Use Context to Deepen Our Writing From AI Slop

For as long as I���ve been a writer, I���ve heard the advice, ���Only write if you can���t not write.��� The idea behind that tough-love guidance is that writing can be ridiculously hard work, publishing can be even harder, and that most published authors won���t make enough money to justify all that effort, so don���t start down this career path unless you can���t imagine any other journey. This advice is even more timely given that we���re now in competition with stories ���written��� by generative artificial intelligence, often called AI slop.

The overwhelming number of books on Amazon has ballooned with several get-rich-quick schemes over the decades, from publishing fanfic and using ghostwriters to the current explosion that uses AI to Frankenstein a story mimicking the pirated books it���s been trained on. As Angela has written about previously, far too much publishing advice the last couple of decades has focused on volume������just learn to write faster!��� But no ability to write faster will help us compete with the countless-and-never-ending amounts of AI slop filling Amazon���s shelves.

So how can we compete? Angela has a fantastic post with lots of great advice on how to outshine AI-generated books, but I want to share some thoughts from the perspective of looking at our strengths compared to AI���s weaknesses. Namely, AI doesn���t understand context, so the more we lean into and deepen our use of context within our stories, the more we can make them meaningful, authentic, and different from the slop.

Wait��� What Does ���AI Doesn���t Understand Context��� Mean?

Contrary to the term ���artificial intelligence,��� AI isn���t actually intelligent. It doesn���t understand the context or meaning of the words it spouts. It can recite definitions, but it doesn���t truly understand them.

Rather, the software currently being touted as AI is just a predictive algorithm, a more complex version of the auto-complete suggestions that have shown up on our phone���s keyboards or word-processing programs for years. It merely spouts words that it predicts should follow a prompt or previous set of words.

The generative AI that we���re talking about today (as opposed to the types of useful-and-ethical AI, such as analyzing images for finding cancer cells or wildfires, etc.) generates strings of words that seem like they go together. That ���seems like��� is based not on the context of the words or situation, but on the percentage of times it���s seen those words clumped together before. That���s it ��� it���s just predicting what should come next based on percentages of previous occurrences it���s seen in its training data (i.e., pirated books).

In other words, our current generative-AI technology gives the most predictable answer or story beat. It���s the ultimate source of stereotypes and obvious plot points, dialogue lines, and character traits.

In addition, because generative AI doesn���t understand the meaning of words���the context of words or situations���it can���t make unobvious or unpredictable connections. It will only generate ideas that it���s seen before, and it will only connect ideas that it���s seen connected before.

Those weaknesses of generative AI hint at how we can differentiate our work: making unique story choices and using context to deepen meaning.

Tip #1: Avoid the Obvious and Predictable

Long ago, an agent���s blog post (that I can no longer find) advised that writers should never rely on the first idea that pops into their head. He said that our first idea is often the easiest, top-of-mind, most predictable, and most obvious. He pushed authors to dig deeper and consider using our third, fourth, or fifth idea instead, as those would more likely be less obvious, and thus, less predictable. This guideline may be even more applicable now.

The less predictable our writing and storytelling, the more our book will be unique. In turn, our story will be unlike all the copy-and-paste AI slop that���s been cobbled together from all the stolen stories in the AI���s training data.

While this advice won���t necessarily help us immediately stand out from the bursting Amazon shelves, unique stories will earn us a reputation among our readers and reflect in our reviews. That reputation for unpredictable���yet logically, thematically, and/or emotionally connected���storytelling will eventually help us stand out from the crowd.

Tip #2: Deepen Context and Connections

Think of all the ways that a story authentic to the human experience includes context and situations that have no ���one true answer��� and involve nuance beyond an AI���s understanding:

How does a character���s past affect their current thinking and emotions?How does a character���s backstory wound affect their ability to relate to others?How does a character���s past affect their dreams, goals, and motivations?How is a character���s past connected with their choices?How do the story���s obstacles interact with those aspects of the character(s)?How do the story���s events affect the character���s growth?How do all those aspects of the character and the story interact with other characters and their histories?How do the antagonist and protagonist���s character traits connect and reflect each other?How is a character���s success or failure in overcoming obstacles connected to their choices?How is a story���s theme reflected through all of the above���from goals and motivations to choices and success?How will readers relate on a deeper level to the emotions, choices, and themes of the story?

And so on���as the list is endless. Every element of storytelling that makes a story genuinely human and relatable is a source of context and connection that AI doesn���t actually ���get.��� It can make predictions and generate words based on what it���s seen before, but it doesn���t actually understand how past affects present, or the complexities of human relationships, or how growth changes us, etc.

Because it doesn���t understand the nuances and emotions of situations, AI can���t create writing that explores those connections in unpredictable, unique, or unobvious ways. It can���t dig into what those situations feel like and what they mean to our characters and their hopes for the future���or what they mean to readers. It certainly can���t pick up on subtext or hidden ideas to deepen readers��� insights into how previous events connect to every aspect of a character or why a character may make an unpredictable or unexpected choice.

In other words, a story generated by AI can go through the motions of those story and character elements and connections, simply by mimicking what it���s seen before. But since it doesn���t truly understand the human experience, it can���t go into depth and explore the situations and emotions in unexpected ways. It can���t reveal new insights into the human experience that are authentically relatable to readers.

Tip #3: Focus on Deep, Relatable Characters and Stories

Every character should be unique, and their personal context of their past and their goals, motivations, and choices should therefore lead to unique outcomes. As I���ve said since the fanfic-get-rich-quick days, if we replace our character with someone else, but nothing in the story changes, then we haven���t actually created a three-dimensional unique character.

The same goes for the story as a whole: If an AI could write our story, then we haven���t created the story and characters that only we could write���with our own context, history, and understanding of the human experience. It would mean that we not only haven���t created three-dimensional unique characters with their personal context guiding everything they do and choose, but that we also haven���t dug deeply enough into how all that context affects the story overall to create a new and unpredictable���yet relatable���story for readers to enjoy.

Final Thoughts: Differentiating Our Stories from AI Slop

None of this is new advice, obviously, but it���s more important now than ever to lean into our human understanding to differentiate our work from that of AI. (And yes, none of this will help us get our work the exposure we���d like, but Angela���s posts have good advice for that issue as well as specific tips for leaning into our human experience.)

There are many aspects of our career that we don’t control���historically, the traditional publishing path or our level of success have been obvious examples���but ensuring that we write good, unique stories with deeply human and relatable characters is something we can control. *smile*

Have you struggled with motivation to write given all the news about AI���s entry into the publishing world? Do you have different thoughts about AI or its weaknesses or our strengths? Do you have additional ideas for how we can differentiate our work?

The post How to Differentiate Stories from AI Slop appeared first on WRITERS HELPING WRITERS��.