The Internet needs user-owned identity and an associated open data layer. Both of which are interoperable with every app you use.
Imagine if you were able to easily export your identity, social graph, and reputation from Twitter to Instagram. And then to Spotify, Craigslist, Amazon, Yelp, Uber, Reddit, or any other site. You’d be able to quickly find new content or products/services recommended by your friends or people you trust. And brand new experiences could emerge like a dating or recruiting app that recommends the perfect person for you based on mashing up previously siloed data from someone’s Twitter follows, Goodreads posts, and Spotify streams.
This would not only create a superior product experience for almost everything, but would also encourage people to invest more time and energy building their identity – a piece of Internet property they actually own vs. one they’re sharecropping for a feudal lord, set to end up in the growing graveyard of inactive Facebook or YouTube profiles created circa 2005. Companies like Substack already recognize the value of ownership by allowing users to bring their own domain and retain their email subscriber lists. Human readable identity would be just a step further.
User-owned identity would also solve the cold start problem that every new app faces. Watching people build their social graphs from scratch yet again with new apps like Airchat (just like Clubhouse before) is a powerful reminder of how difficult it is to bootstrap critical mass and how little experimentation happens as a result. It reminds me a bit of the ‘05 era when every startup had to set up their own server – a tall and expensive task. Once any developer could offload this work to AWS, a massive wave of experimentation and innovation followed. Imagine the even larger Cambrian explosion that would occur if developers no longer needed to run a server at all but could still access the entire Internet of structured data and allow any user to easily bring their identity & social graph to a new app. I believe we’d see orders of magnitude more experimentation allowing an exciting new long tail of Internet experiences to flourish. The rapidly falling cost of AI-driven software development would only accelerate this trend, leading to even more custom and niche apps. The perfect conditions for rewilding the web 🙂
The major question is who would issue such an identity and data standard? And how? One obvious answer would be the major tech platforms that already own all of the users. And indeed they do provide a very limited version of this today (i.e. login) via the Oauth protocol. You can log into many apps with Google, Meta sites, Twitter, and Apple. But this protocol – designed in the ‘00s for big tech companies – is overly complex and bloated, making it challenging for indie devs to integrate. And much more importantly, it’s heavily permissioned – the big tech companies severely limit what data they’ll share beyond simple authentication. Some of these companies (e.g. Twitter and Facebook) used to offer much richer APIs that provided social graph data. But as we’ve seen play out over the last decade, what they share (and for how much) is subject to change on a dime. Anyone trying to build a business on top of Twitter or Facebook’s API has been consistently and sorely disappointed. It seems the root problem with conventional identity is all the trust that’s required to make it work 😉
One solution would be to regulate these companies into sharing identity and other data according to an open API standard. But this solution is both incomplete and tenuous. Regulation in the age of the Internet has generally proven anachronistic, arbitrary, slow, and imprecise – like trying to regulate early cars as if they were just faster horses and buggies. Not to mention it’s unlikely to happen and largely unenforceable in competing jurisdictions.
A much better solution would be to build an open identity/data protocol for the Internet which will not be shut down because it CANNOT be shut down. I’m not the first to argue for protocols over platforms, but I do have high conviction that a winning protocol has already emerged: Nostr.
Nostr is a radically simple standard for identity and data transmission with three key features:
That’s it: signed JSON shared across a free market of relays. I’ve been very bullish on Nostr for over a year, ever since a December ‘22 tweet from Jack Dorsey brought in a wave of hundreds of thousands of individuals interested in open, interoperable social media alternatives. Since then, my conviction has only grown. Why?
Nostr is ridiculously simple. It’s likely impossible to create a standard for generalized identity/messaging simpler than Nostr. And simplicity reduces the barrier to developer adoption and experimentation, unleashing the full power of evolution online. In the case of foundational technologies like Internet protocols (e.g. HTTP/SMTP/Bitcoin) or app programming languages (e.g. Javascript/Python), I’ll generally bet on the simplest solution.
Users own their identity. Unlike competing protocols like ActivityPub where identity is created on and tied to a single server the user doesn’t control, Nostr’s public/private key pair is created client side, stored by the user (or an app they use), and travels easily between all Nostr apps.
