During Spotify’s Q2 earnings call this morning, CEO Daniel Ek teased a few ways the streaming service could introduce additional AI-powered functionality. Ek discussed how AI can be used to create more personalized experiences, condense podcasts, and generate ads.
Earlier this year, the company launched a DJ feature that delivers a curated selection of music alongside AI-powered spoken commentary about the songs and artists you like. Ek says consumers can expect similar AI-powered features in the future that aim to contextualize and personalize content on the streaming service.
“DJ is a phenomenal product,” Ek said during the interview. “It’s probably one of my personal favorites from the last few years that we’ve developed, and we’ve seen really strong consumer interactions with it. And that’s just talking about the ability for us to contextualize and personalize all the great content we have on the Spotify platform. So I think you’re going to see a lot more of that where we can contextualize and personalize content across the platform to make it more accessible.
One of the ways Spotify can use generative AI is to use it to summarize what podcasts are about, as it can be somewhat difficult to get into new podcasts. Ek said this could make it easier to market new podcasts to consumers, which in turn would lead to more engagement and more growth for creators.
Ek says another way AI can be used to make Spotify more efficient is through AI-generated audio ads.
“By using generative AI and our tools here, I think you’ll be able to see that we can significantly reduce the cost advertisers need to develop new ad formats,” said Ek. “And of course that means that as an advertiser, instead of having one ad, you can imagine having thousands of them and testing them on the Spotify networks, things that you could easily do today with text, but haven’t been able to do through video or in audio.”
Ek’s comments come as Spotify seeks a patent for an AI-powered “text-to-speech synthesis” system. The patent was published on July 20 and filed in February. The technology can take text and convert it into human-like speech audio that incorporates emotion and intent. The system can create realistic speech that can convey emotions such as anger, happiness or sadness, along with intentions such as sarcasm. It is also capable of doing this in a whispering or yelling manner and with accents.
The patent indicates that Spotify wants to go beyond its DJ feature, which simply says a few AI-powered sentences between songs. The text-to-speech synthesis system can be used for things like narrating audiobooks using AI in a way that sounds natural. It’s worth noting that Apple launched AI-powered narration for select titles on Apple Books early this year.
The patent filing and Ek’s comments come as Spotify has invested in AI speech technology. Last year, the streaming service acquired Sonantic, a London-based startup that built an AI engine to create realistic-sounding human voices from text, for an undisclosed sum. Spotify leveraged the acquisition to power its AI DJ feature.
Spotify’s second quarter results showed that there are currently 220 million people worldwide paying for a Spotify subscription – that number is up 17% year over year. In total, Spotify now has 551 million monthly active users. The company reported nearly €3.2 billion ($3.5 billion at the current exchange rate) in revenue for its most recent quarter. Sales are up 11% year over year. However, Spotify also reported €247 million ($274 million) in operating loss.
Spotify also announced a price increase for its premium subscriptions yesterday. In the US, an individual premium plan now costs $10.99 per month instead of $9.99. The duo plan costs $14.99 instead of $12.99 and the family plan costs $16.99 instead of $15.99.