What is Corus trying to fix?
Spotify's Discover Weekly works for most people. For anyone who works with music professionally, or who cares deeply about what they listen to, the algorithm has one persistent failure: it shows you more of what you already know. Gabe Jacobs built Cymbal on the premise that social discovery was the answer. Now, with Corus, he is going further, stripping out the engine entirely and replacing it with people.
The pitch is simple and deliberately human-scale. Post what you are listening to, or what you think someone else should hear. Follow people whose taste you trust. No feed tuned to keep you scrolling. No "because you listened to" chain. No ad profile being assembled from your listening history.
"Discovery is not a problem that needs to be solved by a machine. It's a problem that needs to be solved by people who actually care."
How does it compare to Letterboxd?
The Letterboxd comparison is useful and slightly imprecise. Letterboxd works because film criticism has a culture of considered recommendation, and a film is a complete unit with reviews, lists and context attached. Music has the same potential and has historically done it better through physical culture (zines, record shops, crate digging, pirate radio). What digital platforms mostly killed is the trust component: you don't know who built the algorithm or what it is optimising for.
Corus is betting that if you remove the algorithm and replace it with visible human endorsement, the trust comes back. Five posts a day on the free tier is a soft constraint that forces a degree of curation. You cannot spam the feed with your entire listening history; you have to choose what you think is worth sharing.
Will it survive?
The honest answer is: the graveyard for human-curated discovery platforms is enormous. Last.fm, Hype Machine, Bandcamp's old discovery tools, Cymbal itself have all tried versions of this and found the economics hostile. A $2.99 per month subscription for music enthusiasts is not a lot; scaling it to a viable business is a serious challenge.
What Corus has is timing. There is genuine and growing scepticism about algorithmic curation in music communities, particularly among DJs, collectors and label A&R staff who have watched playlisting concentrate power in a handful of playlist editors at three or four streaming platforms. That audience is niche, but it is also the most vocal and most influential part of the online music conversation.
