This article is adapted from a talk originally given at the 2023 Urbit Volcano Summit.
Balaji claimed at Assembly that Community is Urbit's killer app. But I would argue that it is something more fundamental than that.
Trustlessness has its place, don't get me wrong, but in the modern era people have more than their fill of trustless and untrustworthy interactions. We don't need even more of this in our lives. As humans we want to be able to trust those around us. It makes us feel good to be surrounded by honest actors. And when it is lacking it hurts our soul.
From a purely practical sense this lack of trust is a centralizing force and has gotten us to the mess of the internet that we experience today. A technical inability to trust peers drives us towards institutions which then abuse the trust we place in them--and even those that are attempting to be honest actors are abused by malicious actors in order to take advantage of the trust we are forced to place in these centralized silos.
What does trust enable? Let's look at it in context of a multi-polar matrix of Prototype vs Protocol vs Product. To understand the rough shape of how to categorize a point on this matrix, we can look at how different functional definitions might map to this matrix:
Of course, the most straightforward way to look at this is taking different popular apps / software and mapping it to the matrix:
As we see the shape of application development mapped out, we can ask what incentives look at how different "revenue/funding models" might map out:
And of course as we think about what unlocks 'funding' or developer resources, we can look at how different success metrics are likely to be assessed along any one of these axis:
Notice anything in this mapping? Yes, there is a huge gap in the middle, and one that likely has a lot to do in our current headaches with web2:
The layer of trust that Urbit enables isn't just a warm and cozy feeling. It actually can do an immense amount of heavy lifting for managing uncertainty at a practical scale, reducing risk and driving down the floor for the cost of worthwhile interactions. It enables undertakings in this 'forgotten middle' that would previously have been impossible or impractical to fund, distribute, or implement because the cost of things like capital formation, outcome specificity, and assessment of project completion were too amorphous.
Aera is a protocol for establishing inter-node reputation between urbits where that reputation is not globally stated, but instead set by each node individually. The goal is to provide a minimally opinionated substrate for which reputation will aglomerate ontop of the Urbit ID system.
It is a non-canonical per-user global graph. While we agree there are times where you want a system that has a canonical perspective of global state, for most algorithmic reputation mechanisms you want what you want (fundamental truth be damned). By being recentralized on the individual, the graph acts as though it is cannonical (replicable given the same state for the same user), without being global.
There have been reputation systems in the past which attempt to do this, but to do this in web2, you have to work with their identity system, and you need to bend to their API. In contrast to trying to do this in web2, Aera allows Urbit developers to:
Bringing this functionality to developers allows us to take the technical affordances of Urbit (namely the p2p, encrypted, and cryptographically sovereign identity elements) to the next level with use cases that will provide a better user experience than what is available in web2 walled gardens. For example:
Quartus built %fund, an escrow agent + wrapper library in less than a week, and raised 2k in 48 hours from ~25 donors to fund an upgrade to our Wordle clone. People pledged, we trusted them to follow through on their end, and they are trusting us to deliver. The Dalten Collective treasury is serving as the escrow and the dalten membership will vote to approve the payment transaction when Quartus delivers on the work proposed in the crowdfunding campaign. Future versions will include built-in crypto transactions and an escrow smart contract with options for escrow agent fees. Aera would offer to %fund the mechanism by which developers, donors, and escrow agents/arbiters could garner and share reputational information in a landscape that is not predetermined and thus trust enables the ability to manage a certain level of uncertainty, providing answers to questions such as 'how good at following through on their promises is this particular donor?' This enables a certain scale of both fundraising projects that is not currently supported by DAOs, VC funding, and could also be applied to smaller multiparty contract work for completion of a given software spec.
Predicto! is a proposed integration with chat by ~rabsef-bicrym wherein you would see a friend make a prediction in chat, and then be able to request that the message be resurfaced to you at some future date. When presented with this past prediction, you would then personally judge the accuracy of the prediction. Over time, you would garner an understanding of the predictive capabilities of your cohort, and then see if they were good at predicting the predictive power of other users. This would then offer an emergent view of what events are being predicted by reputable sources, providing you valuable insight into possible futures and their probabilities.
Urbiters can often get caught up in rejecting megacorp services and forget that there often is a value that gets delivered to users around algorithmic recommendations and content suggestions. This is real human value, but it gets perverted to the extent that megacorps sell access into your feed to advertisers or other entities seeking to manipulate your behavior. And even with all that, it is still valuable to people. But even with all that, issues still arise for how these recommendation engines work. For example, with stuff like Yelp, Uber, or AirBnB:
Aera would give devs the ability to surface rating systems across applications and have ratings weighted to how their users percieve them. Does your brother give out 5-star ratings with literally every review? Perhaps the way that you understand his ratings to relate to your experiences with similar things--and the ratings of all your other friends--can actually be adapted to the way that you are recommended a novel experience or piece of content. And critically this offers a mechanism by which people can get what web2 offers (i.e. dopamine-infused feeds, if that is what they want), but also have an option to exit if that is more their style.
"A certain level of piracy is desirable." - ~harden-hardys
At Vaporware, we are working on using Urbit as an infrastructure layer for Non-Fungible Tokens (NFTs). Instead of just being jpgs (deserving of right click, save as mockery), NFTs can be used for software provenance/distribution/legitimacy including a vast new array of artistic creativity. But as is alluded to with the current ridicule of million-dollar rocks and shitty monke pfps (ridicule in which I whole-heartedly participate), the ability to fully lock down data is not possible, and may not even be desirable. Using Aera within networks of digital goods, we get a soft enforcement layer of reputation prompting a healthy level of creative destruction. Think of this in context of music sharing in the early 2000s, where there was a scale of what was socially acceptable:
Now, this roughly represents the 'base case' of a normie social network, but you can imagine networks and subcultures where these activities have different interpretations. And this 'looseness' or 'leakage' of data/content prompts some level of emergent creativity as ideas and information propagate through a network, while providing an appropriate scaling of enforcement for how to accrue resources to original creators.
Obviously these use cases are just a drop in the bucket compared to what developers will come up with for %aera, or from what affordances end users will desire as they begin to understand what it means to have proper sovereignty over their own social graph. But, in a world where threats of global financial crises, surveillance and censorship come from central banks and manipulative practices, we need to be diligent about bringing sovereignty back into the hands of individuals. Technologies like blockchains, ZK proofs, and decentalized exchanges have their roles to play in the coming epoch, but it is important to remember the mechanisms by which we maintain the humanity of our interations and the subjectivity of our experiences.
Full financialization of all interactions (by over-indexing on tokenization and global state) will threaten our ability to have the types of digital relationships that will result in serendipitous experiences.
By building %aera on Urbit we will further the ability for humans to bring trust into a digital first world.