Mormon Lit Lab Artificial Intelligence Philosophy

Friends,

Fifteen years ago, we started the Mormon Lit Blitz to connect writers with readers as they explored a shared religious culture in literature. We’ve been quite pleased with the results. Our contest’s short length limit and online publishing model have helped writers experiment and made it easy for readers to get to know a wide range of voices. Even for many people who haven’t been finalists, the Mormon Lit Blitz is a chance to try out new ideas and find a personal voice. We believe that there’s real spiritual as well as cultural value in that kind of exploration.

As we prepare for the 15th annual Mormon Lit Blitz, the landscape of online reading and writing is changing as Artificial Intelligence systems evolve. With these changes, we will be adding new AI disclosure requirements. The key points are:

• In a literary contest, readers expect and deserve a direct human encounter.

• Going forward, contest entrants will be required to disclose AI use.

• Using AI to generate text in a piece for the contest is not allowed.

• Some AI uses in the research process or for coaching are permitted. 

This moment, though, calls for more than rules. We also want to share our philosophy on AI use and join an ongoing conversation about how AI should fit into our culture. We welcome thoughts and feedback from our writers and readers about these issues as we move forward.  

OUR AI PHILOSOPHY

I. Why AI is Interesting

For someone who loves language, right now is an interesting time to be alive–even if it’s also unsettling. At the Mormon Lit Lab, we’re trying to bring the same curiosity to our interactions with AI as other subjects.  Literary writers have good reasons for concern about artificial intelligence. Fiction was key in the training datasets for today’s large language models–and in the process, many writers had their work used without their consent. Given the evidence of artificial intelligence companies’ disregard for ethical and legal concerns while developing the technology, it’s reasonable for writers to anticipate that companies will place a higher value on getting ahead in the AI race and marketplace than on the interests of writers or human culture. 

At the same time, writers are in a unique position to recognize just how impressive it is that a technology can now interact with and through natural language. We are in a unique position to understand just how complex human language is. It’s wild that computer programs can now identify patterns in language well enough to respond to context and connotation. It’s amazing that they can use pattern-recognition and prediction techniques to produce clear language. And we’re probably only beginning to understand what it might mean now that computers can guess at the intention behind natural language rather than relying only on keyword searches or explicit commands. 

As writers, we can celebrate the way that AI’s ability to interact with language serves to remind people of the power of language. We can appreciate how AI reminds us of the intricacy of the patterns humans leave in the words we organize. 

As religious writers, we’re also in a unique position to understand that there’s a difference between the craft we bring to writing and the agency we exercise as we write. The writings in the Mormon Lit Blitz are not only made of well-arranged language. They also reflect writers’ conscious choices about what is important, about what to emphasize when depicting people’s thoughts and feelings, and about the tensions that define our inner lives. 

Stories and poems are less about facts or events than about how we choose to see and interpret them. Literature is made by devoting an overabundance of attention to a subject, because literature is fundamentally about the agency at work in how we choose to use our attention. Literature, then, requires both an awareness of existing patterns and an intentional effort to create new ones. 

Many writers have used their work in the Mormon Lit Blitz to reflect on types of stories we tell all the time in our culture–and then interrupt readers’ rote habits of response. Writers use their agency to invite readers to consider the limits of our normal ways of seeing things. Writers invite readers to make more space for their souls. Writers in this contest understand the interplay between novelty and tradition, the power of rooting a piece of art in the familiar while also reaching toward fresh vision. 

As champions or writers and caretakers of culture, we at the Mormon Lit Lab are interested in what large language models can do–and conscious of their limitations. Our deepest commitment is to human agency and vision. We feel strongly that AI is best used in ways that are complementary to that human core. 

II. Guiding Principles for AI Use in Literary Work

With that philosophy in mind, what do we see as the best and worst uses of artificial intelligence in the writing process? 

In general, we do not believe that AI should be used to make decisions for writers. If used, AI tools should be focused on helping human writers to see possibilities, which the writers then make choices about. Here are some examples of how we see that principle playing out in the writing process

Ideation: 

Writers should not use artificial intelligence to replace their own intuitions about what is worth writing about. But once they start playing with an idea, writers can use artificial intelligence to see what’s been done before, to explore different formal possibilities, to help think about how to scope their idea for the word count, and so on. 

Acceptable: A writer is interested in writing a poetic lament inspired by the Book of Ether and asks AI to point to examples of previous literary writing and scholarship about that book.

In this example, the writer is using the AI to learn about context and possibilities. By comparing their own work with previous work, the writer can use their agency to do something more distinct and original within the larger cultural tradition.

Acceptable: While trying to adapt a larger idea for the Mormon Lit Blitz, a writer asks an AI how many characters appear in a typical flash fiction story and what examples of flash fiction stories with higher character counts are. 

