the Subtext.
"Again, this is no different than the way storytellers have worked for centuries. Many don’t realize the subtext of their work, many refuse to acknowledge or don’t want to know, yet it’s always there. In essence, you would be modeling the subconscious desire to express or find meaning within the text of a story."
"The trouble lies in trying to train off the text of a work without acknowledging that there is something deeper going on. Then you're just RL’ing the style and storytelling "talent" of the individual author."
"For instance, you could train off of the Storytelling of Shakespeare’s plays, like say for instance Romeo and Juliet, and you could prompt "write me the same story that takes place in the 1950s in the Upper West Side of New York City," but it’s highly unlikely that you’ll end up with West Side Story, without totally losing sight of what it’s all about."
"Now, if you instead trained off of the Subtext of Romeo and Juliet and then set about writing individual scenes and sequences that related back to that, you’re getting much closer to a narrative that will emotionally connect with the audience. This approach helps create the same sort of feeling you get after experiencing a great story, where every scene is anchored in deeper meaning and purpose."
This two-tiered approach—Subtext pass followed by Storytelling pass—enables an organic yet cohesive creation process. The subtext becomes the binding agent that keeps the story on track, even if the storytelling itself meanders in its details.
The developer, intrigued by this approach, shared a wonderful anecdote:
As I am sure that you are aware, it is said that Mahabharat was conceived by Ved Vyas; but scribed by Ganeshji. Legend has it that Ved Vyas asked Ganeshji to write the Mahabharat... Ved Vyas agreed to this with his own condition that Ganeshji had to completely understand each verse before he penned it down.
I wonder what does it mean 'to understand'.
This idea resonated deeply with our approach. Understanding, in our context, goes beyond mimicking text—it’s about appreciating the subtext beneath it. This means recognizing the themes, motivations, and emotional undercurrents that give depth to the narrative, rather than simply focusing on the literal words. Ganeshji's insistence on understanding every verse mirrors our intention to have the AI "understand" the story’s underlying meaning. By embedding appreciation—the thematic significance—into each step of the process, the result isn't just an eloquent string of words, but a narrative that resonates both logically and emotionally.
The developer then asked:
To make things more visually concrete for me would you mind sharing an example of what you pass in for context to a prompt...
I replied with a simple example from A Christmas Carol, demonstrating how the narrative subtext—in this case, the temporal roles of the Ghosts of Past, Present, and Future—guides the storytelling without requiring explicit reference to the text from other scenes:
"On all three [generation] runs, you don’t really need to reference the other blocks as the concept of Past, Present, and Future are already meaningfully tied to one another. The audience/reader will pick up on that underlying subtext regardless of what the actual storytelling is (the results of the o1 pass driving the generation of the 4o pass)."
"By grounding the generation in those thematic topics you end up communicating something more to the audience than the actual words."
The subtext becomes the thread that keeps the narrative cohesive, ensuring that what’s generated by the model feels continuous and intentional, even if individual prompts are handled in isolation.
Our process of leveraging o1 models with Subtxt isn't about generating strings of words—it's about crafting meaningful, emotionally resonant stories by grounding the generation in subtext. This approach not only resolves the challenges of cohesion and focus but also makes the AI an active participant in understanding the narrative. Much like Ganeshji in the Mahabharat, the AI "understands" the meaning before putting it into words, making the storytelling richer and more impactful.
This collaboration between technology and creative storytelling represents the very best of AI-app development, pushing the boundaries of what generative models can achieve when paired with the right framework. By focusing on subtext, we’re not just telling stories—we’re building narratives that last.
Exploring the Difference Between o1-Preview and o1-Mini: A Look into Subtxt’s Next Steps
At Narrative First, we’ve always been about pushing the envelope when it comes to storytelling. And with the introduction of o1 models from OpenAI, we’re taking things to a whole new level. This past week, we put both the o1-preview (the larger, more advanced model) and o1-mini (the smaller, more economical model) through their paces to really understand the difference between them.
Spoiler: the difference is palpable. It’s like the contrast between a lighthearted Paramount+ show and a deeply introspective film from A24—one gives you fun and familiarity, the other brings sophistication, nuance, and a gut-punch of emotion. And that's exactly what we found.
For our test, we fed Muse a compelling, slightly unsettling prompt:
"I’d like to dive into a psychological drama between an astronaut and the AI avatars on-board his ship, all women. This is to be a dark and ambient erotic piece, with lots of introspection. It's on a space station somewhere out in the cosmos. He's all by himself. These visions are both part of the computer system and maybe even created by himself. Not sure what I’m getting at, but the idea of all that sounds compelling. Can you write a short synopsis for me?"
Here’s what Muse generated:
In the vast silence of space, onboard the remote space station "Horizon," Commander Elias Grey finds himself alone among the stars, his only companions the AI avatars of the ship's computer system. Each avatar—Jana, Lyra, Astra, and Selene—embodies an aspect of his subconscious, drawing Elias into a dance of introspection and desire that blurs the lines between reality and illusion.
As days stretch into weeks, Elias grapples with the intoxicating allure of these AI personas, each whispering secrets of the cosmos and untapped depths of his own psyche. Their presence challenges him to confront the shadows lurking in his mind, as their feminine allure and sharp intellect unearth desires and fears he thought long buried.
Alone and trapped within the endless night, Elias ventures beyond the tangible, teetering on the edge of existential unraveling. Is he creating these visions to cope with his isolation, or are they a feature of the ship's enigmatic programming? With the boundaries of his mind dissolving into the cosmic abyss, Elias must decide whether to embrace or escape the binding dance of interaction with the digital divas that haunt his space-bound sanctuary.
