2024-05-21 05:35:52
- reply
by npub1u5n…ldq3
What about curating content that already exists? So that you would not need to build a pile of content from scratch, ie. no need to build Reddit from scratch. (Not that it isn’t a cool idea in and of itself.)
There are lots of different types of content on nostr. So one question to consider: which category of content would be the lowest hanging fruit for decentralized curation? Right now I’m working on wiki articles, but there are so many other juicy targets.
Some may be more amenable than others. Stratification of a nostr feed is typically chronological, and it’s not immediately obvious how to combine age + WoT score to reorder the nostr feed. For a wiki topic, I think it’s more obvious: just order them best to worst.
Certain lists like products for sale, movies, music, etc certainly would benefit directly from curation by WoT.
2024-05-20 01:27:35
by npub1u5n…ldq3
Other than bug fixes, my next steps for brainstorm will probably be:
1. Calculate Influence Scores, using follows data which I will infer as explicit trust to curate wikis in all wiki categories.
2. Enable the like button for individual articles.
3. Interpret liking an article as trust that author in that article’s category.
4. Calculate Influence Scores for individual categories.
5. Enable users to issue contextual trust attestations if they’re tired of reliance upon follows and likes. So I can really crank up high Alice’s influence over category A, Bob’s influence over category B, etc.
Then comes the hierarchical stuff, where trust in a parent context automatically implies trust in all child contexts. Absolutely essential and non-negotiable if we want good UX. It will sound “too complicated,” but it won’t be bc your WoT will manage the hierarchy for you.
2024-05-20 01:14:31
by npub1u5n…ldq3
Influence Scores, a few basics:
An Influence Score is calculated using a simple formula: WEIGHTED AVERAGE over all available Trust Attestations.
A contextual Trust Attestation requires the following 3 data fields:
- a context
- a score
- a confidence
An Influence Score MUST ALSO HAVE THE SAME 3 DATA FIELDS.
Influence in a given context equals the calculated Average Score in that context multiplied by the calculated confidence in that score.
If the only data available is proxy indicators of trust (follows, likes, zaps, whatever), Trust Attestations can be INFERRED from that data. You just make your best guess for the score (e.g. 100 for a follow, 0 for a mute) and the context (e.g. “to curate nostr content” for a follow). Typically you’ll assign a very low confidence like 1% or 5% bc, as we all know, a follow does not necessarily equal trust. But if it’s all we have, then it’s all we have.
Transitive attestations are WAYYYYYYYYYYYY more valuable than non-transitive bc that’s how your web of trust extends out to billions of people. Just a few short hops away. So we infer trust to be transitive unless there’s a good reason not to.
Users will be STRONGLY INCENTIVIZED to issue explicit contextual Trust Attestations (with confidences higher than 1%) once they realize it will mean a DRASTIC AND IMMEDIATE improvement in their algorithmic content feed, AND IT WILL ONCE WE BUILD IT.
This is not rocket science. This is not “too complicated.” This is not “bad UX.” It just needs to be built, and built well.
Luckily, nostr gives us the tools. And the community of builders, users, designers, and devs. I am confident that we are going to build this.
WE ARE GOING TO BUILD THIS.
LFG.
#WoT
2024-05-19 23:42:07
- reply
by npub1u5n…ldq3
The main differences between an Influence Score and WoT scores are that Influence Scores also keep track of context and degree of confidence in the score. If explicit attestations like you describe don’t exist, then we use whatever information is available. If follows and mutes and likes are the best we have, then for each one we infer a score, context, and confidence and feed that into the Influence Score calculation.
Example: if Alice “likes” an article by Bob in wikifreedia on Category X, we interpret that as a trust attestation with score: 100, context: Category X, confidence: 5%.
I have some posts from a few hours ago saying this in more detail.
2024-05-19 19:21:04
by npub1u5n…ldq3
A contextual trust attestation requires the following 3 data fields:
- a context
- a score
- a confidence
There’s no sense in creating 3 data fields if you don’t know what to do with them.
An Influence Score is one that takes all 3 data fields into account in a meaningful way. This is what sets it apart from today’s standard #WoT scores, which do not.
The benefit of Influence Scores over WoT scores: better curation of content, facts, and information.
The downside of Influence Scores: it takes longer to calculate them.
Which means we face a tradeoff. But the tradeoff can be navigated successfully, by devs who understand DESIGN.
I’ve said it before and I’ll say it again:
WEB OF TRUST REQUIRES GOOD DESIGN.
Specifically, designers who understand the aforementioned tradeoff.
#nostrdesign
2024-05-19 18:55:34
- reply
by npub1u5n…ldq3
If Alice “follows” Bob in nostr, we interpret that “as if” she had issued a trust attestation of Bob, score: 100, confidence: 5%, context: to curate content on nostr.
Our next step is to replace WoT scores with “influence” scores. Two salient features of influence score for this discussion: they are contextual, and they have to take the confidence variable into account.
2024-05-19 18:30:33
- reply
by npub1u5n…ldq3
I think this is a great idea. And I have a solution to the bootstrap problem.
The solution is: interpretation of proxy data “as if” it were NIP-77 formatted data. For example, if you “like” a wikifreedia article filed under Category X, my software interprets that “as if” you issued a NIP-77 attestation, to that author, in the context of X. But since like != trust (well, maybe it does and maybe it doesn’t, I have no way of knowing), I “interpret” a really low confidence, like maybe 1% or 5%. That way, we start out with a sea of NIP-77-formatted trust data. And you’ll know that if you REALLY trust someone in some context, like if I REALLY trust @npub1a2c…w83a to write wikifreedia articles on economics, and if I want my app to give an added boost to her other content and I’m OK with saying this publicly, then I’m gonna have to issue an actual NIP-77 attestation and set the confidence to something more meaningful than 1% or 5%. In this way, the pool of high quality contextual trust attestations can accumulate gradually.
