AI Thinking Process Narrator

A unique tool that combines Groq's LLM capabilities with OpenAI's Text-to-Speech to create an audible, stream-of-consciousness experience of an AI's thinking process, complete with creative tangents and unexpected connections. The tool specifically focuses on narrating the AI's thought process, intentionally ignoring the final response to maintain the stream-of-consciousness experience.

Features

🌀 Creative Tangents: Explores unexpected connections and parallel concepts
🗣️ Real-time Audio Narration: Converts thoughts to speech as they're generated
📝 Streaming Text Output: Visual representation of thoughts in real-time
🔄 Continuous Exploration: Multiple iterations of increasingly creative connections
🎯 Process-Focused: Only narrates the thinking process, not the final answer
⌨️ Interrupt Handling: Clean exit with Ctrl+C
🧹 Automatic Cleanup: Temporary audio files are removed after execution

❤️ Join my AI community & Get 400+ AI Projects

This is one of 400+ fascinating projects in my collection! Support me on Patreon to get:

🎯 Access to 400+ AI projects (and growing daily!)
- Including advanced projects like 2 Agent Real-time voice template with turn taking
📥 Full source code & detailed explanations
📚 1000x Cursor Course
🎓 Live coding sessions & AMAs
💬 1-on-1 consultations (higher tiers)
🎁 Exclusive discounts on AI tools & platforms (up to $180 value)

Requirements

groq
openai
pygame
termcolor

Environment Variables

The following environment variables need to be set:

GROQ_API_KEY: Your Groq API key
OPENAI_API_KEY: Your OpenAI API key

Configuration

Key parameters at the top of the script:

PROMPT = "Explain why the sky is blue, but think very short about it."
STYLE_OF_THINKING = "very jovial and fun and daydreaming like"
CONTINUOUS = True  # Enable creative tangent exploration
MAX_ITERATIONS = 3  # Maximum number of thinking iterations
VOICE_MODEL = "tts-1"
VOICE_NAME = "onyx"

How It Works

Initial Thinking:
- Converts thoughts to speech in real-time
- Only processes content between <think> tags
- Stops processing when thinking ends (at </think> tag)
Creative Tangents:
- Each iteration explores new connections like:
  - Metaphysical implications
  - Historical perspectives
  - Cultural connections
  - Scientific parallels
  - Artistic analogies
  - Psychological aspects
  - Philosophical considerations
Audio Generation:
- Completed sentences trigger TTS conversion
- Audio plays sequentially while new thoughts generate
- Only thinking process is converted to speech
- Final conclusions are shown but not narrated
Continuous Exploration:
- Uses previous thoughts as springboards
- Discovers unexpected connections
- Maintains consistent thinking style
- Each iteration only processes new thinking, not conclusions

Important Note

This tool is designed to focus exclusively on the AI's thinking process. While the model will generate both thoughts and a final answer, only the content within the <think> tags is processed and narrated. This design choice emphasizes the stream-of-consciousness nature of the tool and allows users to focus on the journey rather than the destination.

Example Tangents

Starting with "Why is the sky blue?", the AI might explore:

Connection to human perception and color psychology
Cultural significance of the color blue across civilizations
Parallel with ocean depths and spatial perception
Musical analogies about frequency and wavelength
Philosophical implications about reality vs. perception

Usage

Set up environment variables:

export GROQ_API_KEY='your-groq-key'
export OPENAI_API_KEY='your-openai-key'

Run the script:
```
python deepseek_groq_audio_thoughts.py
```
Listen as the AI explores your prompt through multiple creative angles!

Customization

Modify STYLE_OF_THINKING for different thinking personalities
Adjust MAX_ITERATIONS for more or fewer tangents
Change VOICE_NAME to alter the narrator's voice
Edit PROMPT to explore different topics

Error Handling

Graceful handling of API errors
Clean termination with Ctrl+C
Automatic cleanup of temporary files
Informative error messages with color coding

Notes

Audio files are temporarily stored in system temp directory
All files are cleaned up after execution
The system uses streaming for both text and audio
Memory efficient with queue-based audio processing
Each iteration builds creatively on previous thoughts

Dependencies

Groq: For LLM capabilities
OpenAI: For text-to-speech conversion
Pygame: For audio playback
Termcolor: For colored terminal output

License

MIT License

listen-to-deepseek-r1-thoughts

README