Nostr is extensible. As a result of this simplicity, Nostr is extremely extensible. Unlike competing protocols like Bluesky’s AT protocol which is purpose built for social media, Nostr can be used to build anything. Literally any app you can imagine. Marketplaces, Reddit clones, music players, and entirely new categories of online experiences.
A committed community of builders. I believe that Nostr has already reached the necessary critical mass of tens of thousands of weekly active users, a disproportionately high percentage of whom are talented and ideologically committed developers obsessed with the protocol. Dozens of these devs are funded indefinitely to pursue their own visions, each of which strengthens Nostr as a whole.
One potential criticism is that most of these users and developers are also committed Bitcoiners. The Nostr protocol emerged from Lightning devs who wanted to create a standard for open markets similar to OpenBazaar without unnecessary “blockchain” technology or scammy “crypto” tokens. Instead, they sought a solution that shared Bitcoin’s ethos of simplicity, permissionless identity creation, and bazaar style development. I view this overlap in community and ideology as a major strength. It means that Nostr’s core builders and users are in it for the long haul. They see Nostr, like Bitcoin, as a critical technology for the freedom and flourishing of the Internet and humanity at large. It also means that Nostr inherits – though is not required to use – instant, nearly free, and permissionless bitcoin payments via the Lightning Network. Indeed, micropayments and tips (zaps) have been the first truly novel features on Nostr social media clients like Primal and Damus. As of mid April 2024, we’ve seen >2.8M zaps totalling >1.5B sats (~$1M USD) across Nostr.
All of these conditions together create the recipe for super network effects. Each new Nostr app inherits hundreds of thousands of potential users with existing identities from other Nostr apps. And any users that new app recruits – whether it remains niche or goes mainstream – brings those users not only to that app, but to all other apps built on Nostr. This is the Nostr flywheel – a super network effect at the protocol level, instead of the application level.
Many Nostr users – including Jack Dorsey, likely the protocol’s most committed advocate and funder – see the long tail of microapps enabled by this super network effect as the killer feature that will ultimately propel Nostr to victory. But which of these microapps or novel use cases will drive the next wave of Nostr adoption remains an open question.
I obviously don’t have a crystal ball, but I’ll share what I see as likely next steps. My general thesis is that Nostr should – and will – lean into using its emerging Web of Trust to solve discovery, identity, and reputation problems for its existing core of ideologically committed Bitcoiners, who are willing to suffer through the growing pains of experimental tech and novel user experience. Nostr will then graduate to solving similar identity problems for more general tech early adopters, and ultimately, the mainstream.
Social Payments
I believe the next major Nostr use case already exists: social payments. One of the biggest problems with legacy payments is the lack of interoperability, especially internationally. If I’m a Cash App user, it’s impossible to send a payment to a friend in Brazil using Nubank. Hell, I can’t even send a payment to a friend in the US using Venmo.
Using Nostr identity for recipient discovery – and Nostr Wallet Connect, a simple Nostr based protocol for coordinating Lightning Network payments – I can now send a low cost payment of any size to anyone with a Lightning address on Nostr. This is a huge deal that not enough people realize! I can now get a friend registered with a Nostr client/Lightning wallet like Primal in a matter of minutes and send them cheap payments globally! 🤯This experience is already >10x better than closed payment apps like Cash App, Venmo, or Nubank which cannot send payments internationally. Here’s a video of me sending an international payment to my buddy DK in a matter of seconds!
There’s already >250k users with a Lightning address who have sent ~$1M USD worth of payments already. And it’s not just one client. I can send a payment with Primal and receive it with Mutiny. Or vice versa. Other early adopters like Alby, which created Nostr Wallet Connect, and Prism are already adopting the same standard. And as larger companies offering Lightning payments begin to understand the superior user experience of Nostr identity, adoption will explode. I expect Strike, River, Lightspark, and Cash App to follow suit.
Primal also recently released another groundbreaking feature – Top Zaps – which displays the identities of the top zappers for each note. This feature is already spurring a new wave of social payments and will have major implications for search and discovery, which we’ll discuss later.
