Topics
a la mode
AI
Amazon
Image Credits:Pocket FM
Apps
Biotech & Health
Climate
Image Credits:Pocket FM
Cloud Computing
Commerce
Crypto
go-ahead
EVs
Fintech
fund-raise
Gadgets
Gaming
Government & Policy
Hardware
layoff
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
societal
blank
Startups
TikTok
shipping
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Lightspeed Ventures - backed audio recording platformPocket FMannounced it has partner with voice - cloning companyElevenLabsto quickly convert text content , such as script , into audio series using AI .
Pocket FM , which raised$103 million in Series D financing in March , severalise TechCrunch at the time that it was already experimenting with the ability to convert text depicted object into audio usingElevenLabs ‘ technical school . Now , the India - base company has lucubrate the partnership to make the conversion putz uncommitted to all God Almighty over the next few week .
In the mental test phase , Pocket FM already produced 30,000 hours of audio series using ElevenLab ’s AI tech . With the unexampled rolling wave - out , the inauguration expects to treble its contentedness library of over 100,000 hours of audio capacity this yr . Pocket FM also said that during the experimental phase angle , the AI - powered puppet helped it edit out the cost of producing audio by 90 % .
Pocket FM ’s co - father and CTO Prateek Dixit told TechCrunch over a call that with this partnership , the company want to make it easier for author to exchange their written material into audio series .
“ We have over 250,000 writer ( including the ones on the fellowship ’s Pocket Novel indite plaform ) and this partnership minify the cost of setting up and recording audio for them , ” he suppose .
“ Even with a good stage set up of recording instrument and equipment , writer can produce roughly 30 minutes of high - quality audio frequency subject per day . With the AI pecker , this output can be 10 time more , ” he add .
Pocket FM has built a tool integrating ElevenLabs technical school , through which it is tender 50 voices for writer who want to change over their content . ElevenLabs ’ co - beginner Mati Staniszewski said that his company ’s tool translate the circumstance of the writing and infers emotion through the voice mechanically .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
“ Working with Pocket FM , we are deploying our young model that understand the music genre of writing and are emotionalism well , ” Staniszewski said .
Dixit noted that found on data from drug user ’ involution with this variety of content , the platform also plan to suggest voices that work well for author in a particular music genre .
Pocket FM is not the only audio serial platform experimenting with AI - powered tools . Google - back Kuku FMis using GPT-4 , Claude , BandLab and even ElevenLabs to help its writer with different stage of creation , including refining script , father thumbnails , adding sound effect and converting text into audio .
Kuku FM told TechCrunch that it is also experimenting with using visual generation tools such as Midjourney and Runway to create ads related to content .
Quality of content and impact on artists
The hope of AI - power tools is to engender more content faster , but that does n’t mean the content is good . Pocket FM ’s solution to aiding uncovering and surfacing quality content is make its uncovering algorithm advanced and experimenting with exploiter meshing .
“ If a author publishes an audio serial publication , we surface that content to a select bit of users and observe mesh metrics . If these metric unit are positive , we further circularize that , ” Dixit said .
Kuku FM said it is working with its quality control team to ensure only high - quality content is promoted on its app , even if Jehovah have used AI in the unconscious process .
“ We recognize the importance of having a human Quality Control squad at the heart of our decision - qualification when it occur to audio content yield . We have germinate a core team of Content Producers who have eminent possession & sanction on the aesthetic standards , ” the fellowship ’s atomic number 27 - foudner and CEO Lal Chand Bisu said .
Utilizing AI could lead to quicker results and a bigger content library for these platforms , but it will also reduce the roles of voiceover artists act with them . India ’s Association of Voiceover Artists ( AVA ) has expressed its concerns about AI hold over .
“ If AI takes over , we are stop . As voice artists , we call for to get some ordinance in shoes so that our livelihood is protect , ” Amarinder Singh Sodhi , the tie-up ’s general secretary , tell Indian publishing Scroll .
Sodi also told Scroll about incident where voiceover artists were called into the studio apartment to immortalize samples to train AI without receive their consent or inform them .
“ On an worked up grade , it scares me . By using AI , you are essentially debase the human experience of storytelling . You lose out on an emotional connection , ” Delhi - based voiceover artist Aditya Mattoo told TechCrunch .
He added that give access to premium voices to people who do n’t have the taste and skill to produce quality substance will conduce to the market place getting oversupply by bad content .
Voice creative person inother partsof the human race have also evoke concern about AI impacting their job . And despite play with some of the AI companies , they feel uncomfortable about their voices being altered .
When we ask about the shock of AI - power spokesperson propagation on Pocket FM , the company did n’t directly answer the doubtfulness . However , Dixit noted that engagement with AI - generated content in its experiment is “ as good as human voiceover output . ” Notably , the company is also working on engineering to incorporate multiple voice in one audio output .
Both Pocket FM and Kuku FM do n’t presently label their subject matter to indicate if AI has been used in the initiation mental process .
