Last updated 05-08-2024
Category:
Reviews:
Join thousands of AI enthusiasts in the World of AI!
Earkind
Earkind is an innovative platform that revolutionizes podcast creation through artificial intelligence. It offers AI-generated podcasts with dynamic and engaging content that keeps audiences entertained. The process starts with a text file containing a selection of news and URLs to recent arXiv research papers.
Earkind then crawls and extracts pertinent information, which is processed by the chatGPT API to create podcast scripts. The scripts are turned into audio using Azure's text-to-speech services, with cutting-edge research backing the technology. Earkind stands out with unique characters in its podcasts, providing a lively discussion format.
The result is a complete podcast episode that combines information, entertainment, and a high level of production quality, including background music, sound effects, and polished editing. Committed to community engagement, Earkind encourages feedback and plans to share its behind-the-scenes creation process.
AI-Generated Podcasts: Craft engaging podcast episodes using AI technology.
Character-Driven Content: Utilize distinct characters to provide a conversational feel.
Text-to-Speech Technology: Employ Azure’s advanced text-to-speech for audio content.
Automated Editing: Combine audio clips with effects and music for final production.
Open-Source Aspirations: Intention to make the code public and provide detailed explanations.
1) How does Earkind create podcasts?
Earkind uses a combination of language models (LMs), neural expressive text-to-speech (TTS), and programmatic audio editing to generate full podcast episodes.
2) Who are the characters in Earkind's podcasts?
The show features hosts like Giovani Pete Tizzano, and collaborators Robert, Olivia, and Belinda, each with unique characters to keep the content engaging.
3) What is the process of selecting content for podcasts on Earkind?
To start, a .txt file with news and links to arXiv research papers is used. Then the title and abstract are extracted, and further information is obtained from the raw PDF text.
4) Does Earkind plan on making its code available to the public?
Yes, aside from using Azure’s TTS for audio generation, Earkind uses Pydub for editing and plans to release the source code publicly.
5) How can I provide feedback on Earkind's podcasts?
Sergi at sergi@earkind.com welcomes feedback and is keen to hear from the audience regarding their interest in Earkind's podcasts.