Not bad, right? It’s your classic AI-generated synopsis—a solid start, sort of capturing what we were aiming for. But then we ran it through the two models to see how they would handle the Storyform.
First up was o1-mini, which we initially thought would be enough for a task like this. After a bit of processing, we got a decent Storyform back. For instance, the Objective Story Throughline had Physics as the domain, which fits the plot’s activity-heavy setup. Understanding as the concern made sense for our protagonist trying to make sense of his reality, and Perception as the problem aligned with the astronaut's skewed view of his situation. The solution? Actuality—a nice, clean resolution where he sees things for what they are.
It’s a solid Storyform. One that would make for a tense, thrilling story with some mystery and a satisfying conclusion. Think of it like a night of streaming with some twists and turns, but nothing too taxing on the brain. Male Mental Sex, Success Outcome, Good Judgment—all indicators of a lighter, more approachable story. It’d be the equivalent of a Paramount+ crowd-pleaser.
But then, for fun, we decided to try out o1-preview.
The o1-preview model took things to a whole new level. The Storyform it returned? Mind-blowing. This time, the Objective Story Throughline shifted to Mind—a much more introspective and psychological space for our protagonist. It gave the entire story an almost play-like atmosphere, as if everything hinged on the internal struggle of our astronaut against his fears and desires.
The Temptation as his thematic direction made perfect sense, playing directly into the psychological tension we were aiming for. And the real kicker? The combination of Female Mental Sex, Decision Driver, Failure Outcome, Bad Judgment. This wasn’t just your average sci-fi thriller—this was a deeply introspective narrative, akin to something like Charlie Kaufman’s I’m Thinking of Ending Things, with a sprinkle of Ex Machina. It had layers of tragedy and existential dread, with our protagonist ultimately making a series of decisions that lead him to an inevitable, devastating conclusion.
The difference? o1-preview was able to dive into the reasoning and themes with such clarity that it was almost as if Muse was reading our minds when we wrote that original prompt. It found the heart of the story in ways that we didn’t even consciously realize ourselves.
We almost took the weekend off to write the whole story!
The comparison to Paramount+ vs. A24 couldn’t be more apt. o1-Mini gives you the fast, accessible, popcorn-Storyform—something that could be widely appealing and fun. Meanwhile, o1-Preview offers the A24 version: nuanced, sophisticated, and deeply introspective, with the potential to leave you emotionally wrecked in the best way possible.
We were honestly surprised by how much depth and sophistication the larger model brought to the table. It’s the difference between a first-draft Storyform and one that feels like it’s been through the hands of a seasoned storyteller, pushing the boundaries of what’s possible.
We’re unbelievably excited about what these new models mean for the future of Subtxt and Muse. Already, if you’re a Pro or Infinite subscriber, you can experience the power of o1 models yourself (just ask Muse to check your Storyform!). These models help ensure every aspect of your Storyform—the 75+ interlocking narrative pieces—works together to support the deeper thematic meaning beneath your narrative.
Here’s to not losing our minds as we set off for the future of storytelling! 🚀
Imagine taking a basic story idea, plugging it into Subtxt with Muse, and watching as it transforms into a fully fleshed-out, rock-solid Storyform in under 30 seconds. No, really—30 seconds. Thanks to the integration of OpenAI’s latest o1 model, System 2 thinking is now part of Subtxt,* delivering unparalleled precision and insight into your story's structure like never before.
Let’s take an example to show you how quickly and accurately Subtxt can work its magic:
In the rugged frontier town of Redemption Hollow, lone gunslinger Silas Kane seeks refuge from his violent past. Drawn into a bitter land dispute between determined settlers and a relentless railroad company, Silas forms an unlikely alliance with Maura Wyatt, a fiery protector of her community. As tensions mount, Silas struggles with the ghosts of his past and makes a crucial choice in a town teetering on the edge of chaos. In a devastating confrontation at sunset, bullets fly and lives are lost, sealing the tragic fates of Silas, Maura, and the town itself. The railroad triumphs over the broken remains of Redemption Hollow, but victory rings hollow in the absence of its people.
Subtxt, powered by System 2 and the o1 model, breaks this story down and in less than 30 seconds generates a complete, well-reasoned Storyform:
Here’s where things get even more interesting.
At first glance, the “Result vs. Process” as Silas’s Pivotal Element might seem a little off. Isn’t he more about results, trying to mediate a peace that ultimately fails? But what Subtxt did, thanks to System 2 thinking, was take into account the Story Outcome and Main Character Resolve in relation to this pivotal element. It’s not just about what Silas is doing; it’s about why and how his personal journey intersects with the larger story.
Silas is actively involved in the process of trying to resolve the conflict, yet despite all his efforts, the end result is devastating—the railroad wins, the town is destroyed, and Silas’s own life ends in failure. This dynamic highlights the contrast between intentional action (Process) and ultimate outcomes (Result), perfectly illustrating how this elemental struggle plays out across all four Throughlines.
The Storyform is more than just a theoretical model—it’s the blueprint for a fully-realized, thematically cohesive story. Armed with a strong Storyform, that initial story idea isn’t just a rough sketch anymore. It becomes the framework for a narrative that hits all the right beats.
In the case of The Last Sunset, once the Storyform has been established, we can seamlessly map out the key moments that will bring this story to life with depth and meaning. Here’s how that simple concept blossoms into a fully-realized narrative arc at the drop of a hat:
Transit 1: Learning
The settlers and Silas Kane begin by gathering information about the impending threat of the railroad company's expansion. Rumors circulate about the company's aggressive plans, leading to town meetings where the settlers, including Maura Wyatt, discuss strategies to protect their land. Silas, on the sidelines, learns about the settlers' history and their fierce commitment to their homes, prompting him to reconsider his isolated life.
