Show HN: A social feed with no algo where communities decide what gets seen(veridonia.com)

3 pointsby smnkgv7 hours ago2 comments

matzalazar4 hours ago
Love the decentralized and open source approach. I'm building Exogram (Django/AGPL-3.0) to solve the same 'engagement' problem, but using semantic similarity of book highlights instead of voting.
No popularity metrics, just readers connecting through ideas. Would love to exchange 'failure mode' notes!
Repo: https://github.com/matzalazar/exogram
- smnkgv4 hours ago
  Good to see a fellow anti-engager! If you send me an invite, I will definitely give it a try. As for the "readers connecting by using semantic similarity of books" - That is a great idea. I wonder, how have you addressed the "formation of echo chambers" problem? Or is it not the goal of this particular project (now/ever)?
  - matzalazar3 hours ago
    Actually, the invitation tree (with its user-defined depth levels) is our primary defense for privacy and trust, ensuring you only interact with people within a certain 'social distance'.
    To counter the echo chamber effect, we rely on the semantic engine. Since Exogram connects you through shared ideas across different books, it often surfaces perspectives from outside your immediate social circle. It prioritizes intellectual affinity over social or popular consensus.
    Looking forward to seeing you on the waitlist! :)
smnkgv7 hours ago
a few years ago, when I was trying to "return back" to social media, I realised there was kind of nowhere to go. All social feeds were either engagement-driven and wanted too much of my attention at the cost of content quality, or chronological, where the burden of sorting was on me, or editorial, where there were fixed elites doing all the sorting/filtering.
This is our attempt at making a feed that answers a more human-centric question. Instead of collecting likes, upvotes, views, etc. to answer the question of "What is getting the most traction today", or "what is most popular today" - it does none of that and just lets users decide what gets published (and, consequently, seen by everyone) through randomised majority voting. Basically, it tries to answer the question: "What does each community as a whole decide is worth attention today?"
Answering this question was not easy. After trial and error, we can reliably say that it does work mathematically, on simulation, and "on the paper". But it still remains to be seen whether there are other failure modes we did not think of, such as psychological, game theory-based, etc. After all, as far as we're aware, such a design hasn't been tried before, and there may be solid reasons for why - please help us see them!
on this experimental note - I hope you'll like it, and I am here to answer any questions you may have! Cheers!