Why Nostr? What is Njump?
2024-08-18 02:05:58

Greg on Nostr: Hello everyone, I've been reading through some of Nostr's ideologies, the problems it ...

Hello everyone,

I've been reading through some of Nostr's ideologies, the problems it solves, the problems it presents, etc. Among the problems that are solved:
- A censorship resistant communication network. This is without a doubt an extremely important aspect especially for third-world countries/communities under an authoritarian regime.
- Destruction of big-tech data silos. A decentralized network of relays which store information means that no big-tech company owns your data, which opens up a massive number of avenues which I'm posting to discuss
- There are of course others, like client interoperability, crypto (bitcoin) integration, user specific choice of which content they'd like to see, etc.

But there are many issues yet to be solved:
- With no centralized store of data, how do we guarantee data safety? In other words, how is it guaranteed that my post from 10 years ago will in fact still be on the relay(s) it was originally posted to?
- How are relay operators incentivized to store data? They can of course require a fee for data fetching/storage, but at least to my knowledge it will not be guaranteed to offset the electricity/hardware costs to run the relay.
- How is content recommended to others? With no central authority to provide an algorithm, the discovery issue on Nostr clients is very obviously prevalent

I think its important to address these issues. Nostr can geniunely change the world, but not if the adoption rate is as low as it is now. There have been proposals to fix these issues, but I very rarely see them discussed on Primal unfortunately.

I have some ideas of my own and would like to know what's already been fleshed out or what can be improved. If anyone knows of a blog/repo where this type of stuff is welcomed, please let me know.

First, data safety. There was a recent proposal called Replicatr, which not only defines the protocol for a relay but also the infrastructure of the network. It relies on the Internet Computer, which from my understanding is a collection of independent datacenters, of which anyone can join, that operate on top of the blockchain to ensure content distribution is sufficient and content remains available. This would solve the issue of data inconsistency (two users connected to relays seeing two different feeds), and data safety to ensure content does not go randomly missing.

The incentive to store data is a tough problem to solve but I think there are viable avenues. In the case a user is viewing a post provided from a specific relay and there is an advertisement, a small percentage of the ad revenue would go to the relay operator. This would require direct contact between the relay operator, the advertiser, and the client developer(s), and would probably rely on smart contracts to ensure proper payment.

Content recommendation I think is the biggest opportunity that Nostr has going for it, and could vastly modify user experiences on social media platforms. I was reading through some of the other protocols that have been proposed besides Nostr, and one is called Solid by MIT. It proposes that everyone has a POD (a Personal Online Datastore). This is almost like their online identity, and it contains all the info about that person (their information, their browsing history through certain clients, video games played, etc.). Of course this is sensitive information, but the point is that if the user owns their POD, they can distribute information only with their consent. This opens up so many doors that it would be hard to list all of them in one post, but there are some big ones worth mentioning. Jack Dorsey has recently mentioned that the problem is not free speech, but free will. Big tech with their large data silos decide what content you do or don't see, with really no input at all from your side. But what happens when that data isn't in a silo? The posts you want to view are publicly available across a huge network of relays. And most importantly the data about your interaction with those posts is something that you own. This means that you can apply your own algorithm. An algorithm that is more biased towards educational content, gaming content, content that is more or less recent, content you've watched before or not, or content that you aren't directly subscribed to but others connected to you are (social graph), etc.. Open sourced algorithms become a thing, and user feedback of these algorithms (upvotes, discussions, etc.) also becomes prevalent.

What could become possible if you are able to temporarily assume a recommendation feed? For instance, what if we are able to upload the feed that is recommended to us on our homepages (without any disclosure of your specific viewing history)? I can see a use case where users try their best to cater a recommendation feed through a careful viewing history, so that others can enjoy the interesting feed of content they have to offer. This could even evolve into a money-making opportunity.

Nostr is great for now to act as a censorship-resistant protocol but how do we bring everyone else over? I think it would be a promise of a feed that they control and have influence over, not just for their own free will but because it actually recommends interesting and relevant content.

There is little standing in the way of moving from: closed-source, limited customization, a monopoly on platforms that prevents small devs from making their own tweaks, censored speech; and into: open-source, a free-market in regards to clients because anyone can access the data, a free-market in regards to algorithms and recommendation feeds because we own our data, and free speech

The biggest issue I can see is the incentive for users/organizations to host relays/data centers for our use. They need a way to make money and in a relatively easy way. I would love to look at some of the ideas that people have dreamed up for this and/or other similar protocols, these are just my own. What are the other economic models that are more sustainable than just running ads on every page and post?
Author Public Key
npub1k9sx8s7hfmxmujwtutl2y4tjj6uz74rrxawsjh3dmn4myuh2hxvqzcgdk7