Lightspeed Ventures-backed audio platform Pocket FM has announced that it is partnering with voice cloning company ElevenLabs to quickly convert text content, such as script, into audio sequences using AI.
Pocket FM, which raised $103 million in Series D funding in March, told TechCrunch at the time that it was already experimenting with the ability to convert text content into audio using ElevenLabs’ technology. Now, the India-based company has extended the partnership to make the conversion tool available to all creators in the coming weeks.
In the test phase, Pocket FM already produced 30,000 hours of audio streams using ElevenLab’s AI technology. With the new launch, the startup expects to triple its content library to more than 100,000 hours of audio content this year. Pocket FM also said that during the pilot phase, the AI-powered tools helped it reduce audio production costs by 90%.
Pocket FM co-founder and CTO Prateek Dixit told TechCrunch after a call that with this partnership, the company wants to make it easier for writers to turn their writing into audio series.
“We have over 250,000 authors (including those on the company’s Pocket Novel writing platform) and this partnership lowers setup and audio recording costs for them,” he said.
“Even with a good setup of recording tools and equipment, writers can produce about 30 minutes of high-quality audio content per day. With AI tools, that output can be 10 times higher,” he added.
Pocket FM has created a tool that incorporates ElevenLabs technology, through which it offers 50 voices to writers who want to convert their content. ElevenLabs co-founder Mati Staniszewski said his company’s tool understands the context of writing and infers emotion through voice automatically.
“Working with Pocket FM, we’re developing our newest models that understand the type of writing and are better emotional,” Staniszewski said.
Dixit noted that based on data from user engagement with this type of content, the platform also plans to recommend voices that work well for writers in a particular genre.
Pocket FM isn’t the only platform in the audio lineup experimenting with AI-powered tools. Google-backed Kuku FM uses GPT-4, Claude, BandLab and even ElevenLabs to help its writers in various stages of creation, such as improving the script, creating thumbnails, adding sound effects and converting text to sound.
Kuku FM told TechCrunch that it is also experimenting with using visual production tools like Midjourney and Runway to create content-related ads.
Content quality and impact on artists
The promise of AI-powered tools is to create more content faster, but that doesn’t mean the content is good. Pocket FM’s answer to helping discover and display quality content is to make its discovery algorithm sophisticated and experiment with user engagement.
“If an author publishes an audio series, we present that content to a select number of users and observe the engagement metrics. If these readings are positive, we spread it further,” Dixit said.
The use of artificial intelligence could lead to faster results and a larger library of content for these platforms, but it will also reduce the roles of the voice artists working with them. Voiceover Artistes Association of India (AVA) has expressed its concerns about artificial intelligence taking over.
“If AI takes over, we’re done. As voice-over artistes, we need to make some regulations so that our livelihood is protected,” said Amarinder Singh Sodhi, general secretary of the association. said Indian publication Scroll.
Sodhi also told Scroll of incidents where voice artists were called into the studio to record samples to train the AI without getting their consent or informing them.
“On an emotional level, it scares me. By using artificial intelligence, you are essentially reducing the human experience of storytelling. You lose out on an emotional connection,” Delhi-based voice artist Aditya Mattoo told TechCrunch.
He added that giving access to premium voices to people who lack the taste and ability to produce quality content will lead to the market being flooded with bad content.
When we asked about the impact of AI voice generation on Pocket FM, the company did not immediately respond to the question. However, Dixit noted that engaging with AI-generated content in her experiments is “just as good as human voice production.” Specifically, the company is also working on technology to integrate multiple voices into one audio output.
Both Pocket FM and Kuku FM do not currently flag their content to indicate whether AI has been used in the creation process.
