Spotify Tests Voice Translation Feature for Podcasts

Small figurines are seen in front of displayed Spotify logo in this illustration taken February 11, 2022. (Reuters)
Small figurines are seen in front of displayed Spotify logo in this illustration taken February 11, 2022. (Reuters)
TT

Spotify Tests Voice Translation Feature for Podcasts

Small figurines are seen in front of displayed Spotify logo in this illustration taken February 11, 2022. (Reuters)
Small figurines are seen in front of displayed Spotify logo in this illustration taken February 11, 2022. (Reuters)

Spotify Technology is testing an AI-powered feature that will translate podcasts from the likes of Dax Shepard and Lex Fridman to other languages, the audio-streaming company said on Monday.

The feature marks the latest attempt by the Swedish company to capitalize on generative artificial intelligence, the technology that has taken the world by storm after the rise of ChatGPT, to tap new users and boost revenue.

The translated versions, powered by Microsoft-backed OpenAI's newly released voice generation technology, would mimic the original speaker's style and will be more natural than traditional dubbing, Spotify said.

The company had also worked with other podcasters including Monica Padman, Bill Simmons and Steven Bartlett for the feature.

The voice translations would be available in languages including Spanish, French and German for a select number of catalog episodes and future episode releases, said Spotify, which could expand the audience of the shows.

Spotify has spent aggressively in the past few years on its podcast business, hoping the higher engagement levels offered by the format will bring in more advertisers.

But that drove up the company's operating expenditure, which surged at twice the speed of its revenue growth last year. Besides, businesses have also been dialing back spending on ads due to rising interest rates and high inflation.



AI Chatbot Startup Character.AI Launches New Calls Feature

Figurines with computers and smartphones are seen in front of the words "Artificial Intelligence AI" in this illustration taken, February 19, 2024. (Reuters)
Figurines with computers and smartphones are seen in front of the words "Artificial Intelligence AI" in this illustration taken, February 19, 2024. (Reuters)
TT

AI Chatbot Startup Character.AI Launches New Calls Feature

Figurines with computers and smartphones are seen in front of the words "Artificial Intelligence AI" in this illustration taken, February 19, 2024. (Reuters)
Figurines with computers and smartphones are seen in front of the words "Artificial Intelligence AI" in this illustration taken, February 19, 2024. (Reuters)

Artificial Intelligence chatbot startup Character.AI said on Thursday it has launched a calls feature that allows users to have voice conversations with their AI characters, ratcheting up efforts to compete with rivals including OpenAI.

AI-boom has prompted startups to release new features to their chatbots as competitors such as Microsoft-backed OpenAI, Google and Amazon.com-backed Anthropic are looking to gain market share by engaging new users.

The availability of free two-way voice calls on Character.AI's app has come following OpenAI's announcement on Tuesday that the ChatGPT maker's rollout of the advanced voice mode of its latest model GPT-4o is delayed by one month.

Character.AI, co-founded by Noam Shazeer, who helped invent "transformer" AI architecture while at Google that underpins systems like Gemini and ChatGPT, enables users to create customized AI companions with specific personalities and values.

The Menlo Park, California-based startup said the new calls feature prompted more than 20 million calls from over 3 million unique users in its early rollout, suggesting a strong engagement.

Character.AI launched Character Voice, a suite of features that allows users to hear characters speaking to them in one-on-one chats, for all users in March.

Alphabet's Google was in talks to invest hundreds of millions of dollars in Character.AI last year, two sources briefed on the matter had told Reuters.