Interview with Andre Staltz on Decentralized Networks

Hey all! I’ve been aware of Edgeryders for a while, but hadn’t yet made an account over here. I didn’t realize there was so much actual discussion going on.

Some of you may know me from Scuttlebutt (SSB) circles. Among other things, I also write in my blog about tech and society. Perhaps the most relevant to this forum is “a plan to rescue the web from the internet”. I also recommend this talk I gave this year: “the internet in 2030” (hint: it’s a lot about climate crisis).

@zelf is who encouraged me to finally make this post. We recently sat down in Helsinki for a podcast/interview. Link and transcript below.

Nice to meet you all and happy holidays!

Interview audio


Welcome to the fourth episode of the Edgeryders podcast on the Human-Centric Internet. Today, we have Andre Staltz with us. Andre Staltz is a thinker, writer and a programmer. You’re particularly well known for having done Cycle.js.

Yeah, I’d say like maybe two years ago, that was like two or three years, that was the top thing. So I got invited to a lot of JavaScript conferences related to that Cycle.js. Nowadays, I’m doing more scuttlebutt things.

Especially nowadays: Manyverse.

Yeah Manyverse is my main project, actually, they’re kind of related so I used Cycle.js to build Manyverse.

We’ve had a few hours of conversation before this.

Yeah, actually.

Which is awesome. Yeah. To start off in general, though, do you want to give your take on Scuttlebutt because I don’t think I’ve interviewed anyone about it as part of this series.

Yeah, um, my take on scuttlebutt. So scuttlebutt, I discovered it, like as a, as a process that I just suddenly, you know, I want to move on and build new stuff I wanted to like join some project that’s like very promising or interesting and purposeful. And I looked at a lot of projects, and many of those were related to blockchain. And because I saw that, you know, the control that tech giants have over the current situation of the internet is quite, quite strong and relevant. So it seems like a problem to solve and I saw, you know, looked a lot of blockchain projects and wasn’t particularly convinced that they are, you know, there was that time when the IPOs were just like, everyone know, ICOs, I’m sorry, we’re everywhere. And so this ICO for everything, and you kind of felt like, they just want the money here, you know, and not so there was some content as well anyway. And then I stumbled upon scuttlebutt, and I was like, this is interesting, but what like kept me in scuttlebutt and what fascinated me in scuttlebutt was that it, it worked in a very different way of things that I’ve seen before.

So it it’s a protocol for, let’s say, a decentralized social network. And I’ve worked with social networks before, like, I had a startup that did a forum discussion platform. And, you know, startups usually fail, mine failed as well. But in in that one, I wanted to get right, the idea of moderation, like, you know, like, how can we have moderation without that being like a fixed dictator model, which is, like the model that exists in many forums. It’s like, Well, here’s the community. And here’s the dictator. Like, obviously, that’s going to run into a lot of problems for the community because there’s a dictator and so I wanted to solve that problem. And I came up with like this fluid model of moderation. But what really fascinated me and scuttlebutt is that I was always looking for, “Okay, what is the moderation model here like where’s The where’s the power? And how are people, you know, policing What’s going on?” And I just couldn’t find it. And I was like, why is there no model and at the same time that I couldn’t find it, it also felt like the nicest place that I’ve been on the internet so far. And I was like, How can this be? And it just fascinated me. And I was like, there’s something going on here. And like, the more I got involved with it, the more interesting it became. And then there was also like, a really strong community. So I think like, the things that Scuttlebutt are doing is doing in the space of social networks is very interesting. I think it’s bringing a new narrative to trust, for instance. So far, like people have trusted the platform or trusted the moderators. And now in scuttlebutt, it’s kind of like, well, you have people in your life that you trust. Let’s just use that. Like, we’re not going to introduce any platform trust any sort of dictated moderator trust is just people you know and that’s it. Yeah. So that’s my take on Scuttlebutt and what really fascinated me and, and I decided like, okay, I want to help this and there wasn’t like a mobile app. So I thought, okay, mobile is very big. So I think I can help in that sense. Yeah. So Manyverse is Scuttlebutt on mobile.

you mentioned it in passing there. There’s a lot of conversations that happen on Scuttlebutt. And there’s also a lot of like, ideas that float around. I’m thinking, number one, there’s a lot of cool conversations happening on Scuttlebutt. But yeah, you write a lot of great longer posts. And if we take this back to a conversation we had previously on collective intelligence, we were talking about how to utilize it to map together our understanding. Do you want to bring back what you described to me?

