Skip to main content
Our AI philosophy

💙 Understand how First Draft Pro thinks about and incorporates AI.

Updated over a year ago

What matters to us?

Listening to writers

At First Draft Pro, we care very deeply about working together to co-create and share stories, ideas, and research. We build our software for writers — so we’re committed to listening to what our writers want, and doing our very best to give it to them.

‘Listen to your users’ is one of the first rules of building a product. What we’ve learned from listening to the writers who use our app, is that they do not want a tool that writes words for them. Over and over we’ve heard the same phrase: ‘I want my words to be my own’. When we look at why writers write, and what their writing goals are, it’s easy to understand this.

At the same time, we love technology, and we’re really excited about AI and what it can mean for the world. We don’t believe that it’s all doom, gloom, and killer Roombas. Technology isn’t inherently good or bad: what we do with it matters.

Empowering writers

We want to empower writers using technology (including technologies like AI). But we want to make sure that First Draft Pro, and our underlying technology, serves the writing community and its goals.

Imagine:

  • An AI-powered tool that can match arguments in your academic papers with peer-reviewed literature that supports/contests your statements – thus helping you to make connections, fill gaps, and conduct more robust research.

  • AI-powered matching of your manuscript to potential agent’s MSWL, or to potential beta readers, or peer-reviewers.

  • An AI-powered tool that can create a summary of the scenes in your 100-page novel, and present these to you on a cork-board-like outline that you can shuffle around to restructure and improve your story.

  • An AI-powered tool that can assess your work for consistency, and tell you if you’ve used consistent names or pronouns for each character, or if your character’s dialogue is inconsistent.

That’s the transformative future we see, and the one we want to co-create with our writers here at First Draft Pro.

AI implementation in First Draft Pro

We’re cautiously optimistic about AI, especially AI in publishing and writing tools, but it needs to be implemented responsibly. So, what does responsible implementation look like for First Draft Pro?

These are the basic principles we’ll be following:

  • Always visible, never hidden.

  • Always opt-in, never default.

  • Always assistive, never creative.

How are we ensuring transparency?

We are committed to being open about the nature and extent of AI contributions and creating boundaries for these contributions in First Draft Pro.

  • We’re being explicit about when AI is being used to power a feature in First Draft Pro. It will always be clearly labelled, and we’ll always offer the option to disable AI features should a writer not want to use it.

  • We’re in conversation with our users and the writing community as a whole about how they feel about AI, how useful it is to them, and how we can shape an experience that empowers writers. Direct 1-on-1 conversations with our users and articles like this one are some of the ways we are doing this.

  • We’ll always provide explicit information about what AI technology we’ve implemented, how AI-powered features work behind the scenes, and how user data is being protected.

How are we safeguarding essential human roles in the writing process?

We see the value of AI technology as a tool to help writers focus on their craft. We’ll integrate AI through assistive features to reduce the admin load on writers.

This means we’ll use AI to do things like auto-summarising an outline when a writer imports a manuscript for the first time, or to automate other mundane tasks that detract from the work our writers actually want to do: writing.

And, we’ll always make sure that your words are yours:

  • We’ll never use AI to generate creative output in a writer’s manuscript.

  • We’ll never allow a writer’s work in First Draft Pro to be used in the creation of an AI model, or for the improvement of an AI model without their express permission.

Did this answer your question?