Varun Srinivasan
@v
Warpcast now uses an ML model to order casts. If you have many unread casts, the model decides which ones go at the very top of your feed. It's main purpose is to learn your preferences and show you the most interesting stuff first. I'll go into some more details on how it works in the thread below.
49 replies
42 recasts
307 reactions
czar
@czar
Relying purely on the metadata of a cast - likes, author, frequency - instead of the content of the cast somehow doesn't seem complete. Wouldn't that work against surfacing relevant content and new users casting relevant content?
2 replies
0 recast
0 reaction
Varun Srinivasan
@v
This model isn't (yet) focussed on finding new stuff. It's mostly reordering the stuff that is already in your feed to present it in the best possible order so that the most interesting stuff is at the top. That seems to make feeds a lot more useful for people, both qualitatively and quantitatively.
2 replies
0 recast
1 reaction
czar
@czar
I don't doubt it makes the feed useful and you guys will keep iterating. Just airing a stream of thoughts. - You liked a cast and followed an author most likely because of the content of some of their casts. Think the current algo may be missing that key signal. It is looking at the outputs (likes/follows)...
1 reply
0 recast
0 reaction
czar
@czar
....not the inputs (content of the cast). The latter is likely also very hard. The challenge with relying on these outputs IMO is that, you get the good and the bad of the author. Though I like the casts of the author and followed the author, I do not necessarily want everything the author casts.
1 reply
0 recast
0 reaction
czar
@czar
One day they cast good stuff I wont want to miss, another day they cast something that is totally irrelevant. I was asking these questions the other day in the feed. I am now relating some of my comments here to that experience. https://warpcast.com/czar/0x2c5da17b https://warpcast.com/czar/0x9fa89357
0 reply
0 recast
0 reaction