Yeah. I think there’s all kinds of ideas like that, you know, how we could utilize scuttlebutt and not just like, Hey, here’s a tool we can utilize it. But there’s a lot of like concepts and ideology. I wouldn’t say it’s not yet ideology, it’s more like ideas that can become ideology, sort of like in scuttlebutt, what matters are the people that you’re connected to. And that’s, that’s what creates like substance and content. And it’s it’s also the infrastructure so like, the network topology in scuttlebutt is your network of friends. So network topology means it’s kind of like a computer infrastructure concept, that let’s say you have a bunch of computers, and how are those computers connected to each other? So in something like let’s say Facebook, you have actually in Sweden, you have a data center for Facebook that has a bunch of computers there, and those computers are talking each other, like in that data center, they’re also talking to a data center all around the other side of the planet in California, let’s say to, you know, update the data. And that the way that those computers are connected to each other, has almost nothing to do with how the human relationships are connected to each other. So let’s say I’m friends with with you and you’re in Sweden, the way the the computers are going to negotiate that and how the computers are connected to each other has nothing to do with that is just add, they just, you know, it’s social. Yeah, I mean, in the end, they do deliver the data for us and that kind of thing. But like, yeah, it doesn’t work. And something like a discussion forum, you know, you have a server that is somewhere. And that’s where all of the data is. And the network topology in that situation is everybody who’s in the forum, has some laptop or some mobile phone and those computers are talking to that server. So it’s like a kind of like a star situation where you have this server in the middle, and a bunch of you are connecting to that. And that’s the network topology.

So what Scuttlebutt is like is, if I’m friends with you, then my computer connects with your computer. And that’s it. There’s like no computer in the middle that we connect to. You see? So what that happens, what that does is that it sort of so the network topology in the other services are usually invisible. We don’t see them, we don’t see how are the computers talking to each other? It doesn’t kind of we don’t care about them, but in scuttlebutt, because the network topology is the social connections, suddenly, the network topology is visible, we know what it is, it’s because it’s the same as the social connections and we know what the social connections are. So I think these ideas are very promising, because suddenly, computer connections start becoming more like a day to day concept for us, because our computer You know, let’s say my computer represents who I am on the internet, and it gets to connect with your computer. So it has a bit more like, there’s more substance to these connections. I mean, these connections are not, they become more important, because because let’s say if I stop being your friend, my computer’s not connecting to your computer anymore, and then I might get less data. So it becomes more real. I mean, there’s more substance to it. For instance, if I would like, let’s say, block you on Twitter, and I wouldn’t do that. But anyway, if I would block you on Twitter, you could make another account and follow me and you still get my data. Right, it didn’t really matter that I blocked you. It didn’t have any sort of like that blocking didn’t have like a substance in the network topology, whereas was scuttlebutt and if I would block you, you would not have that connection to get my data anymore, and you would have to get it from other people. So what was the initial question about collective Intelligence?

Yeah, you just described it. And what you described is basically a system that replicates the human interaction.

Yes, yes. Yeah.

Yeah. Which means from a network topology, you can see the direct connections also between the data and for collective intelligence. What could that mean?

So I think the best example that we have, well, one of the best or most visible examples, we have a collective intelligence on the internet today is say Wikipedia. So a bunch of people get together and put their intelligence together and they make one single article about this thing. And when you use Wikipedia, you don’t often see the authors you just see the outcome. You know, you see the, you know, this is what this intelligent thing wrote. And it’s really like some intelligent thing, wrote it or some intelligent crowd wrote it, I think we can we can get better at doing that kind of thing, because there’s a lot of obstacles in writing on Wikipedia.

Yeah, like I submitted once scuttlebutt article for Wikipedia, and it took like weeks to be reviewed. And then when it was finally reviewed, they were like, this is not suitable for Wikipedia. And I was like, What is suitable then? Like it was so vague of an answer. So you kind of noticed that there are better ways how we can do this. And I think that this, let’s say, to come together as a crowd and produce an intelligent an article such as something on Wikipedia. So I think we need something like Wikis on scuttlebutt. And what would that mean? So for instance, let’s get something simple, like a cookie recipe. I could have my cookie recipe, and this is like the best that I’ve found so far. And then let’s say I make friendship with someone like let’s say you, and then I suddenly would have access to all the articles that you wrote. And maybe you have a recipe for cookies that’s actually better than mine. So I can decide like Okay, I could possibly, let’s say, click like “accept your recipe” and then I would update my recipe to be your recipe. But, except that you add like the say Cardamom on the cookie. And I don’t like that. So I’m just going to remove that. So there could be like, new ways of adapting each other’s intelligence and merging them in ways that don’t require global consensus. So one of the aspects of scuttlebutt is that it doesn’t require a global consensus.