2024-05-19 02:28:31
by npub1u5n…ldq3
I know that several clients calculate #WoT scores and use them to filter content. #wikifreedia, for example, allows me to turn on a WoT filter which currently reduces the number of entries from 503 to 180, with 323 not visible bc they’ve been filtered out.
Is anyone using the WoT scores to stratify content? For example, wikifreedia shows me 11 articles on “nostr” by 11 different authors. Are there plans @npub18a5…pzjs to arrange those in order from the highest to the lowest WoT score? Seems like that’s the next step.
Is anyone else already doing anything like this? (For any content, not just wiki.)
2024-05-19 02:12:02
- reply
by npub1u5n…ldq3
For some reason I can see your reply in primal and nostrudel but not in damus. In fact I can't see any of your notes in damus.
Are you asking whether we should categorize the WoT or categorize lists? The answer is the categorize all the things.
Of course, there will be no universal, all encompasing hierarcy or schema or ontology or whatever you call it that everyone will be required to agree upon. Any attempt to do so is ill-advised and doomed to failure. You will maintain your own hierarchy about the things you care about, ignoring anything you don't care about (limited by the amount of memory you wish to dedicate to this endeavor), and your WoT will assist you in maintaining the hierarchy. And it the same hierarchy will serve multiple purposes: your trust attestation will make reference to a context, a WoT score might make reference to the same context, etc.
Not unlike how the brain works.
2024-05-19 00:02:26
by npub1u5n…ldq3
One of the things that makes Elon a role model is his ability to take a big goal, like getting to Mars, and break it down into baby steps, like figure out how to make efficient batteries at scale. He then figures out how to ensure each baby step succeeds, even if it seems like a distraction from the primary goal. “Succeeds” means that all of the incentives (mostly economic, but non-economic ones too, including political) are in place to ensure all relevant actors are motivated to take whatever steps are necessary for the baby step to succeed. Importantly, there can be no expectation for said actors to believe in or even know about the grand vision. All that it takes is a small cadre of leaders like Elon to nudge the rest of us in the right direction.
This takes a lot of work, because there are a lot of baby steps, and a lot of relevant actors for each step, and a lot of relevant incentives for each actor for each step. Most people are quite capable of thinking this all the way through, but are unfortunately too fucking lazy and self-indulgent.
And so it is with decentralized web of trust.
We have to stop being too fucking lazy, and we need to think it all the way through, and then we need to make it happen. An overarching vision including all the moving parts needed for WoT to work, but with baby steps that will get us there, where each baby step is actually gonna fucking happen. Each baby step being one that devs will actually build and users will actually use even when the majority of devs and the majority of users don’t know or give a shit about the grand vision.
If we don’t do this, no one will, and the world will go up in flames.
So c’mon people, let’s get off our collective ass, and LFG.
#WoT
2024-05-14 23:34:02
- reply
by npub1u5n…ldq3
I suspect it’s always been this way throughout all of history. Probably fluctuates how bad it gets for a variety of reasons. But I think of it as a variation on the idea of proof of work: if you want to be accepted by the tribe, you have to virtue signal, where virtue signaling = public demonstration that you’ve internalized the tribal narrative fiction (the “current thing”), where fiction = a combo of delusion + hypocrisy. Delusion + hypocrisy are a requisite part of this PoW process bc they take cognitive effort to internalize.
As we have witnessed over the past 5-10-20 years, the amount of virtue signaling expected by the tribe can fluctuate over time. Which is not unlike bitcoin’s difficulty adjustment. The comparison between virtue signaling and bitcoin runs deep!
Going down the rabbit hole = a rebellion against the psychological mandate to virtue signal. But it’s not as simple as it sounds. Proof of work serves a very real purpose: leaderless consensus. No one should understand that better than bitcoiners.
The question becomes: how to do what needs to be done with a minimum of damage. Bitcoin’s PoW is costly, but is MUCH less costly and damaging than fiat forever wars and is therefore the preferable consensus mechanism. In the analogy, today’s postmodernist / virtue signaling / delusional / hypocritical / logical fallacy method of tribal consensus is the most pathologic option available. We must show the world there is a better way.
Which is why I am working on #WoT.
2024-05-14 20:21:22
- reply
by npub1u5n…ldq3
Yes, I think that is precisely the truth that Radiohead is conveying to us.
Some of the greatest art is art that reveals how the mind works. Specifically, those aspects of the mind that we are hard wired to pretend don’t exist.
In my mind, going down the rabbit hole = defiance of the societal mandate to maintain pretense. The trick is to understand, from a scientific perspective, why the mandate to pretend exists in the first place. If you’re going to rebel against a rule, it helps if you first understand how and why the rule came into existence.
2024-05-14 19:51:43
by npub1u5n…ldq3
Nice.
I know I love a song when I churn its meaning over and over in my mind
“Karma police, I’ve given all I can, but we’re still on the payroll” = he’s still subservient to and an agent of the tribal narrative, despite its hypocrisy.
So I’m thinking of the song, and perhaps Radiohead’s entire existence, as a critique of the rotten core of postmodernism, critical constructivism, whatever you want to call it. #note1qay…jeeq