What is this?
Ever have a random question pop into your head while doing the dishes, walking the dog, or trying to fall asleep? That's what this project is all about. I capture these spontaneous prompts using my voice, send them through an AI processing pipeline, and turn them into full blog posts complete with audio episodes and AI-generated responses.
How It Works
- Voice Capture: I record a question or prompt using the Voicenotes app
- AI Processing: My voice is transcribed and cleaned up by Gemini Pro 2.5
- Response Generation: Claude Sonnet 4.5 generates a comprehensive response
- Audio Creation: The response is converted to speech and combined with my original voice recording
- Publication: Everything is automatically published here with images, audio, and text
Why?
There's a gap in AI tooling for preserving and sharing interesting LLM outputs. This is my experiment in:
- Building a personal knowledge base of AI-assisted insights
- Creating multimodal content from voice-first interactions
- Exploring the future of human-AI collaboration
- Making weird, random questions into something useful (or at least interesting)
⚠️ Fair Warning
This content is AI-generated. While I do curate and review posts, some responses may be incomplete, inaccurate, or just plain weird. That's part of the experiment! Think of this as a public digital garden where AI and human thinking intersect in unexpected ways.