I’m maybe getting too much ahead of myself. But what I’m trying to get is to is that Wikipedia, it is about global consensus, essentially. And there’s a lot of other things on the internet that our global consensus, For instance, blockchain is global consensus. There’s like, The Blockchain is like gigabytes I think, or maybe terabyte I don’t know how big it is right now. It’s very big. And what it means is like it’s the global ledger of all the Bitcoin transactions, and it’s a global consensus system.

Global consensus is a very difficult problem to solve I think. I think we’re getting too much ahead of ourselves. I think we should try to solve like smaller cases of consensus. And that’s kind of the space where Scuttlebutt operates in because you only need to think about, okay, what does this mean in my social space, like the people that I am interacting with? Maybe, you know, if you make a cookie recipe on Wikipedia, there will be all kinds of opinions from all kinds of experts and people, that it’s actually a very difficult process to, you know, review all of those things. And there’s a lot of friction, I think. So I think scuttlebutt helps us have a smaller barrier to participate in collective intelligence because the collective is not so huge as like global. It’s basically like the people that you’re in interaction with the people that you trust and it’s very, very liquid for that sort of, like, flexible and liquid. Because let’s say someone that was friends with like, had good cookie recipes and stuff and good let’s say recipes to not just food but like let’s say how to build stuff right. Like, like Let’s say 3d printing stuff. If relationships change, I could later like not be connected to them, and I would sort of not receive their ideas anymore. So it’s a bit more fluid in the sense that we don’t need to fully agree. But we can, like choose who we want to agree with. And yeah, it’s a bit more fluid. So I think it is important that we find ways of making that that collaboration faster. Because obviously, like if I submit an article to Wikipedia and takes weeks to review, that’s, that’s not good.

It’s almost like it sounds to me almost like you’re suggesting a way to form a collective brain.

Yeah, or collective neurons. I mean, beginning somewhere small and Yeah, actually, the brain is a huge thing. So I think the brain is like the ultimate goal if we could create a collective brain. But yes, it’s going towards that direction, basically, sort of like some people have some insights. Some other people have other insights. How can we get there? sites and put them together. What’s very interesting about the brain is that there’s no central part, like hierarchically distributed, that this is the ultimate neuron that tells those neurons how to interact with each other. It’s like not a tree is definitely like, is actually So, so mixed up and constantly evolving that, like, we don’t understand the brain, like people don’t understand the brain yet. It’s like literally like a huge mesh of chaos. And that’s your brain.

Yeah, a great example of that, when people split the brain. And different parts of the body.

Oh, yeah. Yeah, so I’ve been reading some books and like brain is not. So we have this idea that we are individual and that we have like a simple idea of who we are and things like that. But like the brain is so diverse, like it, like how it functions that the left side and the right side can take different decisions. So some scientists, I think psychological and the psychology field. They did some experiments that, you know, if you shut down, like, I think if you induce some waves and sort of little bit shut down the influence of one side of the brain, then it will take one decision. And if you shut the other side down, then the right side will choose yet another decision. I think it was something like on the question, they they put someone in the test chair, their visual field was controlled by one side of the brain, and it showed them one image only on one eye. And then they asked what you want to be when you grow up? And they said one answer and then the other in the other case, they did the opposite. And the person answered, I want to be this thing when I grow up. So it’s literally if you ask someone like who are you? The question depends on what part of the brain are you talking to? So yeah, the brain is a very complex thing. I don’t even want to pretend I know about that.

I want to just throw in one comment because you talk about left hand side right side of the brain and Society of left and right.

Oh, yeah, that’s kind of interesting.

And if you cut off communications between them you get two different worlds. Oh, yeah.

And also it makes the point that there is no such Central Party. Right? Yeah. That I never thought about that. But we are getting to a point as society where the decisions of the collective sometimes are more interesting than the decisions of individuals.

I actually… is it okay if I bring you back to when we were on collective intelligence - you mentioned this system of sharing neurons almost became because… or sharing understanding of the world or whether the cookie recipe or 3d printing. Do you think those possible now we’re now just like expanding, thinking big, do you think it’s possible to extend that kind of collective and Intelligence into how we build infrastructure and society?

The answer is yes. Because there’s a lot of knowledge and that is locked in, in many institutions. Basic know how. And that knowledge is is it like the idea that we share information is an idea that the internet started like promoting that, you know, we get information, we put it out there and we, you know, share it? Well, I don’t know exactly where it started. But like, one of the main examples was, for instance, the p2p movement, you know, like, by the time that like people started sharing mp3 files, you had this idea that Well, I have, like this music file, but it’s just information I can share that information and is we just all enrich each other, you know, this idea that we can enrich each other. And then came like the a open source movement, which is basically this idea of like, you know, institutions such as companies and like universities, they all had like information such as code that was valuable to them, but it was locked, like in their institutions.

