Musixmatch moves beyond songs to the spoken word. It has launched a new podcast platform in beta that, similar to its music service, allows users to follow an episode with lyrics generated by artificial intelligence and varied using a crowd-sourcing model. Musixmatch says the move to podcasts will help shows with discovery since the transcripts are not only more search engine friendly, but also make it easier for a show’s fans to share clips on social media.
“We believe this new experience will make audio accessible to a whole new group of customers who couldn’t previously enjoy it: people with hearing disabilities or those who aren’t fluent in the podcast’s spoken language,” the company says in a blog post. It says it will also create the most accessible podcast experience to date for the deaf and hard of hearing.
For the past year, the Musixmatch AI has transcribed, indexed and categorized the most listened to podcasts around the world and created a graph that ranks them according to what’s being talked about and who is participating in the discussion. The company says each episode is fully transcribed and dubbed, with speakers tagged in each paragraph.
“That way, podcasts are easier to scan, and listeners can jump to the parts that really interest them, and it’s easier for them to ‘read through’ the episodes,” it explains. “It’s like watching a movie with subtitles.”
The new Musixmatch podcast portal features transcripts for many of the biggest hit shows such as The Daily by The New York Times, Stuff You Should Know by iHeartRadio and Crime Junkie by Audiochuck. However, it does not include The Joe Rogan Experience, Armchair Expert, Call Her Daddy or other Spotify exclusives.
Musixmatch admits that the accuracy of the AI-generated transcripts is “okay but not perfect” as artificial intelligence alone is great but not enough to guarantee quality content. That’s why Musixmatch relies on an open-source environment that allows a community of podcast hosts and creators to edit and tweak the AI transcriptions using dedicated online software. If the AI system hasn’t already created a transcript for an episode, the community member can use the software to create one in about five minutes.
“Basically, our goal is to take ‘podcast captions’ to a whole new level of quality and ensure that they are universally adopted and accessible,” says the company. “The result of this joint work will be verified, time-synchronized transcriptions, provided with speakers and topics.”
The Musixmatch Podcasts experience is not limited to its own website. The company says publishers and podcasters who claim their show can take their episode transcripts and embed them into their own web channels and apps. Embedded transcriptions are SEO-friendly, so they can be easily found through a Google search.
The Financial Times has already used Musixmatch to work on their website, as part of what Business Development Director, Virginia Stagni says, her vision is to provide her audience with a useful learning tool through our podcast content.
“Accessing a new interaction with our audio through the Musixmatch products is an exciting opportunity to offer a 360-degree UX that not only touches on ‘hearing’, but offers a holistic digital experience of our brand,” said Stagni. “We are a hub of innovation here at the Financial Times: Musixmatch is a disruptive partner to work with to experiment and take our products to the next level.”
Italy-based Musixmatch is already working on the music side, transcribing lyrics and running streaming services including Apple Music, Amazon Music, Spotify, Tidal and YouTube Music. Chief Product Officer Marco Paglia says they were also able to use their AI algorithm to analyze song lyrics to understand a song’s mood, themes and subject matter.
“We are incredibly excited to leverage our 12+ years of music metadata expertise to support the larger audio ecosystem, starting with podcasts. While this is a very important milestone for us, this is just the beginning,” says Paglia. “Our vision is that in the near future, you can easily ask for the latest and greatest on any topic from your car, and listen to that content in your own language, for the duration you want, and with the ability to hear more effortlessly insights. The line between what is written and what is recorded blurs in both directions.”