AI Usage Policy ✨

Introduction 

I use AI tools across different formats: texts, podcast, images, and video. This page explains where the tool helps technically, where it influences the creative process itself—and how I mark that for the reader.

Where I use AI 

• Texts. From proofreading to co-authoring on phrasing and structure.
• Podcast. Transcripts, brief episode descriptions, show notes, notes inspired by episodes.
• Images and video. Generation and processing of visual materials for projects.
• Other. Translation, search, summaries, outlines, and organization of materials.

Basic labeling principle 

Minor assistance is not labeled: spelling and punctuation in texts, basic transcription and small description edits in the podcast, non-essential processing, conversion, and export for images and video. If AI’s involvement affects the way I work or the final form, I label such materials separately.

Labels for texts 

I mark only three cases—when AI is materially involved in the writing process.

Architect 

The text is written in dialogue with a machine. Here AI is not a proofreader but a mirror and conversation partner. I formulate the idea, and it helps unfold it: suggests new angles, brings order, connects parts, clarifies the line of reasoning. The voice remains mine.

Dictation 

The source was a voice recording. Sometimes a note is born from live speech. For example, I may dictate it into my phone during a walk. AI turns the audio into readable text: removes repetitions and filler words, smooths out the rough edges of spoken language, and helps organize the thought into paragraphs. My voice, rhythm, and intonation are preserved.

Based on the podcast 

The text is based on a key idea from an episode of my podcast. I take the main idea, expand it into a text, add clarifications and examples. AI helps adapt and structure it, and the podcast serves as the starting point.

Labels for other formats 

• Images and video. I mark AI involvement when I consider it important for context and perception.
• Podcast. If AI created the transcript or description, I note that when it matters for the listener.

How this appears on the site 

• Texts. I place the ✨ emoji next to the date, in the title meta tag, and add a short note at the bottom about AI’s role.
• Images and video. I label in the caption or directly under the work—for example: “AI-generated” or “AI processing.”
• Podcast. I indicate it on the episode page when it matters for the listener.
• Projects. I indicate it on the project page when it matters.

If there’s no label, it means AI only helped technically and didn’t influence the content—or I didn’t deem it important.

Principles 

• Transparency. I don’t hide the tools that help me work.
• Authorship. I am responsible for meaning and form. The tool does not replace my voice.
• Benefit to the reader. I label only what affects perception.
• Minimal noise. I don’t mark technical cleanup.
• Privacy. I don’t publish drafts, source recordings, or chat history unless necessary.

Updates 

• 2025-09-10 — Clarified that I don’t always label AI-generated images and videos.
• 2025-09-02 — Created a new structure and fully rewrote the text.
• 2025-07-29 — Created the page.