So if you could get that know how, and just like release it out there. I think the easiest case, of course, is with software because the information is the software, right? So it’s a bit more direct. And it’s also because they are so involved with computers that are very obviously involved with the internet. So it’s obvious that like the open source movement group, but I’m thinking what would be the open source movement for other institutions such as, you know, maybe health institutions or educational institutions that like they have know how, and it’s sometimes locked in people sometimes locked in special institutions, but I think there’s a lot of things that start happening once you start releasing all that knowledge. And it has to happen on something. So for instance, open source happened on GitHub, or basically a platform. And it has to that’s where, you know, people have to find the each other, that’s the thing, right? How do they find each other, then we start talking about things such as either social networks, and GitHub is a social network or protocols for social network. And that’s where sort of scuttlebutt comes into the picture.

Actually, you know, I think we should really emphasize that email is a decentralized protocol. And it has helped for decentralizing information and sharing information also, between institutions. I mean, email is actually a really big thing that helps us share knowledge. And we need to remember that, you know, that’s just one of the things that we can do, we could do much more than that we could like have much better systems than just email.

To add on to that you you mentioned how we start sharing that information. And for the infrastructures of society, we also already do have computer controlled mechanics, which means that building even hardware or building physical things still so computer data can The open source it isn’t as code, which is pretty amazing.

Before we round up, you mentioned also that scuttlebutt is essential in the sense that it’s a social network. And it’s interesting because it is so easy to build collective intelligence upon a social network that also replicates the human structure that we use to share data between humans.

Are there any shout outs? Any universe here we do this any anything else that you’d like to throw out there before we run this out?

universe? Hear me? Um, what kind of thing?

It can be. I don’t know. Anything. Thanks to your ManyVerse backers.

Oh, yes. Well, definitely. I have to thank them. That’s that’s that’s one thing that Yeah, I’m actually very happy that like people like some some of them have donated an amount that if I would do it myself, I would find it like this is this is ridiculous. And I’m really thankful for those people.

And I also think that I need more help in general. One of the things that I’m noticing is that there is actually a scarcity of people and resources in one of these fields that seemingly looks like it’s, it’s on fire and it’s like awesome, such as open source, for instance, it seems like everything like open source is the hottest thing, but there’s actually scarcity of people and resources and funding and stuff. So for instance, I’ve talked with a couple of other people who are concerned about open source funding we have this call where we were like talking you know, what can we do and stuff. The situation is looking pretty negative, but my my shout out is this is the future and it’s like, it’s not because I want it to be the future because it’s actually changing world like crazy way. We need to sort of think okay, how do we fit this into? Not necessarily how do we fit this into The society that has existed is so far, we need to think, Okay, how do we build society with this or on this? And I think we need a lot more people and resources. And we also need to get together with more people from other fields, I think, like these, you know, programmers who are doing open source and stuff, I think they should be talking to anthropologists and people who have been thinking about, you know, I don’t know, actually I don’t know. But the point is, like, we need to get better at finding people from other fields. And I’m not blaming anyone, I’m not saying that the programs are too stuck into the corners. I’m just saying that we also, we not only miss people in funding, we miss opportunities to bridge with other communities, other fields of research and other kind of stuff. And we need to create a dialogue with that.

There’s just this potential for things to happen. And for things to get better in this field that seems to be taking fire, beyond fire and like in really being hyped and stuff. I just want to be part of that in some way.

Thank you guys for listening. Thank you. And thank you, Andre. for sharing, fantastic worlds. For those of you who see the potential in the structure: shout out. There definitely seems to be a chance for a different world.

So that wraps it up.

Thank you for inviting me to the podcast.


Hello @andrestaltz , we know each other (a bit) from Scuttlebutt. Nice to read you! This made me think:

When I think about collective intelligence, I like to use metaphors from evolutionary biology. I like social systems with an engine of variation (like mutation or sex in biology), and an engine of selection among the variants (like natural selection in biology). In the right conditions, these systems converge: some variants are superior to others, so they outcompete them, and suddenly the species has evolved. I am aware of the problems posed by taking Darwinian evolution as a blueprint for social processes (for one, the variation and selection processes are not independent from each other in most social processes). But still, there is something very appealing in this schema.

Back in the days of the unMonastery, I tried to apply it to Protocol, a set of low-level rules of interaction for working together. Sometimes it seems a lot of Egderyders lives by it, other times it just seems a failure. Good one, though :slight_smile:


Great to have you here @andrestaltz!