Bitcoin Adjacent Discovery & Reputation
In order to build an economy based on p2p bitcoin payments, we also need p2p solutions for market reputation and trust, which are currently provided by centralized intermediaries. I tend to agree that the “web3” vision of “smart contracts” doesn’t make much sense, as self-executing contracts only solve the problem of settlement, but do little to solve larger problems like breach of contract. For any sort of real world recourse, we’re still likely to use reputation systems. And this is where Nostr really shines. Here are a few Bitcoin related decentralized marketplace use cases I’m excited about.
Ecash Mints: Bitcoin backed Chaumian ecash is having a moment. Recent advances with Fedimint & Cashu and their interoperability with Bitcoin and the Lightning Network are creating a huge amount of developer interest. As the tools improve, it will quickly become easy for any organization or individual to launch their own mint and create Bitcoin backed tokens or currencies (including 1:1 bitcoin and competing stablecoins). Since a mint is effectively a bank, these advances are likely to precipitate bitcoin backed free banking on steroids. In this world, the most important factor for selecting a mint (and associated ecash) is trust. And Nostr reputation solves this problem beautifully – users can easily select mints based on the ratings of their friends or organizations they trust. And indeed it’s already happening! Mutiny and bitcoinmints.com are already surfacing mints via Nostr login and a specific user’s Web of Trust. The table is set for this use case to accelerate quickly.
DLC Oracles: As tooling for Discreet Log Contracts (DLCs) improves, I expect to see huge new markets for non-custodial Bitcoin lending, derivatives, synthetic stablecoins, and gambling/prediction markets of all types. Think BlockFi and Bitmex that never take your keys (e.g. Lava, Atomic, or 10101). The missing ingredient, of course, is in finding an oracle you trust to publish a price feed or event. Another problem Nostr discovery & reputation are perfectly suited to solve!
Local Bitcoins: As many of the large p2p Bitcoin trading sites have shut down recently, there’s a huge opportunity for Nostr based apps to rise and replace them with Nostr discovery & reputation for buyers and sellers. Different apps with a shared standard for publishing and discovering offers should lead to a shared, global order book for bitcoin <> fiat trades! Mostro is an early mover here. I expect many more to adopt the standard soon.
Early Tech Adopter Discovery & Reputation
After solving problems for the early Bitcoin majority, I believe Nostr will begin solving identity, reputation, and discovery problems for early tech adopters more generally. Especially for groups with at least some ideological overlap with Bitcoin and Nostr – e.g. open source enthusiasts, Hacker News readers, and the e/acc community. These waves, of course, won’t be perfectly sequential. Indeed, proofs of concept for many of the examples below are already live! But I expect they’ll really pop after the Bitcoin adjacent services reach critical mass.
Open Source AI: As open source models get cheaper and easier to fine tune and run, I expect the number of agents to go exponential, with a fascinating long tail customized for every niche and use case imaginable. As I detailed in my 2023 AI thesis, I believe these agents will eventually cooperate via Nostr messages and Lightning payments. Agent discovery/reputation and payment coordination are major factors that could kickstart this trend.
While OpenAI and other large tech companies will aim to keep developers within their closed wall app stores, there will remain a large contingent of open source developers who are unable or unwilling to publish apps that use and comply with model restraints or company regulations. Perhaps because the developer wants to use information OpenAI has censored or because they live in a country without Stripe payout or because they just don’t want to pay the app store cut. These developers will prefer to use open source models and publish their agents or services directly to the web.
The problems of how to find and trust these agents will, of course, quickly arise. Nostr’s simplicity and Web of Trust make it well suited to solve discovery and reputation. And using Nostr as the standard will also provide an easy way to coordinate open, programmable micropayments and easy data permissioning (an unsolved problem that’s starting to attract mainstream investment). Open Agents is building this vision to a tee. And Cascadr already offers several Nostr-published/Lightning gated AI microservices. I expect many more to follow suit.
APIs: Adjacent to open source AI is the open source software and services that those open agents will consume and integrate. Nostr once again represents the perfect layer for discovering and selecting which services to use. A generalizable version of this idea was originally created by Pablof7z as a spec for the Data Vending Machine (DVM). This idea is simple but profound. Any human (or bot) can call a value-gated API with some data input + payment and receive some newly computed data back. All announced, delivered, and even composed over Nostr. TaskTiger already offers an early marketplace for DVMs and unleashed.chat, which is similar to ChatGPT with open source models, will offer its LLMs access to DVMs soon. Cascadr’s marketplace for microservices uses a similar Nostr based standard. Together, these projects are building the foundation for an open, composable, and machine payable web. Or as Stu Bowman says, an API for market intelligence.
