AI UsePolicy

In short, our advice is this: Proceed with caution.

In writing our policy on the use of LLMs and image generation models in creating submissions to Refactor, we knew we had to balance these two ideas:

We believe human work is valuable, irreplaceable, and deserves to be fairly compensated.

We believe using assistive and augmentative technologies can yield interesting, novel, and labor-saving results.

That’s why the theme of Issue 1 is Augment. And that’s why we’re running a story contest about augmentation.

Help figure out the future of augmentation here. And you can apply here to be an editor.

For now, here’s our AI policy. It’s subject to change based on what we learn together. We’ll show you the different versions as they arise (we’re sure they will).

With thanks to Scott Cunningham for inspiration from his class AI use policies.


Feel free to use AI, or not, to generate your works. Sometimes we can tell. But we are sure there are also times we won’t catch. It might or might not have long-term impacts on your writing. We have sympathy for the purist view. But we also recognize that that is largely from a bygone era.

The effort is the thing.

Something can be AI generated and effortful. Something can be human generated and slop.

The quality is the thing.

That is what we are looking for, and that, we believe we can consistently discern.

The originality is the thing.

Modern frontier LLMs are all built on pirated content. All human creativity is derivative. Plagiarism is wrong. These are all true.

We advise that you Steal Like An Artist.

If you do use an LLM to generate, edit, or modify your works, we strongly encourage you to submit your chat logs. We reserve the right to reject works for being low-effort or low-quality, whether they were human-generated or not. We find that this is an important and valuable part of your thought process. Let us meet your rubber ducks.

Notes

Kosmyna, Natalia, et al. "Your Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant for Essay Writing Task." MIT Media Lab, 2025.