In this example, the writer is trying to understand conventions and see options for pushing their limits. The writer then uses their agency to choose what is best for their piece.

Prohibited: A writer starts their Mormon Lit Blitz entry by asking an AI, “What should I write about?”

In this example, the writer is transferring their agency to the artificial intelligence. Even if the writer still chooses between options that the AI offers, the AI has set the frame.

Research: 

Writers should not replace their own skills of analysis and discernment with the uneven capacities of artificial intelligence. Writers can productively use artificial intelligence as a part of a large research process to identify leads and examine scenarios.  

Acceptable: While writing a historical fiction piece, a writer asks an AI for examples of common foods from that time, place, and class as a starting point for further research on a meal description.

In this example, the writer is using a natural language search to find ideas and terms that might be useful in follow-up keyword searches. The AI helps find things that might fit at an intersection of ideas, but the writer still looks to human-written sources to confirm and for additional details.  

Acceptable: While writing a science fiction piece, a writer carries on a conversation with an AI about how a task might unfold under zero-gravity conditions and what physics principles are at work in that process.

In this example, the writer takes advantage of a natural language search to ask a scientific question about the possible effects of a specific situation that might be prohibitively difficult to research via keyword search. By asking for the scientific principles in play as well as for predications about the interaction, the writer has leads for follow-up research to confirm the AI-generated possibilities.

Prohibited: A writer asks an AI what a character would do in a given historical or scientific situation. 

Asking what might be typical in these situations would be background research, but asking what the character would do crosses a line. The writer gives up their agency and authorship if they are asking the AI to choose character behaviors or define character reactions.

Drafting:

Acceptable: A writer asks an AI for coaching tips on different ways writers set and keep drafting goals.

In this example, the AI is helping with drafting by showing the writer possibilities for their process. The writer is using their agency to decide which process ideas make sense for them.

Prohibited: A writer uses an AI to generate language for all or part of their piece. Whether it’s the full piece, a scene, or even a single description, AI-generated prose is unacceptable for the contest.

In literary writing, the texture, voice, and pacing are all part of the experience. A person who outsources choices about those things is not doing literary writing.   

Revision: 

Acceptable: A writer is interested in using simple language and concrete images in a poem. The writer asks an AI to review the piece and highlight examples of more complex words or abstract images.

In this example, the writer is choosing their own goals about the style of the piece and using the AI to help them identify possibilities for where the writer might further pursue those goals.

Prohibited: A writer asks an AI to fix or clean up the language in their piece.  -In this example, the writer is deferring any judgement about what makes language good in the context of this piece to the AI. The writer is also inviting the AI to directly change the text, which violates our rule against drafting through text generation. 

When using AI tools in any stage of the process, writers should recognize that these systems were not refined over generations, like a hammer or saw, to get a job done in the best way. They were made by companies to be habit-forming. Left unchecked, AI tools will nudge you to use them more and exercise your agency less. 

We encourage writers to use their own creativity to reduce AI models’ natural tendencies. To discourage dependence on the models’ judgements, you can instruct an AI not to include the praise for your work. Remember that the praise was only there to promote user engagement in the first place. You can also tell an AI model not to offer to write scenes for you. Left unchecked, an AI will tend to seek roles for itself without regard to whether they are good for your work, for the culture, or for you as a person. When you use this tool, try not to be used. 

III. New Requirement for AI Disclosures 

Starting in 2026, writers who submit to the Mormon Lit Blitz will be required to agree to follow these principles and to disclose AI use in each of these four areas (Ideation, Research, Drafting, and Revision). If writers used AI in these areas or any other part of their writing process, they’ll be asked to make a brief statement explaining their use. 

These disclosures are a good self-check for writers. We hope they’ll help you be intentional in your own process. They’ll also help us maintain a relationship of trust with readers. Most readers pay closer attention to poems or short stories as they read than they do to the language in genres like billboards, social media posts, or user agreements.  As a reader, it’s a terrible feeling to start reading a story closely and then realize that it has the generic texture of an AI-generated work. No one wants to read a piece of literature more closely than the author did. At the Mormon Lit Lab, we want people to know that the work they read came from a human mind, heart, and spirit. We want them to know they’re interacting with human attention and human agency. 

Feedback

We hope these principles will help the writers and readers in our community experiment with new tools while maintaining the basic trust that they’re having a human experience. If you are a writer or reader of our contest, we’d love to hear from you about this philosophy of AI use. What do you think of this philosophy? What do you think of the policy aspects? If you’ve used AI systems on a creative project, what was your experience like and what advice do you have for other creators? 

This is an important moment in the history of writing. We hope that we can work together to face it well.

Sincerely,

James Goldberg

Kathy Cowley

Liz Busby

Merrijane Rice

Mormon Lit Lab AI Committee