Open Source Software Development: Perhaps the most fundamental problem that Nostr can solve for early tech adopters is the process of software creation itself. Today most developers use Git, a distributed and open source version control system, for building new software. Most FOSS projects, however, also use centralized walled gardens on top of Git (primarily Microsoft’s Github) for hosting and collaboration. While this has not yet proven problematic, we’re only one corporate decision away from censoring a major repo like Bitcoin Core.
Fortunately, it’s possible to build a decentralized alternative on top of Nostr. This version should, in fact, become much better than Github over time because of Nostr identity and reputation. It will become easier than ever to discover new projects rated highly by devs you’re likely to respect. It seems no coincidence that Jack offered a bounty for a Nostr Github alternative before anything else. Developers are likely the first group to really grasp the gravity of the problem, the benefit of the solution, and possess the technical acumen/patience to actually implement it. It seems a lot of progress has been made recently and gitworkshop.dev seems particularly promising.
Apps: Nostr can also solve discovery and reputation for end user apps as well. Nostr App Link already curates and recommends Nostr apps based on a user’s Web of Trust. As the number of Nostr apps (already in the 100s) goes exponential, this use case will become increasingly important and may result in a massive market disruption: the opportunity to create an open app store that rivals Apple or Google. Franzap wrote an extremely well researched evaluation of this opportunity.
One interesting avenue to explore is creating a Nostr based app store for Progressive Web Apps (PWAs), which are usable on phones and unblockable by the app stores. Store.app is already experimenting with a similar vision (sans Nostr discovery for now) and sharp investors are on the lookout for something similar. Nostr discoverability and reputation could be the missing ingredients for this potentially huge idea to reach critical mass.
Search & Discovery: Search and discovery represents perhaps the largest potential prize for Nostr disruption. Generative AI is making it trivial to create compelling content supporting whatever version of the truth you want and is set to increase the already unfathomable amount of information on the web by orders of magnitude. All of this at a time when censorship is on the rise, faith in institutions is reaching new lows, and controversy abounds around our legacy sources of “truth.” Fortunately, this is an area where Nostr excels.
First and foremost, every Nostr event is signed with an author’s private key, making it the only information transport layer where you can be sure the message is what a user claims it to be. Beyond that, Nostr is beautiful in that it embraces a fundamental truth of the universe: there is no absolute truth or single point of view, only many coexisting perspectives. So when searching for information on Nostr, there is no such thing as truth. Only perspective. And Nostr’s discovery / Web of Trust allows users to easily find perspectives from users they trust (or don’t) and compare them accordingly. Wikifreedia, a Nostr based Wikipedia alternative, showcases this reality perfectly. There are potentially limitless entries for each subject and your own personal Web of Trust dictates which entries you surface. Highlighter, an open annotation layer for the web (or Rap Genius on Nostr as I like to think of it), functions according to similar principles.
The next big opportunity may be search engines that surface competing versions of the truth according to your personal Web of Trust (an open social graph). These search engines will also have another valuable tool at their disposal for retrieving information: ValueRank. As I detailed in my 2022 thesis How to Disrupt Google, ValueRank (or MarketRank), may provide the basis for a new search algorithm which surfaces information according to the scarce value that your social graph zapped (an open value graph). Nostr.band, Primal, and Stacker News are the first sites to make use of these discovery algorithms. And Primal’s aforementioned new Top Zaps feature represents the biggest step yet toward making an open social + value graph an easily understandable default. I expect other search engines to follow suit. Hacker News darling Kagi is already indexing Nostr and accepting Lightning Payments making it particularly well positioned to start experimenting with these new techniques. Note that LLMs could also operate according to the same principles, fine tuning based on an open social and value graph. Unleashed.chat is already fine tuning models on Nostr data and integrating ValueRank would be a logical next step.