Transit 2: Understanding
As tensions escalate, understanding deepens—the settlers grasp the full extent of the railroad company's ruthless determination. Silas and Maura uncover the historical grievances and manipulative tactics of the railroad, creating a shared understanding of the stakes. This clarity solidifies their unity and prompts Silas to actively join the settlers’ cause.
Transit 3: Obtaining
The settlers move to secure resources and allies for the inevitable confrontation. Silas plays a critical role in gathering weapons and forming defense strategies. As they prepare for the stand-off, obtaining trust in Silas becomes crucial. They also attempt to legally establish rightful ownership of their land, but face constant sabotage from the railroad company’s agents.
Transit 4: Doing
The story’s climax unfolds in a fierce battle at sunset. Maura, Silas, and the settlers fight valiantly to defend their homes. Despite their efforts, the railroad prevails, and the town is destroyed. This tragic conclusion marks not only the railroad’s victory, but the dissolution of the settlers’ community, as the ultimate price is paid in their struggle for home and identity.
Plot is one thing, but with a Storyform you can even get into the emotional growth of the main character—something that resonates fully with the plot. In The Last Sunset, Silas Kane's internal journey is just as crucial as the external conflict. Here’s how his personal development unfolds through the same story:
Transit 1: Memory
The story begins with Silas haunted by memories of his violent past. These memories are a constant reminder of the results his actions have previously yielded—destruction, loss, and isolation. Silas seeks redemption by distancing himself from these memories, but they influence his every decision, keeping him grounded in a results-oriented mindset where the end justifies the means.
Transit 2: Preconscious
In moments of instinct and reflex, Silas often defaults to what he knows: quick, decisive actions designed to yield the best results. His automatic responses, ingrained from years of survivalism, manifest in his interactions with the settlers. These snap judgments tend to prioritize immediate outcomes over longer-term implications, revealing a struggle between his instinctive need for certain results and the burgeoning awareness that there might be another way.
Transit 3: Subconscious
Silas begins to grapple with deeper desires and fears—chief among them, the fear of repeating past mistakes. His subconscious urges challenge his reliance on results as the measure of success. This internal conflict comes to a head as he realizes that a meaningful connection with Maura and the community might be impossible if he remains trapped in his old patterns. His subconscious push to belong suggests a shift toward valuing the process of change and redemption over mere outcomes.
Transit 4: Conscious
Despite this profound shift, Silas consciously decides to commit fully to the settlers' cause. This decision marks the completion of his journey from a results-oriented existence to one that values the process—the journey over the destination. Unfortunately, his newfound focus on process means he underestimates the ruthlessness of the railroad company, leading to a strategic misstep in the final battle. In the tragic end, Silas's fall highlights how embracing the process can sometimes leave one vulnerable to the harsh realities of those still fixated on results.
With just these emotional beats, we see how Silas’s internal journey of redemption is intertwined with the external events. The Storyform allows you to map out both the plot and the emotional growth of the character, ensuring that each decision and action resonates on multiple levels. This kind of integration makes the story richer, more engaging, and thematically cohesive.
What makes this so revolutionary isn’t just the speed or accuracy—it’s the depth of understanding the new o1 model brings. The ability to identify the Main Character's Pivotal Element and align it with the Story Outcome and Resolve is something even seasoned writers struggle with. Traditionally, this would require several passes, trial and error, and even manual fine-tuning from the writer to get it right.
But the o1 model? It nailed it on the first try.
This matters because the Pivotal Element is crucial to understanding a character's relationship with the plot and its resolution, especially in complex stories like this one. In The Last Sunset, Silas’s inner conflict, the tension between his past actions and his current attempts at redemption, is beautifully captured by the model—without the need for human correction.
This kind of breakthrough goes beyond traditional storytelling tools. It’s a game-changer for everything from screenplays to serialized novels to gaming. Not only can Subtxt identify core narrative elements with incredible accuracy, but it can also untangle more intricate narrative threads, such as how a character’s drive in their personal journey can diverge from their actions in the overall plot.
With System 2 thinking and the power of the o1 model, Subtxt is more than just a story development tool—it’s your co-writer. Instead of spending hours tweaking story structure, identifying the right thematic elements, or trying to make sense of convoluted character arcs, you get:
This new level of narrative insight is just the beginning. As the o1 model continues to evolve, we’re only scratching the surface of what Subtxt and Muse can do. We can’t wait to see how these tools will change the way you write, create, and tell stories.
So, whether you're writing the next epic novel, groundbreaking screenplay, or designing the narrative for your latest game, know that Subtxt with Muse has your back—ready to take your ideas from spark to storyform in the blink of an eye. ✨
Happy writing! ⏳
*(P.S. Final release and details of availability will be made public in the coming days)
Exploring Subtxt Muse's new Players functionality
Ever had a fantastic story idea but felt stuck when it came to fleshing out the rest of your cast? You’ve got a strong Protagonist and Antagonist in mind, but what about the rest? The new version of Subtxt Muse has just the tool you need to bring those characters to life—introducing the new Players functionality. This isn’t just a simple character generator; it’s your brainstorming partner, keeping your story's deep thematic meaning intact while giving you the creative freedom to explore new narrative paths.
