Necessary but not sufficient imo
Discussion
What do you think it's missing?
Maybe augmented slightly with some web of trust componenet?
I think PoW is necessary to tilt the balance of power toward defenders in the spam war. But there's still the question of what is "interesting" content. WoT is probably the best thing we have right now, i.e. the "note that was upvoted by trusted user is considered popular" approach. But even in a web of trust, the long term solution can't just be total number of followers (even when considering followers of followers of followers) because it's trivially easy now to create an army of fake people. A determined adversary could boost their own signal. On a fundamental level there needs to be a source of scarcity to ground the web of trust in finite reality (like how btc uses the energy conservation principle). The way I see it, scarcity in a social context can be:
1) A hand-picked list of "trusted root nodes" of the web of trust (so not actually decentralized) or,
2) Derived from something else that is scarce (so probably money)
This is my current understanding — by no means do I think like I have this all figured out!
When applying a Web of Trust difference between a follow and a follower is critical. A Web of Trust based on followers will never work, at least not in a decentralized way, for the reasons you mentioned.
I like how you are thinking of it in terms of a fundamental source of scarcity. With a Web of Trust based on follows that scarcity starts with the user's follow list, everything else is derived from that input.
The nice thing when everything is derived from that 1 input is that any problem in the output can be solved with a simple 'Unfollow'.
> With a Web of Trust based on follows that scarcity starts with the user's follow list
I agree and I think that's a great way to think about it — but obv this only works for users that are already following people. The question of how to make a "general" trending feed (to show by default to non-signed-in users) seems like a different, separate thing.
A general trending feed is not 100% mandatory (like it's totally possible that you just have to start following some people to see new content, nostr will keep functioning) but I think that it could really help nostr for two reasons:
1) In the short term — faster growth. Lowering the barrier to adoption by showing people interesting stuff immediately.
2) In the long term — avoiding stagnation by not having followers be a prerequisite for visibility/virality.
I think it was George Hotz that said something like "algorithms promote class mobility", which I agree with. If you are some anon poster who is totally unknown but you have really amazing content it would be optimal to surface that content so that person can actually gain followers. Tiktok works because their algo identifies "underpriced" stuff and pushes it up. People are looking for opportunity and it seems they're willing to engage with dystopian systems to get it, so I think the realistic move is to work on a credibly non-dystopian alt version of that opportunity machine (that works even better in edge cases where you'd normally run up against censorship)
> The question of how to make a "general" trending feed (to show by default to non-signed-in users) seems like a different, separate thing.
Yeah general/trending is a separate thing and I agree it will help.
> If you are some anon poster who is totally unknown but you have really amazing content it would be optimal to surface that content so that person can actually gain followers.
The surfacing problem is an interesting one, I think it fundamentally comes down to the question: Did you not like this post because you didn't like it or because you didn't see it?
After thinking about it a bit more it seems to me most solutions lead to some centralized recommendation service, but one thing that might be feasible is for a client to keep track of your likes and people with similar likes in your WoT, ie customers who bought this also bought... but for likes from your follows.
It would solve 2 problems: show something you probably like that you would have otherwise missed and the anon gained a like or maybe a follow that would have otherwise never been seen.