Nice Interview!

The great leviathan can be democratic if we think of it as collective intelligence becoming its own deciding entity with the help of developing tech. I am quite curious where that will go and how that will effect current political structures and our concept of the political party.

There are quite a few people engaging in serious discussions and development in the field of collective intelligence on the Edgeryders platform at the moment.

For example this application for the CI incubator to answer the H2020 call:

great idea and team. Maybe you would like to have a look and add your perspective. You could also join their weekly calls:

Edgeryders Team is inviting you to a scheduled Zoom meeting.

Topic: H2020 Collective Intelligence Call
Time: Dec 26, 2019 08:00 PM Brussels
Every day, until Dec 27, 2019, 2 occurrence(s)
Dec 26, 2019 08:00 PM
Dec 27, 2019 08:00 PM
Please download and import the following iCalendar (.ics) files to your calendar system.
Daily: Error - Zoom

Join Zoom Meeting

Meeting ID: 405 426 755


Hey @alberto and @MariaEuler, thanks for the replies. Took me a while to reply because I had a deadline.

On collective intelligence, I realize Alberto has quite different intuitions about it than I do. The things I was envisioning during that interview were specific designs that could shape internet interactions towards being more “meaningful” or “efficient”. Two designs I’d like to share:

Social bookmarking, but not the way it has been done so far. When you browse the web with a classical browser, it highlights visited links in purple, and non-visited links in blue. It shows your footprint of browsing the web, but no one else’s. What if you could see links that others visited too? Now, before you think about the privacy implications of that, we could bypass that problem by only highlighting links that others have shared publicly. And by others, I’m talking about friends, or people you follow. This design is quite achievable with Scuttlebutt: it’s a matter of processing all the web links shared on SSB, indexing them, and then in the browser through an extension we can highlight links with a special symbol, in case they were shared by friends on SSB, and we can annotate those links with the comment they placed when they shared the link. What Wikipedia did for the collective compilation of knowledge, we could do for the collective exploration of the internet.

Another idea is to extend WorkFlowy (or an alternative to it, such as Orgmode) with MMO (massive multiplayer online) capabilities. I use WorkFlowy for all my personal info, notes, journaling, etc. Its slogan is “Organize your brain”. What if we extend the “your” to “our”? With @zelf and other Europe-based Scuttlebutters we started using WorkFlowy to sketch our notes, and the collaboration model is quite similar to a shared Google Doc. But with the difference that with infinitely-nested lists, we can have just one document and refactor that as we go. So far the experiment has been very fruitful, and we’re getting really efficient at using it. One person may make a sublist for a certain bullet point, and insert a “where is web link for this?” and another person may suddenly web search that and fill in that info. When we’re online at the same time it feels like collective lego playing, but with nested lists. It’s essentially mindmapping, but productive, uncluttered, and efficient, with more emphasis on productive than on mindmapping.

Maria, I recall seeing this H2020 initiative with Worldbrain and others from somewhere! I’m curious to hear more, but if I understood correctly from this calendar, the Zoom meetings were in the past? Are they still occurring?

1 Like

I do not have this impression at all. I am simply chewing it at a higher level of abstraction: collective intelligence is best achieved not by recruiting the smartest people, but by investing in collaboration practices. This is absolutely not the same thing as investing in groupware: the world is littered with orgs who spend lots of money tooling up with one or the other project management suites, but when push comes to shove default to unstructured email with heavily formatted MS-Word attachments. Or, even, meetings. No, collaboration practices are really social rules, and require quite some social engineering. I still find this intuition profound and fresh, and realize that my own tribe (this one), while comparatively advanced, is nowhere near taking it onboard to its fullest.

Your reflections on

Continues the same intellectual trajectory. You focus on tools, because your group already has an unspoken agreement that each member is going to make some extra effort to achieve better group-level efficiency. Most groups do not have that. Even those that do are constantly struggling to onboard new people, or contrast the tendency of members to “cut corners”.

Actually, Discourse does this already. It displays small numbers next to links: they are click counts. I have no information on how appreciated this feature is. I use it mostly for my own posts, as a very rough indicator of the usefulness of the links that I used to back my argument.

I do not know WorkFlowy, and their website is just a sign-up-for-updates page. So, I cannot comment on anything useful there.

This has really been at the heart of the exploration into collaboration at The Borderland, which has been a great playground to test the propagation of “best practices by meme” and a bit of nudging through available incentives. I have always been skeptical of anyone who tries to offer collective intelligence through a tool alone.

As I understand it, this is exactly where @BlackForestBoi is trying to go with Memex & Storex?