I’ve spent the past few weeks quietly ashamed by how much I’m turning to Spotify’s “Discover weekly” playlists, which appear to be uncannily good at putting together two hours of music per week that I actually like and haven’t already listened to a million times. At least it turns out I’m not the only one: People have listened to more than 1 billion songs off of those playlists in just its first 10 weeks.

For better or for worse, it looks like Spotify’s lofty hopes for the playlists are paying off for now. How lofty? The company’s initial announcement boasted that the service would be like “having your best friend make you a personalised mixtape” every week.  And then they referenced High Fidelity.

The playlists, of course, can lack some of the humor and eclecticism of a gift from a real friend. But Spotify is pleased with how it’s going so far, anyway: 71 percent of listeners to the weekly playlists have saved at least one song from them to their own playlist, and 60 percent listen to at least five songs, according to Spotify.

The concept of algorithm as music curator is hardly new. Pandora — which just bought Ticketfly — is one of the older such services.  And Google’s Eric Schmidt recently said that he thinks companies looking for real humans to help them sift through all the noise of the body of music available on these streaming services are way behind the times — algorithms are the present and future, for him. The comment, as Billboard noted, was widely read as a criticism of Apple Music, which features the taste of real actual humans.

So how do Spotify’s playlists work? The Verge’s Ben Popper went deep on this in late September. There’s an algorithm there, but as it turns out (whispers), it’s people. Kind of.

The algorithms behind Discover Weekly finds users who have built playlists featuring the songs and artists you love. It then goes through songs that a number of your kindred spirits have added to playlists but you haven’t heard, knowing there is a good chance you might like them, too. Finally, it uses your taste profile to filter those findings by your areas of affinity and exploration. Because the playlist, that explicit act of curation, is both the source of the signal and the final output, the technique can achieve results far more interesting than run of the mill collaborative filtering.

In other words, the playlists you’re building yourself on Spotify are being harvested in order to help the company recommend music to other listeners, who then put the music on their own playlists, and so on, and so on.

Yes, it might be a little creepy. But a lot of people don’t seem to mind.