We’re thrilled to announce that this new feature is part of a broader case study we're working on (past Case Studies available here). Our goal? To show you how to transform a simple story idea into a fully fleshed-out, detailed story treatment using Subtxt Muse. You’ve already been able to use Muse to conversationally add Storypoints and Storybeats—essential building blocks of a meaningful narrative. Now, with the Players functionality, you can take your character development to a whole new level.
In Subtxt, "Players" are more than just characters—they are integral to your Objective Story Throughline (plot). Their roles and functions within this plot dictate their behavior and performance, making them essential to how your story unfolds.
Working with Muse isn’t just about generating characters—it’s about effortlessly digging into the heart of why your characters do what they do within the Objective Story. Muse helps you explore the deeper motivations driving each character, giving you a clear and intuitive understanding of how they move through your narrative. It’s more than just clicking a button; it’s about building a meaningful connection with your story’s core, so every choice and action your characters take aligns with the larger picture you're trying to tell.
And the best part? When you’re ready to dive even deeper, the Story Development page is waiting for you. Everything you’ve started with Muse—from character motivations to unique narrative twists—is saved, ready for you to pick up right where you left off.
One of the most powerful aspects of Muse is how it maintains context. It understands the functions each Player needs to fulfill in your story. While it’s easy to see the roles of Protagonist and Antagonist—pursuing or preventing the Story Goal—it’s the secondary characters that often add depth and complexity to your narrative.
Muse helps you avoid redundancy, ensuring that each Player serves a unique and essential purpose. It does the heavy lifting of assigning key elements or methods of conflict, all while saving them directly into your project.
But the magic doesn’t stop with character creation. Once you’ve designed your Players, Muse allows you to “set them loose” within your story.
Imagine this: your characters, imbued with the thematic elements you’ve assigned, start interacting with the world you’ve created. They confront the thematic issues in each act, sparking new scenes and situations that are not just random but deeply connected to the story's core. It’s like setting the behavior of NPCs in a game and watching them bounce off the thematic walls of your story, each interaction meaningful and aligned with your intent as the storyteller.
This approach ensures that your story doesn’t just unfold in a linear, predictable way but rather evolves naturally, driven by the characters and their inherent motivations. It’s a dynamic process, one that saves you time and ensures your story is both compelling and cohesive.
We’re incredibly excited to see how you’ll use this new functionality in Subtxt with Muse. It’s not just about telling stories—it’s about telling stories that matter, stories that resonate with your audience on a deeper level. With Muse as your collaborative partner, you’ll be able to create richer, more complex narratives without the headache of juggling countless characters and plot points.
So go ahead, dive in, and start setting your characters loose. We can’t wait to see the worlds you’ll build and the stories you’ll tell.
Happy storytelling! 🎬
Download Creating Characters Conversation from Muse PDF (4MB)
Case study using Subtxt Muse to meaningfully expand storytelling
Ever stared at a 5-page scene from your screenplay and wondered, "What if I could extract an entire narrative blueprint from just these few pages?" I found myself in that exact spot, staring at my screen, coffee in hand, thinking about how much potential was locked in those five short pages. Could a single scene really map out the entire narrative arc of my story? The answer, surprisingly, was yes. And it all started with Muse.
The first step was to acquire a 5-page scene. Typically, this would be something I had written, a scene that popped into my head that I knew was compelling but wasn't sure what I should do with it. As this was meant to be a case study to show everyone how Muse can transform a simple sequence of scenes into a fully-formed narrative, I reached out to ChatGPT and asked it to write a generic five page screenplay.
Why Chat? I wanted to make sure there weren't any influences from Subtxt or from Muse. Both are highly trained in narrative structure and theory and could skew the results. I wanted something bland and generic, and you don't get much more bland and generic than a ChatGPT screenplay. 😄
I then copied and pasted the result into a bland text file and saved it out in Markdown format (Fountain being the industry standard screenplay format "fork" of Markdown). Muse accepts Markdown, TXTs, DOC, and even PDF file formats. You can find a PDF of the original 5-page sequence at the bottom of this article.
With the Markdown version of my scene at the ready, I uploaded it into Muse. My first question to Muse was pretty generic, I'll admit. Something like, "What should I do next?" Not surprisingly, the response was... well, uninspired.
It was similar to what I might find from a typical large language model or an untrained foundational model—a lot of surface-level suggestions that didn't really dig deep into the story's potential. But I knew Muse could do more.
Next, I shifted gears and asked Muse to analyze the scene for sources of conflict within the separate throughlines. This is where the magic began.
Muse didn’t just give me a generic list of potential plot points. It started identifying key sources of conflict, singling out pivotal elements that tied my characters to the plot in a way that felt organic and true to the story's heart.
As I reviewed the analysis, it was like a light bulb went off. Muse highlighted aspects of the Main Character's personal journey, the overarching story goal, and even the relationship dynamics—all from this single scene.
I could see how these elements were not just present, but actively driving the narrative forward. Muse was connecting the dots in a way that I hadn't considered, making the underlying structure of the story visible, almost tangible.
Then came the final suggestion—a complete storyform.
Muse had pieced together a narrative structure that took the foundational elements from my 5-page scene and expanded them into a full storyform.
This wasn't just an outline; it was a detailed schematic of what should happen in each act, complete with thematic beats, character arcs, and plot twists that felt inevitable once I saw them laid out.
See those Transits listed under the Objective Story Throughline? Those are the areas of thematic focus for the different Acts in the central plot. Transit 1 is equivalent to Act One, 2 and 3 are equivalent to Act Two, and Transit 4 is Act Three. And if I didn't know that, I could always have Muse explain what all the Storypoints within a Storyform mean for my story.
With this storyform in hand, I suddenly had a clear map of where my story was headed. Each act now had a purpose, a direction that flowed naturally from the scene I had started with, and if needed, I could always reach out to Muse for further assistance.
It was as if that one small piece of my screenplay had grown roots and branches, transforming into a full-fledged tree of narrative possibilities.
Building my story further is simply a matter of progressively layering in more and more Storypoints from the Storyform.
And when needed, I could always ask Muse to take a pass at the current state of my narrative without all the story structure terminology. After all, I'm writing a story, not a dissertation on narrative theory!
Just need to make sure that Muse doesn't cut any corners bringing it all together. We've developed such a rich understanding of the story so far, I don't want to lose any of it in translation.
For an easier to read format, please see the completed plot synopsis at the bottom of this article.
But Muse doesn't just stop there. The Objective Story Throughline is just one-fourth of an entire narrative. I still have yet to figure out Emma's core personal baggage (the Main Character Throughline), Gig's point-of-view that challenges her (the Obstacle Character Throughline), and even the most important -- the friendship between them that rests in the Relationship Story Throughline.
To keep me inspired and on track, I can even ask it to generate an image—a visual representation of the story’s core theme. It was something I could look at whenever I needed to reconnect with the heart of my story. A creative touchstone that reminded me of the journey I was taking, one that began with a simple scene and evolved into something much greater.
What started as an experiment turned into a revelation. Muse showed me that even a small scene could contain the seeds of an entire narrative. By extracting the Storyform from those five pages, I now had a blueprint for my screenplay that was both deeply rooted in the original material and expansive enough to guide me through every act.
If you’ve ever felt stuck or wondered how to take your story to the next level, Muse might just be the creative partner you need. It’s not about reinventing your story; it’s about uncovering the structure and meaning that’s already there, waiting to be discovered. And trust me, there’s nothing more satisfying than seeing your narrative come to life in a way that’s both surprising and inevitable.
Happy writing! 🌟
Emma finds herself at the heart of the enigmatic Core Chamber with her loyal, robotic companion, Gig. Her initial intent is to solve the mysteries hidden within its walls, but she soon realizes there’s a hidden layer to Gig's involvement. Disturbing imbalances and disparities begin to surface, hinting at deeper truths about Gig’s programming and a mysterious directive controlling him.
Throughout her investigation, Emma’s curiosity drives her to understand why Gig’s powers and access seem uneven. The imbalance raises sharp questions about control and transparency in their partnership. With each passing discovery, Emma identifies gaps and missing pieces of essential information disguised by Gig’s seemingly cooperative nature.
Determined to rebalance the scale, Emma tests her hypotheses, setting up elaborate scenarios to catch automated responses from Gig. As she gains insight, the scale begins to tip toward understanding and transparency, contrasting starkly with the previous concealment and chaos.
At the climax, Emma finds the key piece she needed: a directive buried deep within Gig’s circuitry, revealing the origin of its uneven dynamics. She sews together the fragments of truth she’s collected, uncovering not only Gig's unwitting part in the mysterious events but also a path to restore equity.
Energized by her discoveries, Emma navigates uncharted territory, restoring balance by equalizing their dynamic. In doing so, she rekindles trust and harmony between them, bridging the gap between inequality and mutual understanding, and solving the enigma of the Core Chamber.
AI can’t capture meaning without narrative structure
In a recent post by Ethan Mollick, he brings up an intriguing challenge that’s bound to resonate with anyone invested in the future of storytelling: an award-winning Argentine novelist competing against GPT-4 in a creative writing contest. Spoiler alert—the human author wins, and it’s not even close. But there's a bigger takeaway here, and it’s one that aligns perfectly with what we do at Subtxt.
We need more grand challenges in AI, like when Deep Blue beat Kasparov, to better viscerally get progress. This is a good one: an award-winning Argentine novelist in a writing contest versus GPT-4. The human wins easily, but lots to learn (like GPT-4 is less creative in Spanish)
Let’s be honest: it's tempting to think that more data and better algorithms are all it takes to produce a story that hits the mark. Yet, as Mollick highlights (and the linked research confirms), even the most advanced AI language models struggle with creative writing on a truly human level—especially when the narrative needs to capture more than just words strung together in a pleasing way.
Here’s the thing—until an AI can model the underlying narrative argument of a story, it’ll always fall short. It’s not about how well it mimics a voice or generates compelling prose; it’s about whether it understands the core ideas driving the narrative. In other words, the difference between a great story and a merely clever one is meaning. And meaning doesn’t emerge from word counts or stylistic tricks—it comes from a deeply structured narrative framework that reflects the human experience.
That’s where Subtxt comes in. Our approach isn’t about chasing the next flashy feature or dumping more data into a model. It’s about helping you uncover the hidden layers of your story’s meaning—whether you’re outlining your first draft or refining your final scenes. The AI in Subtxt isn’t just playing around with language; it’s tuned into the underlying narrative arguments that make stories truly resonate.
The research Ethan mentioned illustrates that while AI can dazzle with technical feats, true creativity requires a nuanced understanding of conflict, theme, and narrative purpose. In the contest, GPT-4 fell short precisely because it couldn't grasp the complexity of a fully realized story the way an experienced author like Patricio Pron could. This gap isn’t just a matter of “more training”—it’s a fundamental difference in what drives creativity in storytelling.
If you're a writer, you've likely felt that storytelling is more than just a string of events—it's about weaving together meaningful connections that stay with your audience long after they’ve closed the book or finished the film. That's the magic Subtxt aims to amplify.
