Recommendation Systems

MK and

as a network scales, it is inevitable that you need a way to sort & rank content.

web 2.0 solved it by ranking for popularity but a better approach could be to have a minimal quality threshold but not to try to rank things above that - so the search is not maximization-driven and leaves more room for serendipity.

h/t ken stanley

sari added 10mo

Today most algorithms that recommend or suppress content act purely on the basis of inferred popularity. They look at how much time people spend engaging with a piece of content, and boost it to more people if the numbers look good. The content itself is almost purely a black box. Some algorithms try to classify content with tags like “food” or “fu

Nathan Baschez • Noise-Canceling Filters for the Internet are Coming

sari added 1y

In the rush to copy TikTok, many Western social networks have misread how easy it is to apply lessons of a very particular short video experience to social feeds built around other formats. If you’re Instagram Reels and your format and interface are a near carbon copy, then sure, applying the lessons of my three TikTok essays is straightforward. Bu... See more

How to Blow Up a Timeline

sari added 1y

I’ve never cared much about the presence or not of a blue check by a user’s name, but I do notice when tweets from people I follow make up a smaller and smaller percentage of my feed. It’s as if neighbors of years moved out from my block overnight, replaced by strangers who all came knocking on my front door carrying not a casserole but a tweetstor... See more

How to Blow Up a Timeline

sari added 1y

I wonder whether/why a For You feed is harder to tune in text-based mediums than video ones. Regardless, agree that the Twitter For You feed doesn't deliver.