Earkind

Earkind is an AI-powered podcast generation tool that creates daily episodes about artificial intelligence news and research. It takes plain text news items and arXiv research paper URLs as input, then generates complete podcast episodes with intro/outro segments, sponsor messages, and character discussions. The system uses large language models to create scripts organized as conversations between distinct AI personalities, including an enthusiastic host, a skeptical analyst, and a witty research expert.

After script generation, Earkind uses Azure's neural text-to-speech technology to create voice recordings for each character. The tool then automatically edits the audio using Pydub, adding jingles, sound effects, background music, and adjusting volumes. It creates a complete podcast package including descriptions with timestamps and episode titles, all generated through the same AI pipeline.

The tool is designed for AI enthusiasts and content creators who want to stay updated on AI developments through audio content. It's particularly useful for researchers who want audio summaries of recent papers or creators looking to generate regular podcast content without manual production work. The character-driven approach makes technical content more engaging and accessible to listeners.