At the Edge of Screenwriting
How I’ve Come to View Storytelling After Working With AI For the Last Three Years
I began my quest to learn and master the art of screenwriting 12 years ago. It started as a desperate attempt to get out of my depression from living an aimless life but has since grown into an obsession for the sake of the craft, itself. Sure, I’d love to get paid for my work, but around 2020 a new fascination had bubbled up in my life that began to steer me in another direction: Artificial intelligence.
Okay, yeah, yeah. Go ahead and roll your eyes. Call me a tech bro, an accelerationist, a naive idealist, or whatever you want. I don’t care. AI is scary, cool, stupid, fun, and everything all at once. It’s truly one of the most historical events in human history. So you’d be crazy not to, at least, be fascinated by it. That’s why my brother and I decided to dive headfirst into it and that’s where we’ve been for the past three years, synthesizing our knowledge of screenwriting and storytelling with LLMs to produce tools that can help with ideation and streamlining.
What I’ve realized about storytelling, from working with AI, is that even though they appear elusive and almost “gaseous” in nature, they’re actually made up of very small, meaningfully discrete parts that, when pieced together, form the whole, which is the story, itself.
Now I know I’m gonna get a lot of pushback, so let me try and quell the pitchforks by stating that ultimately stories are whatever we want them to be, and at the end of the day, there is no formula for making them.
And yeah, yeah. They come from the heart and all that. But at the same time, the vast majority of all stories, from terrible to fantastic, seem to also contain individual properties where the meaning they exude onto the audience is different and unique with every story, but their functions and how they serve the viewers are more or less the same.
Take a scene, for example. The entirety of one is seamless, or at least, that’s the aim. But if you look closely enough, it’s really the compilation of several properties that relate to each other. There’s the situation that is happening in the scene, of course. Then there’s the overall tone, which influences its other main properties, specifically, the setting where the situation is taking place as well as the elements and motifs. And then last, but not least, there are the characters who are speaking and taking action.
However, it doesn’t end there. You can get even more granular with each property. Regarding the characters, for example, you have the person that they are, what they believe about the World and how one should live their lives, what they want at this moment, what they know, what they don’t know, what they’re hiding from others, as well as what they’re not hiding from others.
Another example is the situation in the scene, which is typically broken down into a kind of mini three-act structure with causes and effect moments that can increase and decrease the tension that is happening. Then when it comes to the tone, you can break it down to choices that accentuate the central message of your story. In fact, all of the choices you make in your story should relate to the central message, which even that can be broken down into a dialectic that your characters embody, allegorically.
So you see, stories are a series of properties that you build, which form a logical matrix that’s connected to create a story that expresses something meaningful about reality. And, it took endless interactions with AI for this to truly sink in for me because I had to sit in front of a screen and carefully break down exactly what a scene is so that I could instruct it properly.
That’s like trying to analyze and explain exactly how you’re able to envision a scenic mountain in your head or understand language and talk. Okay, maybe not that hard, but still. You’re taking what you do all the time and distilling it down into its main components.
Now this brings up a very interesting question. Where did these properties emerge from that seem so pervasive in all of our stories? Well, I suppose it’s obvious. It came from all the greats who wrote before us, right? We copied each other so much to the point of forming informational matrix patterns that could be broken down into properties like the ones I outlined above.
And now those properties can be fed into a machine that can pick up on the patterns to generate stories. Albeit, the outputs are still way off from being great because humans are more advanced, but with the right elbow grease, they’re not bad at creating basic frameworks to build off of. And they’re only getting better.
No surprise, LLMs mirror the patterns we use in stories, which subsequently makes it easier for people who interact with AI to recognize these patterns. The distant future of this technology is uncertain, but the immediate future is quite clear. It’s going to act as a digital exoskeleton for writers, which is why I believe geniuses will use AI to tap into fundamentally new realms of storytelling, both in meaning and in the construction of the story. They’ll use it as a tool to better examine the relationship between the discrete parts, which will make it easier to push beyond the frontiers of storytelling.
What Chris Nolan can make sense of through storytelling, will be achievable for many who put the effort in. That doesn’t mean everyone will be able to create Inception because you would still need to be a creative genius. But, at least, a lot of writers will be able to obtain as deep of an understanding of stories as him, or at least, close to him. If you haven’t seen it already, I highly recommend you listen to one of the latest Scriptnotes podcasts that had Christopher Nolan on. His analysis of stories sounds like a mathematical proof and does a good job of alluding to this point about stories being made up of discrete relatable properties.
So when it becomes easier to see these related parts, it will become easier to manipulate them, and that will birth new patterns for us to explore in stories. We’re great at finding creative solutions, but we’re terrible at analyzing and recognizing meaning from huge amounts of data. This is why it’s so hard for creators to push stories into new frontiers, given how much information is contained in just one of them. Now, imagine trying to extrapolate meaningful patterns from the fundamental properties of every story ever made by humans…Yeah, that’s not happening without AI assistance.
We’re on the cusp of a fundamental revolution in everything, which means there will likely be a breakthrough in storytelling that hasn’t been seen since Shakespeare. But future breakthroughs may one day be purely AI-generated, which then begs the question…What’s the point of even trying, then, if a God-like machine will do it better than any of us?
Well, here’s a couple of good reasons. First, that's a speculative future. Nothing is set in stone, so why make any decision about something that may not even come to pass? Second, AI is a machine. So if it becomes sentient one day and transcends beyond God-like status, do you honestly think they’re going to give a shit about winning the Nicholls Fellowship?
Do you think they’ll want recognition, money, or that shiny golden trophy? Do you think they’ll even consider screenwriting important enough to engage in it? They can dominate and control our lives all day long, but the human heart will always create and tell stories, whether it’s in the rubbled sewers of the last human enclaves or the lush spaces of an AI-run utopia. The only real thing that will likely change is how people use stories to make money. But writing and telling them will continue to be an important skill.
Ultimately, we write and tell stories because we love them. And yes, we want to get paid as we should, but that should always come second to the passion of the craft. Too many of us compromise to become more marketable. Of course, if you’re collaborating with anyone, you have to learn how to compromise and take in other people’s ideas. Otherwise, no one will ever want to work with you!
But, there’s a huge difference between taking notes to enhance the story versus taking notes to make the story more profitable. Ideally, you should do both, but it seems much more common these days that our choices are motivated more by money than by story. And that’s probably because material greed is caked into our culture, from the bottom to the top (at least in America).
I mean, the only real difference between Jeff Bezos and the average person is that Jeff Bezos can scale his efforts to obtain more, which, of course, has a much bigger impact. So rightfully, a lot of the anger is placed on people like him. But the truth is that most of us have this propensity to want more than what we have. And it’s that propensity, which drives this complicated system to do some pretty horrible things like support businesses that use slaves.
If we’re all motivated to blindly want more then we’ll continue to perpetuate these growing negative externalities, and with AI, this could lead to disaster as it will be used to fulfill our desire for material gain. So as we carry along on this adventure, I think we’re going to need to do some real soul-searching so that we can reprioritize what we place most of our value on. Should we only strive for material gain and prestige or should we also strive to elicit real meaning and heart into our stories in such a way that can give us the leverage we need to build the life that we want for ourselves? Be prepared to wrestle with this question as we linger on.
Hope this article was insightful and as always, best of luck in your creative endeavors!
Story Prism
___________________________________