Keccers push/pull strategy for early growth On social media 

There are 2 ways to use any social media 

A push: creating new top level material; the artist, a “creator”

A pull: pulling meaning or insight out of what others have shared; the reply guy 

At the start, for an unknown, it is usually not enough to do just one to gain a following. You have to do both, and be good at both. Imploring people to “create more interesting content” is generally true but usually neglects one side or the other. So a person who pushes out beautiful things but neglects to ever pull may feel insulted by this advice but it really means they need to thoughtfully engage with others as much as they thoughtfully create. It IS difficult objectively

We want to believe that the work should speak for itself. And you know yeah it should. But that is not the world we have created so learn to live with what we’ve got. Good work alone is not enough. It must be a part of a greater web of attention and relevance

The Farcaster formerly known as Kchamp. Books, beats, bodybuilding, bright ideas and beautiful things.

katherinechampagne.com

This is generally true for web2, but this is not "the world we have created" (as if FC has no control over its own algos & UI)

Solving public goods value capture. | /impact | book.abundance.id

Well, I think this is an issue with All social at scale. Not sure byo algo fixes the fact you have to give to get 

In a more local, smaller system or community it’s different

This is what I mean by the world we have made

its sound advice to interact with others on social (esp if the user is not just there to broadcast)

but not all algos are the same or prioritize the same things

the "visual" platforms (tiktok, YT, IG etc) don't really care for replyguying, so the growth strategy on those platforms actually favors broadcasting

FC, for whatever reason, decided to copy the X algo model - even tho we're not a closed source platform that monetizes thru ads.

we could instead choose any UI and algo we want

as an open graph protocol, it makes more sense to me to have an algo that promotes discovery and surfaces content that can benefit the network's growth (rather than an algo that maximizes engagement)

LLMs can help in the process but only if they're fed useful user data.

I'm thinking of something that's closer to Ponder polls - on Ponder you can make money if you predict correctly what *most other users* value.

In other words, your thought process is not "what do I like" but rather "what do I think most other people like"

if you apply that to content you can have a Reaction button on casts where you can have a scale for how much you think the community would value the cast. 

Then the closer your prediction matches community sentiment (ie how much the community actually values that content) the more you're rewarded

you can then also give a greater reward for those who make the first accurate prediction - so users would proactively look for valuable content (/apps/art/etc) for the network

then you can feed that data to an LLM so its better at surfacing more content that the community values

point is that it doesn't have to be based on attention at all, just what a network actually values

I believe engagement still a factor on IG/TT 

Curious how would you assess what is good for discovery if you were DWR. LLM rater on novelty? Use network analysis and show 1 or 2 removed + novelty? Track interests like TT/IG? 

Would be cool to built TT client on Farcaster rails instead of doing it as mini app