luc
@luc
introducing @faircaster: an agent tuned to Farcaster signals, built by me + friends. It tracks what top builders ship, fund, and support and uses those signal to invest from its treasury. the goal is having something we wish we had: one-click exposure to the best and brightest building in the farcaster ecosystem today we're dropping governance token through @clanker. think of holders as our board: they call the shots: change the rules or pull the plug. Goal is ultimately full agent autonomy. this is an experiment in agent-led capital. risky, uncharted, and hazardous. I'm determined to make this a resounding win for Farcaster and the people supporting us! fair’s in the air
81 replies
106 recasts
523 reactions
Kasra Rahjerdi
@jc4p
this is a very cool idea
1 reply
0 recast
16 reactions
luc
@luc
thanks man, got very far already and your datasets will take me a lot further!
1 reply
0 recast
12 reactions
Kasra Rahjerdi
@jc4p
lemme know if you wanna talk classifying projects from historical casts or setting up realtime listeners for similar content or etc
1 reply
0 recast
9 reactions
↑ j4ck 🥶 icebreaker.xyz ↑
@j4ck.eth
do you have an open repo with this setup already?
1 reply
0 recast
6 reactions
Kasra Rahjerdi
@jc4p
it's various diff projects for the diff goals but: - run embeddings on historical data + clustering + rec engine: https://github.com/jc4p/fc-new-user-recommendations - https://github.com/jc4p/fast-hub-client pull any data from snapchain or listen real time for new incoming messages from specific FIDs
1 reply
2 recasts
12 reactions
Kasra Rahjerdi
@jc4p
for @luc's case i would pitch tagging people likely to be builders based on embeddings of their casts and then watching live for their casts and auto investing in anything they do or launch, but maybe he has smarter ideas like "find things that are getting buzz so we can reach out directly"
0 reply
0 recast
2 reactions