An unreliable narrator is one whose perceptions or judgments the reader can't fully trust—but calibrating this in AI requires precision, since the model can easily blur the line between intentionally deceiving the reader and simply breaking character. The work lies in designing prompts that make the deception structural rather than accidental.
Unreliable narrator calibration is the technique of prompting AI to generate fiction where the narrator withholds, distorts, or misinterprets information in ways that serve the story rather than confuse the reader.
Mastering this technique allows writers to craft psychologically complex perspectives that reveal character bias without losing reader trust, and AI can help stress-test narrator consistency across long manuscripts by flagging where the distortion logic breaks down.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.