recommender systems

#topic #lab #AI #🌱

i've always wondered how there aren't self-hosted recommender systems trained on your RSS subscriptions, people you follow in social media, articles/videos you consume, etc.

i've been wanting to train models/host my own algorithms for many years, but still haven't made much progress on the development side, besides the designs i've been exploring with the omnichannel curation feed experiment.

i'm following the developments of a few projects in adjacent spaces, to see how/when they can facilitate that:

social graph/curation-oriented:

trails.social (ex-tweetscape)
sublime (ex-startupy)
hive.one

AI-oriented:

index network - 1-min demo: https://x.com/indexnetwork_/status/1786110169396429093
elicit for scientific papers -> great write-up: https://blog.elicit.com/living-documents-ai-ux/
(...)

holochain-based:

neighbourhoods for composable, distributed social media/coordination tools.
coasys

federated networks:
(...)

other efforts:

bluesky user-chosen algorithms

related: personal algorithms design