I have used Youtube music as my main streaming application for some time, and it above all succeeded in avoiding the assault on AI features that came to other applications. However, thanks to YouTube Labs, it is now the balance sheet for a new AI feature that brings AI hosts to the musical listening experience.
It is no different from the DJ AI of Spotify, which presents a host generated by AI that speaks between songs to provide comments or context for future tracks. With the exception of Youtube music, there are two AI hosts who are transformed as part of a “radio program” called Beyond the Beat.
For some, it may seem boring; After all, who would like voices generated by the interrupting their listening experience? However, I found that I don’t really hate this new feature, and I hope it will remain, with some changes.
Dive into my favorite songs
Beyond the rhythm getting a path in your listening experience all the few songs, with the two hosts generated by AI, Mo and Flow, discussing the song that has just played and / or the song to come. The first time you activate the functionality, they will present themselves even after the first listening to let you know that they are, in fact, the AI, and that they can be wrong.
However, despite their AI, they actually have surprisingly a great chemistry and a personality of oozing. The conversations take place in a fairly transparent way between these two, and if you did not know that it was AI, you might think that it was real radio hosts.
Like Spotify’s Ai DJ, they will not intervene from time to time, and when they do, it will last less than a minute, generally about 40 seconds. Usually, conversations provide a context to a song, like Backstory or Trivia, and I actually find a large part of the comment quite fascinating.
The first time “Beyond the Beat” arrived, the hosts explained how Mariah Carey’s song “Yours” was initially intended for a duet with Justin Timberlake, but the contractual complications resulted in a solo version. This is something I had heard before, but it was surprising to hear it from the hosts.
At one point, they discussed “Jealous Type” of Doja Cat played VMA and how it was the first time of Kenny G on the prize ceremony program. It reminded me how shocked I was to see him open for his performance, so it was a great context to add to what I thought was a fantastic moment.
They also indicated how Kelly Rowland’s “commander” has deliberate grammatical errors and provided the story behind the name of the song “Lionheart (Fearless)” by Joel Corry and Tom Grennan. I was also surprised to learn how people apparently thought that the “Rocketeers” of the Far East copy a piece of Justin Bieber (despite the fact that it was produced first).
I also had a little laugh when the hosts discussed the way “Ain’T abandonment” by Craig David played a decisive role in his comeback, and at some point, Flow commented on how she remembers him “since the return”.
Of course, many of these facts are things that you could learn with a quick search on Google, but I am often not curious enough about what I listen to, at least that would justify research. This is probably a great reason why I appreciate the short interruptions, as I have the impression that it adds, not harmful, of my listening experience.
It’s not for everyone
Of course, I understand that it is not for everyone. If you listen to music, you sometimes want to get lost in the air without being interrupted by anecdotes generated by Ai-Généré.
Even a quick search on Reddit will reveal that people complaining about functionality, with some wondering why someone would like such a functionality, and others expressing frustration as to the inability to undress it permanently. It is compared to the DJ AI of Spotify and notes that he does not significantly improve the listening experience on the popular streaming application.
It should be noted that even if you cannot permanently make the hosts, you can repeat the functionality for an hour or a day. That said, this is just a test at this stage, so it is not yet widely available. Subscribers based in the United States will consciously have to register and allow experience to appear. Consequently, most users will not have to worry about a radio program generated by Ai-Aii rushing their musical experience.
With a little luck, all the folds will be pushed before a large potential version, and the hosts of the AI will be an opt-in function rather than an opt-out. It would also be great to see how Google can develop, similar to the way Spotify has widened the functionality to support vocal requests for songs (instead of simply choosing a song for you).
How to start
If you want to try this, Activating the AI hosts of YouTube Music is quite simple. However, it is currently only available for music / premium subscribers based in the United States, and it is not clear if or when it is deployed in more countries.