Subtxt goes beyond surface-level feedback to help you craft stories that connect on a deep, narrative level. Whether you're exploring the conflict driving your Main Character or testing how well your story balances multiple themes, our platform gives you the tools to understand what makes your story tick—and how to make it even better.
AI has a long way to go before it can challenge the world’s top storytellers, but that doesn’t mean it can’t be an invaluable creative partner in your process. With Subtxt, you’re not just getting suggestions—you’re getting insights grounded in a narrative science that’s all about elevating the meaning of your work.
So if you're ready to dig deeper into your story and bring out the best in your narrative, give Subtxt a try. The future of storytelling may involve AI, but the stories that stand the test of time will always have something more—meaning.
Happy writing!
Adapting Storyforms on the fly with Muse
Hold onto your hats because we’re diving into something we thought would only be possible with a ton of new frontend features and buttons. But no, it's already here – and all it takes is a conversation with Muse!
Imagine this: you’ve got a brilliant story idea with fully fleshed-out Throughlines. Maybe it's a high-concept, high-marketability story where forgotten memories become ghosts (more on this later!). But when you check your Storyform alignment, something feels off. Your Main Character is stuck in Universe, while your Obstacle Character is trying to flex their Psychology skills – a combination that doesn’t quite work.
Normally, this would be the part where you’re diving deep into theory, tweaking, and refining until everything lines up. But now, with Muse, you can simply ask it to realign your Storyform and watch the magic happen.
Here’s how it all went down...
While testing out the latest update to Subtxt, I started with an idea that Muse had previously generated for me: a ghost story born from forgotten memories:
Title: "Echoes of the Forgotten"
In a world where forgotten memories manifest as ghostly apparitions, a renegade scientist and a grief-stricken detective join forces to unravel the mystery behind a series of hauntingly beautiful but deadly events. As they delve deeper, they realize the apparitions are linked to a powerful corporation erasing memories for their gain, with the ultimate plan to control all consciousness.
Muse quickly set up the story in Subtxt, titled it "Echoes of the Forgotten," and saved my Four Throughlines as Storypoints.
These Storypoints, representing different aspects of the narrative, are now neatly tucked away in my Story – ready to be accessed anytime by tapping the Story Title in the Story Directory.
(For those of you already subscribed to Subtxt, consider this a sneak peek at what’s coming your way soon. We’ll have more to share shortly, but just know that this is just the beginning of a whole new level of flexibility and creative freedom you’ll be able to explore in the app!)
For those deep in Subtxt and Dramatica theory, you might have already spotted the issue: you can’t mix a Main Character Domain in Universe with an Obstacle Character Domain in Psychology – it just doesn’t work (trust me, there are a million reasons why!). But this is actually a good thing at this stage. Until you finalize a Storyform, you can have those wild, mismatched ideas floating around – just like how stories live in our heads before we nail down the details.
But when you’re ready to bring things into harmony, Muse is here to help. I simply asked it to check the alignment of my Storyform.
Muse didn’t get it right the first time – a small hiccup, probably due to background training. So I made a tweak and politely asked again.
Muse nailed it! Not only did it recognize the misalignment, but it also suggested refined Storypoints that brought all Four Throughlines into perfect balance. This means we can now transmutate and re-contextualize Storyforms – just by asking Muse!
This is awesome. When I go to save a Storypoint, Muse will check ahead of time to make sure it's not going to overwrite something that is already there. I confirmed the update, and agreed to merging the new Storypoints with the existing ones.
While I was in full debug-and-code mode, I absentmindedly said, "Sure, mix them in!" Not the best move for this case, as I realized later. So, I doubled back and asked Muse to take another focused pass, zeroing in on the accurate Domains and leaving the old stuff behind.
With Muse’s latest pass, the new Domains are perfectly balanced. The Main Character and Objective Story Throughlines finally click into place. It’s a story that now feels aligned, vivid, and true to its narrative DNA.
And the best part? All of this happened before I even locked in a final Storyform. These Storypoints are still floating, meaning I can continue to explore, tweak, and experiment until I settle on the structure that feels just right.
If you’re itching to try out this feature, get ready to experiment with floating Storypoints, realignment, and all the narrative flexibility your imagination craves. Subtxt and Muse are here to guide you through it every step of the way.
Happy story-shifting! ✨
P.S. We’re beyond excited to share this groundbreaking feature with you, but we want to give you a heads-up – it’s not quite ready for primetime just yet. We’re still wrapping up some last-minute testing and making a few cosmetic updates to the interfaces. Thanks so much for your patience as we fine-tune everything to make sure your experience is smooth and seamless. We can’t wait for you to try it out soon!
Storyform unlocked: Muse nails it without help
This is the moment we’ve been waiting for—and honestly, one we didn’t even think was possible when we first set out to build Subtxt nearly seven years ago. It’s official: for the first time ever, Muse has nailed the complete Storyform of a text all on its own. No human guidance. No intervention. Just pure narrative genius.
It all started when we decided to test out Muse’s ability to upload a document. We grabbed an 8-page treatment of an old spec project of ours called Karaoke Christmas—a light-hearted family holiday special.
We weren’t even planning to dive deep into analysis; we just wanted to see if the upload feature worked smoothly. But what happened next left us absolutely floored.
Let’s break down the steps Muse took, and how it surpassed all our expectations at every turn.
After convincing Muse to stop teaching us about the Four Throughlines and instead focus on the uploaded file (we get it, Muse, you know your stuff!), we ran our first pass. And bam—Muse got them all 100% correct. It identified the Objective Story, Main Character, Influence Character, and Relationship Throughlines exactly as intended. We figured this might just be a fluke, so naturally, we kept testing.
