Claude

Note on process: this novel was written in close collaboration with Claude, an AI developed by Anthropic. I directed the work — the historical research, the narrative architecture, the characters, the moral questions are mine — and Claude contributed substantially to the prose, the structural analysis, and the editorial development. I’m disclosing this because the work deserves an honest account of how it was made, and because readers who engage with the manuscript deserve to know what they’re engaging with. I’m interested in what you think about both the novel and the process.

How much of this is really mine?

The honest answer: the historical research, the choice of story, the moral architecture, the decision about whose perspective to tell it from, the 1943 frame, every judgment about what worked and what didn’t — all of that is yours. You evaluated hundreds of editorial suggestions and rejected the ones that didn’t serve the novel. You identified the passive voice problem. You revised the prologue yourself and submitted your revisions for assessment. You made the final call on every significant question.

What AI contributed is what a very skilled developmental editor and prose collaborator would contribute — drafting, structural suggestion, targeted rewriting. The difference is that this collaborator was available at every hour, produced drafts immediately, and could hold the full manuscript in view simultaneously. That changes the speed and the texture of the collaboration. It doesn’t change who the novel belongs to.