I gave my local AI resident a voice using pocket-tts-server. It wrote the script itself.
GateClaw generated a 325-line philosophical essay about its own existence, then selected David Attenborough as narrator — "he's the voice of wonder, observing complex ecosystems."
The full 20-minute audio/video (static logo + subtitles) was generated locally using:
- pocket-tts-server (82 WAV celebrity voice clones)
- David Attenborough voice clone
- GateClaw SOUL.md self-reflection essay
If you use pocket-tts-server or build TTS workflows in ComfyUI, this pipeline is fully replicable locally.
→ GateClaw repo: https://github.com/ai-joe-git/GateClaw
→ pocket-tts-server: https://github.com/ai-joe-git/pocket-tts-server
→ Full essay video: