"We want functional decentralization" Q&A with Wildland Creators

Wildland is a new open data management protocol that aims to give individuals explicit control over their online data, in a way that guarantees its availability, accessibility, and users’ privacy. 

At the end of December 2020, I met with Julian Zawistowski and Andrzej Regulski to discuss the high-level rationale and nuts and bolts of the project.

What is Wildland and what kind of end-goal do you envision for it?

Julian: The modern-day Internet operates within the service-oriented paradigm, which heavily favors providers over users. You may have access to your online data, but your control over it is very limited. Other people profit from your data, and you can be easily cut off from it based on their arbitrary decisions. With Wildland we are challenging this whole paradigm by introducing a data-oriented approach.

Wildland is a new, open data management protocol that offers individuals a novel way of interacting with their data, and provides them with improved control over it.

With Wildland you will be able to keep your data with providers of your choice, without giving them access to your information. You will have the ability to easily move your data from one cloud storage to another, without having to manually set up accounts on any of them. You will be also able to set up many different paths to your data and access it using any of them.

So, the initial value-added of Wildland will be the user’s ability to more efficiently interact with the data and easily combine different backends and migrate between them.

Andrzej: We do not see data management - moving files between various devices - as an objective on its own. The ability to seamlessly transfer data from one backend to another is more of a feature than a goal.

The most important thing here is the fact that there is a growing number of people who are frustrated with the way modern software works. An increasing part of our digital life is spent on interactions with various online services to which you connect using a browser or an app. Your data is stored in a silo-like environment over which you have little control. While it is often possible to retrieve your data from such a silo, and export it somewhere else, the whole process is usually so tedious and cumbersome that hardly anyone is bothered to go through it.

This is where the issue of control becomes important. Because if you truly own your data, you can decide what kind of tools you would like to use to interact with it. Your past interactions shouldn’t determine the pattern of usage for the rest of your life.

Down the road, we can imagine Wildland as an interesting platform for developers who find the vision of user-controlled apps appealing and for anyone who would like to be able to switch between various applications and various services in a more flexible way.

How did Wildland come about?

About two years ago, when Joanna Rutkowska [of the Qubes OS fame] joined the Golem Factory team, which we were leading back then, we started to form some very promising new ideas on how to further advance Golem. We quickly realized, however, that our new endeavors do not converge into a single project. Not in a sense that the new ideas are impossible to reconcile with the original one, in our opinion they realize the very same set of values and goals, but rather that the complexity of the whole thing is becoming too big, to be effectively managed within one organization.

This is why about a year and a half ago we decided to start a separate entity, the Golem Foundation, which focuses its efforts on developing entirely new, innovative approaches that move forward the whole Golem ecosystem.

Wildland is the first fully developed idea that resulted from our initial brainstorms, and I think it combines our experience in decentralization and Joanna’s world-class InfoSec expertise really well.

What kind of demand do you envision for Wildland? Who do you think will be using Wildland?

Julian: You’ll get different answers for that question, depending on whom you’d ask from the team.

For me, the most appealing thing about Wildland is the promise of taking back control over your digital life. And by control, I mean the ability to decide where your data is stored, who has access to it, who can wipe it out, and who can cut you off from it. So, I’m mostly excited about the Wildland properties that are connected to the project’s infrastructure layer.

We’re putting a lot of effort into making data management seamless, almost invisible to the user.

But all these Wildland’s security and privacy features that excite me the most, are possible due to some pieces of technology that allow us to offer users other features as well. I’m speaking about multi-categorization here. Wildland allows users to define many access paths to their data, and browse it using any of them.

This multi-path approach to data addressing fits really nicely into the whole project because it allows you to regain control over your digital life not only in a purely technological sense - so that you’re the only person who can edit and delete your data - but also in a more “cognitive” way. It allows you to organize your data in a much more structured and logical way.

I believe that sovereignty is Wildland’s killer feature, but if you’d asked Joanna, she would probably say that multi-categorization is just as important.

As for who it is for: we’re building Wildland for us, which I think is a good way to approach things. You need to eat your own dog food and learn from it.

So, Wildland’s early adopters will be people like us. People who care about their data, and understand that much of it is stored in places over which they do not have control. People who believe that digital sovereignty is important, but who do not have the time or the skills necessary to set up their own infrastructure.

Once the ecosystem grows and matures, I can see Wildland having more mainstream appeal, especially after more apps will become integrated with it.

The decentralized data storage space is getting a bit crowded at the moment. Could you tell us how Wildland differs from e.g. Ipfs, Filecoin, Sia, Safe network, and other similar projects?

Julian: This is a complex question with a complex answer. The projects you’ve mentioned are not necessarily Wildland’s competitors. We are not looking for ways to store data in a decentralized and secure way - we are working on a solution that will enable you to abstract your data from the backend infrastructure.

Our end goal is to allow users to utilize any infrastructure they have and seamlessly move data between different backends, without having to instruct other parts of their setup that the place where their data is stored has changed. So, for us the question of where the data is stored is secondary, and the projects you’ve mentioned could very well serve as data storage options in Wildland.

What is similar between Wildland and some of these projects is censorship resistance. Our approach to achieving it is rather different, however.

Rather than reinventing the whole infrastructure layer, we want to provide users with flexibility and the ability to move their data around. While the approach might not be as radical as, for example, the one that IPFS utilizes, it comes with certain advantages. The most important one being that it is easier to integrate our approach into the existing technological stack, and simpler to implement overall.

What we want is a functional decentralization. We want to empower individual users and not some organizations. This is how we understand user control.

We believe that technical decentralization is not the only way in which you can achieve this goal. There are some things that you have to decentralize, especially the addressing and the transaction system, but my belief is that as long as you’re able to move your data between different backends in a secure and seamless fashion, the decentralization of the storage itself is not that important.

Thanks to our approach we’re also able to remove the issue of trust from our model. Whichever storage provider you choose depends solely on your trust in this particular provider. You do not have to put additional trust in Wildland itself. We do not have to build complex safeguards into our system to prevent malicious behavior.

Filecoin, for example, has this marvelous security design, which is built into the fabric of the blockchain itself. But their solution is very complex, and right now it is hard to tell whether it is going to work. But if it is going to work, we could just add Filecoin as one of the backends for Wildland.

Could you elaborate a bit on this trust issue?

Julian: What is cool about trustless solutions like Bitcoin or Ethereum is that you can interact safely with other parties without putting any trust in them. This is because there are mechanisms built in the protocol itself that protect you from malicious behaviors of others.

Andrzej: Unless someone executes a 50.1% attack.

Julian: Yes, of course. This mechanism relies on you having the ability to see that someone is misbehaving. For cryptocurrency transactions, this is rather simple. You can backward verify all the transactions on the blockchain, and check whether the transaction that was added to the block is valid or not.

It is not that simple, however, for anything more complex than a financial transaction. We actually experienced this problem at Golem Factory. Verification of results submitted by providers turned out to be a huge challenge and potential solutions were usually use-case-specific and not applicable in different types of computations. All in all, at a certain level of complexity building a truly trustless system is extremely hard or even impossible.

Right now, the development of Wildland is financed in full by Golem Foundation. How will Wildland be sustained once the project becomes available to regular users?

Julian: If you want to build a decentralized system in which it is the individual user who is empowered and not some organization, you need to come up with a model of financing that is not dependent on one organization.

It is a general problem, not particular to Wildland, of course. Organizations tend to go rogue, and even if they don’t renegade on the “don’t do evil” principle, they are not necessarily good at deciding what constitutes a public good. I honestly believe that projects like Wildland need a different governance model than just founders deciding on what to do next.

One of the problems with the services people use right now is that there is a conflict of interest between the users and the vendors. This is especially true for ‘free” services, but it can apply to paid services as well. Even if you pay for a service and your privacy is protected, it is in the best interest of the seller to create a vendor lock-in effect by vertical integration. Apple is a prime example of that.

This is why we came up with the idea of a User-Defined Organization. At the moment we only have a high-level vision of the concept, and we will have to develop it further before it will become fully operational.

