Spotify acquires company that fishes unwanted content from podcasts
Spotify has acquired a company that specializes in developing a technique that can detect hate speech and unwanted content in audio. Spotify wants to use this technique to detect unwanted content in podcasts. According to Spotify, this should make the platform a lot safer.
Ban hate-mongering podcasts
We all know Spotify, of course. It is the largest music streaming service in the world and currently has about 180 million users. The service currently offers about 70 million music tracks and almost three million podcasts, of which our podcast is of course one. Spotify costs, apart from the student subscription, at least 9.99 euros per month. You can read more information, tips and other useful articles about the music streaming service here.
According to Spotify, among all those podcasts on service, there are also harmful and hate-mongering podcasts. They do not disclose what the service means by this, they talk about ‘hate-mongering’, ‘harmful’ and ‘abusive’ podcasts. The company wants to be able to better detect these podcasts and has therefore acquired the Irish company Kinzen. This company specializes in developing machine learning techniques to extract certain harmful fragments from audio fragments. According to Spotify, these advanced techniques can do this in different languages and Kinzen also still uses human expertise. According to the streaming service, all this will have to ensure that the platform is and remains safe.
Spotify – Music & Podcasts
Can you use Spotify for free? Spotify is a popular music app in the Netherlands that currently has 159 million users worldwide, of which 71..
More info
Previous collaboration
Spotify has previously worked with the company, but has now opted to acquire it in its entirety. The streaming service itself says the following about this: “We have had an impactful and collaborative collaboration with Kinzen for a long time and are a good team. Now, by working together as one, we can further improve our ability to detect and address unwanted content, and more importantly, do so in a way that takes better account of the local context.”