Will Nostr put people back in control of their data?
There is an incredible momentum around Nostr protocol. I think the key innovation that Nostr brings is letting people own their identity. In this post, I’m gonna explore the implications of this fundamental shift, and where it could lead.
Most platforms control your data, including your reputation: Twitter controls your tweets and followers, YouTube controls your videos, Instagram controls your photos, etc. That means they can pick who is allowed to post, and what content gets shared. They also use your data to make their product better and of course to make money from ads. These are valuable products, since nobody is forcing us to use them and there are billions of people using them, so I’m not saying that these platforms should do it for free.
However, it’s really hard to compete with them. Imagine you want to create a competetive product. You don’t have access to all that data, and thus you can’t really use any machine learning, which is a very powerful tool. In adition, the exit cost for the users is very high, because they can’t take their identity with them. Their content, connections, and reputation, it all gets lost if they decide to switch to your platform. And so these large platforms are not really forced to innovate. They can play it safe since it’s very unlikely anyone will disrupt them. Well, until someone does.
How does Nostr solve any of those problems? It’s an open protocol, which means anyone can participate. In Nostr world, anyone can build a an app or a relay that speaks the protocol, and join the ecosystem. Most importantly, the users control their identity. This means, a user can switch to a different app or relay, without asking for permission, and they can keep their identity. Yes, they don’t need to sacrifice loosing all their followers, or their content.
For instance, I’m using three Nostr client apps right now. One is better for editing profiles and managing followers, while the other one has nicer UI for reading posts from the people I follow. The third one shows the global feed. My identity (private key) is the same and so my followers don’t need to worry what client I’m using. When someone builds better UX, I can easily switch. The same with relays. If a relay doesn’t scale well, or starts censoring my posts, I can easily switch to a different one, or even start one myself. I don’t need to donate my data to a single platform.
If this is so transformative, why did it not happen long time ago? I’m not sure, but it seems like a few important pieces recently clicked together:
-
Nostr is good protocol. It’s simple and yet flexible. The identity is owned by end-users. There were many iterations before (ie. Mastodon), but each suffered from some fundamental issue.
-
Bitcoin/Lightning brings incentives. I believe this is why many projects like Tor (or even internet itself) couldn’t really scale up without centralization. Nobody would run good scalable relay for free. I think ads was actually a decent solution for people to pay for products indirectly with their attention. Lightning is perfect solution for this. Trustless, border-less, digital.
-
Trust in large centralized systems (ie. corporations, governments) seem to be at all-time low, so even non-technical people might be willing to try something different.
Right now, most of the momentum is around censorship-resistant broadcasting of public messages. Basically, a replacement for Twitter or Reddit, and maybe blogging. This seems like the right start, but I think it could go much further. Here are just some ideas:
-
Once the relays scale to handle large data, this could apply to podcasts, videos, music, etc.
-
Once the payments get integrated, this could apply to markets: Uber/Lyft, DoorDash/Instacart, general jobs and projects, coin joins, LN liquidity management, etc.
-
Once the relays and clients can handle real-time data, this could enable various collaborations: Github++, Figma, or even composing music.
-
Eventually, people could collaborate towards shared algorithms: share data on traffic to create better maps experience, weather forecast, or solving health problems.
The last category – essentially research, where anyone can access the data and compete – will require solving anonymity and privacy, but I will leave that for a different post. Let’s disrupt Twitter first ;-)