We describe the UDO concept in the Wildland paper, but the general idea is that Wildland users will be rewarded with a non-transferable token whenever they initialize a payment within the Wildland marketplace. This token will then enable them to decide on who is to receive bounties for the further development of Wildland. The more transfers you initialize, the more tokens you acquire, and the bigger your influence on how the whole ecosystem operates gets. We call this token Proof-of-Usage, as the power you have to decide the organization’s fate is derived from how much you use the services it provides. Note that the new token is by its very nature a non-speculative asset. We believe that as a result, the governance framework of Wildland will be much more robust.

Our approach is partially inspired by DAOs, but I think that the idea of tying decision rights to the usage and measuring usage with a non-transferable token, is a novel approach that can be adopted by other entities. We actually believe that providing governance solutions to decentralized organizations could be the second project of the foundation.

Could you summarize what you have been up to until today, and what are your future plans with Wildland?

Julian: We have decided to invest a significant amount of time in conceptualizing Wildland and the fundamental premises behind the project.

I believe it was time well spent for two reasons. First, whether or not people are as excited about the idea as we are, it’s hard to deny that it’s logically and technologically consistent. We also believe that it is economically sound. It’s definitely something we can work with. 

Second, while we spent a lot of time on high-level work, we didn’t spend much money doing it. We were a small team back then, and our day to day operational costs were relatively low.

Once we arrived at the solid idea, which was around Spring, we started to scale up the organization, building the team, hiring new people, which unfortunately coincided with the pandemic, and that slowed us down a bit.

The biggest issue is recruitment, not many people are willing to change jobs at the moment. Despite these problems, we have made quite some progress during the last 6 months.

In September we held the WildCon1 event, during which we presented our progress since we published the whitepaper. We have working prototypes for the key components of Wildland.

Since then we worked mostly on improving what we already have and preparing for the beta release of Wildland. We do not have a concrete release date yet, but we are working full steam ahead.

The beta will be usable mostly for technologically proficient people, who can utilize the infrastructure they already have at hand. Later on, we will add to that an infrastructure provided by us to make tests and early adoption easier.

We’re also planning to release a more detailed version of the economic framework of Wildland and the accompanying code for it.

The last thing on the road to the working product will be the launch of the actual Wildland marketplace, with organizations offering storage backends.

We cannot guarantee how much of all this will happen in 2021, but so far we’ve been mostly on track.

Andrzej: In mid-term, we’d also like to engage some external partners in the process of building Wildland, through the system of grants, and other mechanisms we are designing right now. But it’s dependent on how fast we can bootstrap the technology. Without bootstrapping this core protocol, it doesn’t really make sense to engage people from the outside.

Your first project, Golem Network was named after Stanisław Lem’s novel “Golem XIV”. In the Wildland paper, there are references to another of Lem’s novels – "Solaris". You also referenced Lem’s work during the WildCon1 event. I wonder who’s responsible for these references.

Julian: Well, I brought the name to our first project. I also convinced Joanna that Lem is a great writer, worthy of her attention. All the references to his work in the paper were made by her.

The reason I’m asking is this - during his final years Lem became a rather vocal critic of the Internet, and I wonder whether his criticism influenced in any way the Wildland project.

Julian: This is a very interesting topic that goes well beyond Lem. It’s true that Lem was very critical of the Internet, and its influence on culture and society. By the time of his death - he died in 2006 - he predicted a lot of things that would have gone wrong with the Internet. But he was not the first one to make such predictions. There were people who were making similar predictions even before the Internet became a thing. I’m referring here mostly to Neil Postman and Marshall McLuhan.

I think that Neil Postman’s Amusing Ourselves to Death, which was first published in 1985 had more influence on us in this regard. Although there’s only a mention of what then was regarded as modern technologies, almost everything he wrote in his book about television you can extrapolate 1:1 to the Internet.

One of the main problems with the Internet is that every medium can support a certain level of a public debate, or complexity of communication, as Postman observed. It turns out that the Internet cannot really support high-level discussions. The Internet makes it really simple to address huge crowds of people while staying anonymous, and this is probably how most of the fake news originates.

I believe that Wildland can improve the Internet, but we focus our attention on a different problem. What troubles us the most is that people are routinely harvested for data, and their information is locked in silos, over which users have no control. Wildland can mitigate all that.

Thank you for the interview.

All photos courtesy of Golem Foundation.