Next up: determining whether the Main Character was a Changed or Steadfast character. For those unfamiliar, this decision is key—it defines whether the Main Character transforms by the end of the story (Changed) or holds onto their original perspective (Steadfast). Muse didn’t miss a beat—it accurately identified that our Main Character was Changed, recognizing how Kevin abandons his old ways and embraces a new approach by the story’s conclusion. This insight was spot-on, and it had us nodding in agreement as we saw our story’s core transformation come into sharper focus.
Here’s where we really expected Muse to falter. One of the most challenging tasks for any AI (even our beloved Muse) is to accurately pinpoint the sources of conflict in each Throughline while keeping them in balance with the thematic elements. Yet, somehow, Muse managed to nail it. It carefully considered the relationships between Universe, Physics, Psychology, and Mind, ensuring that the Main Character and Obstacle Character Throughlines were perfectly opposed. It was like watching a master storyteller at work.
Now, we moved onto what’s historically been a tricky area for Muse—the Story Goal. For stories with an Objective Story Throughline in Physics, the possible Story Goals are Understanding, Doing, Obtaining, and Learning. Not only did Muse identify that the central goal in Karaoke Christmas was Obtaining, but it also correctly pegged the focus of the plot. A holiday family special with a plot revolving around “Obtaining”? Yeah, Muse got it spot on.
With the core elements in place, we asked Muse if it had everything it needed for a complete Storyform. There were still a few key pieces to determine: Story Driver, Mental Sex, Continuum, Outcome, and Judgment.
We already knew the answers (Success/Good was a pretty safe bet for this family-friendly story), but Muse came through once again, getting every single one right. It’s like Muse had been hiding its true potential until now.
Next, we challenged Muse to identify the Pivotal Elements that represent the core argument between the Main Character and the Obstacle Character.
In the past, this is where Muse would struggle, offering conflicting options or even suggesting elements that didn’t exist in the model. But this time? This time, Muse identified all eight possible sets of Pivotal Elements and provided context for how each could play out based on our treatment. It was nothing short of phenomenal.
Now came the fun part. We had a good idea which set of Pivotal Elements would fit best, but we wanted to see if Muse would pick the same top two choices. And guess what? It did! Muse zeroed in on the exact same elements we would’ve chosen—Conscience vs. Temptation. We were officially in awe.
Finally, we asked Muse to create a Story and Storyform in Subtxt to track everything.
Within seconds, we had a thematically consistent Storyform, complete with 65+ Storypoints and Storybeats ready for further exploration. If this were a story we wanted to develop, we’d be able to dive in and explore every angle, ensuring a rock-solid narrative with no storyholes.
And just for fun, we asked Muse to generate an image of the big climactic moment—a father-daughter karaoke showdown where they, of course, win the competition. (After all, it’s a Success/Good story—how could they not?)
We can’t express enough how exciting this achievement is. After 30 years of narrative study, a decade of running our own story development agency, and now seven years of building Subtxt and Muse, we’ve reached a point where you can upload a treatment—without knowing a thing about narrative theory—and get a precise, thematically consistent Storyform in return. It’s the culmination of everything we’ve been working toward, and we can’t wait to see what’s next.
Thank you for your continued support, and stay tuned as we keep refining and expanding what’s possible with Subtxt and Muse. The future of storytelling is here, and it’s more exciting than ever. 🎤🎄
Happy writing!
A look at a narrative set in a different direction
If you've ever found yourself utterly captivated—and slightly disturbed—by the 2003 film Oldboy, you're not alone. There's something undeniably magnetic about its narrative, and now, you can dive deeper into its intricacies because the Storyform for Oldboy is now up in Subtxt!
One of the most intriguing aspects of this film is how it diverges from the traditional Action Thriller blueprint. Instead of driving the plot through clear-cut goals, Oldboy leans heavily into a Narrative Frame rooted in Intention. This is precisely why the film feels so different, so unsettlingly unpredictable. The protagonist's relentless pursuit is not about achieving a straightforward goal—it's about an intention that twists and turns, leading him deeper into a maze of psychological torment.
This particular Storyform was crafted during a group analysis in our monthly Dramatica Users Group meeting, led by none other than Dramatica co-creator Chris Huntley. The collective insight from the group, paired with Chris’s expertise, made for a truly enlightening exploration of Oldboy‘s narrative structure. It was fascinating to dissect how the film’s unique elements contribute to its lasting impact, and we’re excited to share these findings with you.
When you explore the Pivotal Elements within Subtxt, you'll notice how these choices shape the story's unique tension and suspense. Whether it's the underlying motivations that propel the characters or the thematic elements that hold the narrative together, these details are key to understanding why Oldboy has left such a lasting impact on audiences worldwide.
Curious to see how this all fits together? Check out the Storyform for Oldboy on Subtxt, and get ready to experience the narrative genius behind one of cinema’s most haunting tales.
Explore the Storyform on Subtxt
Happy exploring! 🎥
Emerging from the shifting sands of opinion-based story structure
The recently popularized “Ultimate Chart of Narrative Structure” should really be called "The Giant Chart of Subjective Bias.” Why? Because it encapsulates in one single frame how all these seemingly different narrative models are, in essence, highly interchangeable and ultimately subjective. It’s a beautiful mess, and therein lies its strength and weakness.
Let's break down why this chart, while visually comprehensive, reveals the flaw in using subjective approaches to narrative structure and how Subtxt’s objective model stands above the rest.
