About
Voices are AI-generated
The guidance in Seven Doors is spoken by generated voices, not recorded by a person.
What you hear
The spoken guidance in the app is made with text-to-speech technology. Each line of a practice is written first, then turned into speech and checked before it is published. You can choose between two voices, Voice A and Voice B.
How the voices are made
The voices are made with a Gemini text-to-speech model from Google. The audio is generated in advance and downloaded with each practice, so nothing is generated while you listen, and the app never records your voice.
Why AI voices
Generated voices let every practice exist at three lengths and four levels of guidance, and let us correct a line quickly after review. We tell you plainly because you should know who, or what, is speaking to you.
The words
The practice scripts and method pages are original writing for Seven Doors. Many were drafted with AI assistance, and each one is reviewed for source accuracy, tradition and safety before release.