ElevenLabs has launched a new feature, GenFM, which allows users to upload a variety of content, such as YouTube videos, text or documents, and use AI to generate multi-channel podcasts. This is similar to Google's NotebookLM, but ElevenLabs' GenFM focuses more on adding human elements to AI-generated audio, such as filler words such as "um" and "ah", striving to strike a balance between a natural conversational feel and the practicality of the content. . This feature is now available in the ElevenLabs Reader iOS app, supporting 32 languages, providing users with a more convenient multilingual podcast creation experience.
Artificial intelligence startup ElevenLabs on Wednesday launched a new feature called GenFM that allows users to upload different types of content to generate multi-channel podcasts, similar to Google's NotebookLM.
This feature has been launched in the ElevenLabs Reader iOS app and supports 32 languages, including English, Hindi, Portuguese, Chinese, Spanish, French, German and Japanese.
When using GenFM, users can first upload a YouTube video, text, or document, and the application automatically selects two voices to create the podcast.
ElevenLabs offers more than a dozen sounds for users to choose from. As the app prepares the AI-generated podcast, users may see some interesting prompts, such as "Add some pauses" and "Insert some filler words." In a world where many tools help people eliminate the "ums" and "ahs," ElevenLabs has chosen to add a human touch to its AI-generated podcasts.
"We discussed how much to introduce human conversational filler or overlay sounds like 'um,' 'ah,' 'um hum,' laughter and breathing," Jack McDermott, head of mobile growth at ElevenLabs, said in an interview. . Our goal is to find the right balance between natural human conversation and practicality of content.”
He also points out that the best long-form podcasts tend to have fewer distractions and a more natural, deeper conversational flow as an experience they strive for, aiming to make audio storytelling more accessible across different voices and languages.
In the future, ElevenLabs plans to support more customization options and allow users to add multiple sources to create generative AI podcasts. In September, Google launched NotebookLM’s AI-generated conversation feature, and a month later added the ability for users to customize podcast output.
Earlier this month, ElevenLabs also announced that it would invest US$11 million in the Polish start-up ecosystem and open a research and development center in Warsaw to attract local AI talents. Meanwhile, the company is expanding into India, has hired a business leader and is building the team. Additionally, ElevenLabs has launched conversational AI agents for customers.
Highlight:
ElevenLabs launches GenFM function, which allows users to upload videos or text to generate multi-channel podcasts.
The feature automatically selects two voices and adds human-like filler words to enhance the natural conversation experience.
ElevenLabs plans to support more customization options in the future and expand operations in Poland and India.
All in all, ElevenLabs’ GenFM function provides a convenient and user-friendly AI solution for podcast production, and its future development direction is worth looking forward to. The company's aggressive global expansion strategy also heralds its ambitions in the field of artificial intelligence.