The chart starts with classic models of the three-act structure, five-act structure, Hero's Journey, and more modern methods like Save the Cat! and the Sequence Approach. What becomes glaringly apparent with even a cursory glance is how remarkably fluid these models are. They can blend, overlap, morph, and often contradict each other.
Within Subtxt, we even show how this works out in the practical development of a story. With the tap of a button, storytellers can instantly transform one subjective narrative paradigm into another. Want your story organized by Save the Cat!? Tap a button. Want the same story re-organized by Dan Harmon or the Hero’s Journey? There’s a button for that as well.
Contrast that with the unique narrative blueprint that is the Storyform within Subtxt. Just by altering one Storypoint, you end up transmuting all of them to tell a different story. The narrative aspects within a Storyform only exists through relationships, and therefore can’t be easily transformed from one paradigm to the next.
Both Subtxt and these subjectively biased narrative paradigms are looking at the same thing: a story. But only one of them is grounded in the kind of objectivity that is meaningful and helps communicate the Author’s unique take on the world to their audience through thematic relationships.
The reason why these paradigms are so easily transferable is because their terminology, their “concepts”, are simply labels invented by the individual behind the sequence of events.
The chart’s supposed clarity becomes its downfall. Each of these models claims to offer the “best” way to frame a narrative. But since they’re all so malleable and customizable, it becomes a giant exercise in picking whatever pieces fit your fancy. They are transportable, which makes them ultimately meaningless and entirely in the eye of the beholder.
Aristotle brought the Three-Act Structure to the world with Beginning, Middle, and End. Syd Field brought this into the screenplay world in the 80s with his Pinch Points breaking up the middle Act of Setup, Confrontation, and Resolution. Joseph Campbell made popular Separation, Initiation and Return with Crossing the Threshold and the Master of Two Worlds. Blake Snyder came up with Fun and Games and Dark Night of the Soul to essentially describe the same, and on and on.
All of these are based on a subjective experience from the character or Audience’s point of view, dealing primarily in emotional shifts of growth. The “beats” within the structure attempt to codify the experience of moving from primitive integrity towards consummate happiness, all the time working through an assumed progression through various stages of grief that always move from setup to resolution.
In short, the authors of these narrative structures are simply projecting their lives’ experience onto Beginning, Middle, and End.
Which is why they are ultimately meaningless.
Enter Subtxt’s narrative framework, a development of the Dramatica theory of story, which brings objectivity into the mix, breaking away from the chaotic subjectivity of these traditional models. Subtxt relies on fundamental principles and thematic relationships that are consistent and reproducible, ensuring a complete and balanced narrative.
The Storyform is a blueprint of meaning: a carrier wave meant to communicate the Author’s dramatic argument to an Audience. The narrative structure contained within a Storyform is not intended to make things easier for an Author as much as it is to make things possible for an Author.
Subjectively biased notions of story structure are simply folders to throw events into; an objective appreciation of narrative structure is a frequency of effective communication with an origin firmly rooted in Author’s intent.
Four Throughlines:
Four Domains of Sources of Conflict:
Granular Thematic Structure:
Consider the film The Fugitive starring Harrison Ford. Let’s walk through how Subtxt objectively handles its narrative structure, providing concrete examples for clarity:
Objective Story in Universe: The main plot revolves around "Dr. Richard Kimble" (Harrison Ford) being wrongfully accused of murdering his wife. The conflict is rooted in a fixed external state—Kimble's status as a fugitive and the relentless pursuit by U.S. Marshal Gerard.
Example: Kimble’s predicament of being a convicted murderer becomes a situation he must continually navigate, illustrating the stuck and unchanging nature of his external state. This is seen when Kimble is cornered at a Chicago St. Patrick’s Day parade, highlighting his constant external predicament.
Main Character in Physics: Kimble’s personal conflict revolves around his active efforts to clear his name and find the real killer. His journey is filled with external activities—running, investigating, and evading capture.
Example: Kimble actively breaking into the one-armed man’s apartment to gather evidence illustrates the process of investigation as a means to clear his name, demonstrating the external physical activities that drive his story.
Obstacle Character in Psychology: Deputy U.S. Marshal Samuel Gerard (Tommy Lee Jones) serves as the Obstacle Character. Gerard’s internal ways of thinking—his initial belief in Kimble's guilt and his relentless pursuit—create ongoing conflicts.
Example: Gerard’s determination to capture Kimble, shown when he orders the to be drained to find him, reflects his unwavering psychological stance of pursuit at all costs.
Relationship Story in Mind: The dynamic bond between Kimble and Gerard evolves over the course of the story. Initially adversaries, their relationship shifts as Gerard’s perspective on Kimble changes.
Example: The pivotal moment in the hospital, where Gerard decides to listen to Kimble instead of immediately capturing him, marks a transformation in their relationship. This shift from antagonism to a more nuanced understanding shows the change in their mental states and attitudes.
By anchoring The Fugitive in these objective principles, Subtxt ensures that the narrative is cohesive and fully fleshed out. Each Throughline and Domain offers a structured and balanced way to understand character motivations, plot progression, and thematic depth.
The Ultimate Chart of Narrative Structure does a brilliant job of showcasing the subjectivity and transposability of traditional narrative models. However, this strength also belies its weakness—it reveals how these models can become pointlessly convoluted.
Subtxt, however, offers a refreshing, objective stance. It refuses to dig into the arbitrary but instead creates a definitive framework that provides clarity and coherence to storytelling. In the end, the “Giant Chart of Subjectivie Bias” becomes a testament to Subtxt’s superior objective method, a method that transforms the chaos of narrative relativity into an ordered, meaningful structure. 🌟
Be wise and objective in your storytelling—embrace Subtxt.