Varun Srinivasan pfp
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
207 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
Today's model uses 4 main features to "rank" a piece of content: 1. How often you like casts from the author 2. How often the author casts 3. How many interactions the cast has 4. How old the cast is
6 replies
4 recasts
55 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
If you like someones casts a lot, their stuff will move higher in your feed. This is the primary factor used to rank your feed. The model's predictive power (y-axis) increases as the number of likes increases (x-axis).
3 replies
1 recast
30 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
If you cast a lot, the predictive power begins to diminish. A high prediction score means that the model will bump the cast up in the feeds of people who like it and down in the feeds of people who wont.
1 reply
0 recast
13 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
Importantly, this doesn't mean you should cast less! Your casts still show up in feeds, the model is just less confident about changing its position in your feed. But if you are casting a lot of stuff that doesn't get engagement, trimming some of it might help the model get your content to the right users.
2 replies
1 recast
17 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
The third factor is engagement - the more likes a cast has, the more the predictive power is. Interestingly, the relationship isn't linear. It increases from 1 to 10, then decreases till 50 and then increases again. I'm not entirely sure why this happens yet, warrants more investigation.
3 replies
0 recast
15 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
The fourth important factor is time. This acts as a balance to the other factors - even if something is from someone you like, if its many hours or days old it's unlikely to show up at the top of your feed.
4 replies
0 recast
12 reactions
Varun Srinivasan pfp
Varun Srinivasan
@v
And that's a quick overview of how our model decides what casts to show you. We expect to be adding in more features over time like language, channel data etc which will make the model even better at finding stuff that is important to you.
4 replies
0 recast
12 reactions
Dutchyyy pfp
Dutchyyy
@dutchyyy
I appreciate the hard work and intention being put into this. Just can’t wrap my head around why the model of just following who you want to see, then seeing those posts in chronological order like when social media was healthy and inspiring keeps just not even being an option. Genuinely curious
1 reply
0 recast
0 reaction
Varun Srinivasan pfp
Varun Srinivasan
@v
Imagine that you follow 1,000 people. Now one of those users posts 100 messages in the last 10 seconds. What happens to your chron feed? Consider a less extreme example where 100 people post often, but the other 900 have a bunch of interesting people that post rarely. You'll just miss a lot of good stuff.
2 replies
0 recast
1 reaction
Dutchyyy pfp
Dutchyyy
@dutchyyy
In a world of low attention spans, yes I would assume most people wouldn’t scroll all the way to catch up. But we used to and it was fine & healthy. The loss of attention & patience is a symptom of algorithms. Beyond platform experiences, this reprogrammed how we navigate life offline. For attention currency?
0 reply
0 recast
0 reaction