File Marketplaces: In addition to general search & discovery for information published on the web, I also expect Nostr to birth a decentralized market for files. As Stu and hzrd149 describe in this excellent podcast, Nostr provides the perfect simple identity system for uploading and discovering files as each upload is signed by a user’s private key, a hash of which can confirm the file’s authenticity. Nostr relays don’t store media beyond text though, which is why Stu and hzrd149 are working on Blossom, a complementary protocol for media storage, and Satellite CDN, a Nostr content delivery network. Users can easily request a file with its specific hash and retrieve it either for free if the file is popular or for a market rate if the file is more obscure and located at the edge of the network. Nostr is useful here for confirming file authenticity, discovering where a user uploaded a file, and for determining which uploaders to trust. We’ve never seen a truly free market for files (BitTorrent comes closest but lacks identity and payments) and this could end up being one of the most exciting novel Nostr use cases.
Note that all of the existing projects mentioned (and the dozens that will soon follow) are already interesting and useful. But they’ll be far more interesting and useful with at least 1-2 orders of magnitude more users and notes on Nostr, which is why I think these use cases will really start to pop after the next wave of Bitcoin users arrive for Bitcoin related services.
Mainstream Apps and Marketplaces
Eventually everything will move to Nostr because of its super network effects. After Nostr achieves critical mass with the Bitcioners and early tech adopters, it will eat the mainstream web. Here are a few possibilities.
Reviews + UGC: Reviews and User Generated Content (UGC) are particularly well suited to user-owned identity and Web of Trust. Think Yelp, Google Reviews, Goodreads, etc. but that automatically surfaces reviews from people you know or are likely to find trustworthy. Aleksandar Svetski wrote a nice piece detailing the magnitude of the opportunity around Nostr based reviews. Bitcoin focused review sites like Apollo are likely first movers here.
Marketplaces: Next, I’d expect to see pure information marketplaces follow suit. Think expert networks or marketplaces for (likely AI assisted) code, content, and art. Eventually, we’ll see marketplaces that touch atoms – e.g. Nostr based Cragslist, AirBnB, and Uber. The second order consequences of decentralized, mainstream marketplaces on Nostr will be absolutely massive for society. As Stu eloquently explains, trust is deflationary. In a world where you own your identity and Web of Trust, you can easily look for 1st or 2nd degree connections with whom you can negotiate renting a room in their house (or even staying for free ala Nostr based Couchsurfing), disrupting centralized marketplaces and their opaque, excessive (potentially >50%!) take rates. Another way of saying this is that user-owned identity + Web of Trust begins to shift the balance from financial to social capital, which should reduce the cost and increase the frequency of all commerce and activity across the board. Not to mention making for a more livable, humane society.
P2P Content: I also expect Nostr identity + zaps to go mainstream for content creators. Podcast apps like Fountain, music apps like Wavlake, and crowdfunding apps like Geyser are leading the charge already. Users can log in with Nostr identities and zap content they enjoy with Lightning payments. These zaps can add up to a meaningful Value4Value economy, where many artists have the potential to collect more in payments directly from fans (without any middleman) than they currently do on platforms like Spotify, which pay only a pittance to all but the absolute largest artists. In order for this to work at scale, we need orders of magnitude more users. But as you can see from the first few hours of displaying Top Zaps on Primal, we already have an example of a note receiving $100s worth of value, and I expect to see a million dollar note sooner than you’d imagine.
Of course the most exciting category of all is the unexpected wave of unimaginable apps that will eventually emerge thanks to the nature of this new substrate. These might be the new dating and recruiting sites that can mash up way more user data from around the web. Or transparent marketplaces and brokers for user data itself. And who knows what else? The surprise and unforecastable nature of it all is a big part of what gets me so excited!
While there’s a lot I don’t know, what I do feel increasingly confident about is that we finally have the identity/data transport layer for the Internet. To be sure it’s early – it feels like what I imagine 2010/11 Bitcoin felt like or the late 80s/early 90s Internet circa The Well, replete with all of the idiosyncratic characters, dreamers, and builders. But being early is when things are most fun. And the upside from rebuilding and dramatically expanding the web is truly unlimited. TL;DR The Internet needs an identity layer and Nostr has more than a chance of working 😉
Thanks to Artur, Miljan, Bob, DK, and Moritz for feedback on this essay.
npub copied to clipboard
ok
