pull down to refresh

ZapRank is basically PageRank but with zaps rather than backlinks. We already do this with posts/comments. We don't rank stackers this way though.
Perhaps you're asking for something else though. Can you give more concrete details on the feature you're asking for?
Great, so I'm proposing a zap-rank for users, who have a general rank and then a topic-specific rank.
That is what the topic-bounded or contextual follow function would do, as a passive endorsement with a functional utility.
A bug-bear of mine, when Facebook intentionally broke the sub-dunbar number friend networks that users naturally converged on in order to increase the content pool of each user, and therefore engagement, that you couldn't follow people for particular subjects. That's when everyone siloed into groups, which served the same function but with a lot of negative downstream effects.
With domain authority a network can be more public, open, scalable, without introducing tons of noise and relying on engagement alone.
I'm basically asking for Twitter without the brain damage. I've been trying to build it for years but I don't have the skills myself and have built teams who create part of the picture without ever finding PMF.
reply
0 sats \ 5 replies \ @k00b 1 Oct
Each territory has their own zap graph and there's also a general zap graph.
We had personalization mixed in for awhile, but ditched it when we weren't seeing much difference between a global view and personalized view at our size.
Rather than following territories, what we do currently is have you follow all of them, then you can mute the territories you don't like.
reply
I get that, but I'm talking about following a person, and then seeing their posts in the territories you're interested in.
Run Pagerank across all the follows 54 times and it converges on a score between 0 and 1 for each person on each territory.
Bonus: cross territories with physical places (from Mapbox) and you get a rank order for users in particular countries and cities. Those geo-topic territories would be nested, so everything in Bitcoin-London is also in Bitcoin-UK.
I understand that you have limited resources, bandwidth, I'm just describing what I want to see in the world. I fucking hate Reddit, the subreddit architecture evolved with very little thought, it's a disaster. Topics and places should intersect so they are nested, so Physics-London is inside Science-World.
Then run Pagerank on the users and content and you can get rid of or significantly disempower moderators.
reply
100 sats \ 3 replies \ @k00b 1 Oct
I get that, but I'm talking about following a person, and then seeing their posts in the territories you're interested in.
That's what personalization did but rather than follows we used implicit signals like zaps.
I fucking hate Reddit, the subreddit architecture evolved with very little thought, it's a disaster.
I agree. We imagine a more tree-like structure for topics too, where territories are inside other territories, e.g. ~bitcoin territory has ~bitcoin/mining, ~bitcoin/knots, ~bitcoin/mempool, ~bitcoin/mining/solo, etc.
The geo idea is interesting. Bitcoin and geo are not great bedfellows but I can see the value.
All this stuff is planned - we have just been super focused, to everyone's chagrin, including ours, on noncustodial lightning stuff since we introduced territories. The lightning shuffle is almost over - at last - and we'll be able to work on fun stuff again.
reply
This is definitely the best out there as it is. I remember Steemit was one of the only blockchains that found some product market fit that was independent of Bitcoin's coattails or the unregulated securities market of the eth Ponzi platform, and this has the qualities of Steemit with integrity of bitcoin.
There is so much that this can do.
If in the future you interest territories with physical places, I'm building an LLM-based event aggregator which would compliment that, so you could tie digital public spaces with real world communities, which will make the whole tone much more positive and congenial compared to the competition.
reply
0 sats \ 1 reply \ @k00b 1 Oct
If in the future you interest territories with physical places, I'm building an LLM-based event aggregator which would compliment that, so you could tie digital public spaces with real world communities, which will make the whole tone much more positive and congenial compared to the competition.
That sounds neat! I can see what you mean by congenial.
Have you tried Orange Pill App? I believe they do a lot with geo
reply
Yeah, that sent me to an event that wasn't on. I'm general I don't like their model.
But I'll contact them when it's ready and working within my own non-bitcoin community.
I'm thinking of spinning this off as a SaaS for event data, like Mapbox for events and real time event changes. Never been possible pre-LLM, traditional parse engineering couldn't handle the level of fragmentation and lack of structure in event data.
reply