What Grinds My Gears: Mesh Much? Telecommunications, Topology, and Tokens

Most of us take our connectivity to the internet and other telecommunications networks for granted. But behind the simple and benign act of connecting to the internet lies an intricate maze of physical hardware, wires, and complex routing logic consisting of layers upon layers - and many curious parties sniffing your web traffic. In this episode, we discuss the origins of mesh networking topologies and innovations in using cryptocurrencies to enhance security, facilitate more complex logical ordering, and incentivize resource sharing in these new networks.

WGMG S2E4 Blog Graphic.png

Meltem Demirors
The year is 1962. The United States and the USSR are in the depths of the Cuban Missile Crisis. And both sides are building nuclear systems and figuring out how to use technology to ensure mutually-assured destruction in case of a nuclear attack. U.S. authorities, deep in their planning, are facing a really interesting challenge. How will they communicate in the aftermath of a nuclear attack? In a command-and-control network, which needs phone operators connecting lines, how does connectivity survive that type of attack? Centralized switching facilities like long-distance phone plans, military command-and-control systems, these are all going to be targets.

Paul Baran, who's a researcher at Rand, the private think tank, dreams up a new system. He envisions a network of unmanned nodes that would act as switches, routing information from one node to another to their final destination. Now, these nodes use a scheme that he calls hot-potato routing, or distributed communications. And furthermore, the information that's relayed is going to be divided into message blocks before they're sent out across the network. Each block will be sent separately and then rejoined into a whole when they're finally received at their destination. This innovation is the foundation upon which the worldwide web was built.

Jill Carlson
Baran's work was published in 1962 and expanded upon two years later, in 1964. Using these foundations of packet routing, or, as Baran called it, hot-potato routing, and packet switching, ARPANET was launched in 1969 and has its first node housed at UCLA and another node at Rand. It was intended for scientists and researchers who wanted to share computers remotely. But within two years, the network's users had turned it into something unforeseen, a high-speed electronic post office for exchanging technical and personal information.

By 1989, ARPANET became the internet, which was also described, of course, as the information superhighway.

Meltem Demirors
So as you may have guessed by now, today, we're going to talk about networks. And specifically we'll talk about the architecture of networking technology, the evolution of networking hardware, software, and models in the context of, guess, cryptocurrencies. This is a big and meaty topic, so we're going to do our best to provide an introduction to the technology. And we're really going to focus on mesh networking today, or dynamically self-organizing and self-configuring networks.

Jill Carlson
Let's get meshy, Meltem.

Meltem Demirors
Meshiness, I love it. So before we delve into meshing, I want to talk about how communication works today. And this is something I've always loved reading about, learning about. And relearning it for this episode was really fun. So let's get a little dorky and delve into these mechanics just so we understand how this works and why it matters. Will you indulge me, Jill?

Jill Carlson
I'm stoked. Let's do this.

Meltem Demirors
All right. So I'm sitting at home right now in Brooklyn, New York. And so when I came home today and connected to my Wi-Fi network, what actually is happening behind the scenes? Because the internet's not this magical, ephemeral thing, even though it feels like it. There's stuff happening that makes it possible. So-

Jill Carlson
It reminds me a little bit of back in the '90s, right, what you would have to plug into your phone line and your mom would get mad at you because you were kicking her off the phone and dial-up…

Meltem Demirors
Oh, yeah.

Jill Carlson
It used to feel like much more of a process to connect to the internet. Now today, that's all abstracted away.

Meltem Demirors
Totally. I still remember so vividly my mom yelling at me to get off the damn internet so she could use the phone. And then we were one of the cool houses that got two phone lines.

Jill Carlson
Oh, fancy. All right. So today, though, what's happening behind the scenes when you connect to your Wi-Fi, Meltem?

Meltem Demirors
Okay. So if I'm on my phone, what's really cool is my phone has a radio on it. And what my phone does is it takes chunks of data that I put into my phone, like a block of text from a text message, and it'll take this chunk of data and turn it into a radio signal, which it transmits. Now, everyone in their house usually has a router or some sort of physical piece of hardware that's plugged into the wall. And that router has an antenna and receives these radio signals and decodes them into packets of information.

Now, what's really cool is most routers have two radios on them, one which receives 2.4-gigahertz waves and one that receives five-gigahertz waves. And more professional routers will actually have multiple radios. And so the 2.4-gigahertz wave form is longer. It's for longer range. But it's slower because the wave form is longer. And then some of the newer Wi-Fi has five-gigahertz radios. These waves are smaller, more compressed, therefore, travel faster. And so sometimes when you're having trouble connecting to a printer or you're having trouble on your network, you'll switch between the two. Most routers now switch automatically between these two radios. But it's pretty interesting.

I finally actually figured out through researching this episode why I can never connect to the printer in my office. And I fixed it. So professional win for me, Jill.

Jill Carlson
That's hilarious. I'm so glad something is coming out of this episode already.

Meltem Demirors
Such a dorky thing. I'll stop.

Jill Carlson
Okay, great. So you may have heard people talk about TCP and IP. And in the Bitcoin world, people often compare this to BTC and then someday perhaps LN or LND, right, for Lightning...

So let's go back to the beginning. TCP, the transmission control protocols, these are used to direct packets of information, like you said, to specific applications on a computer using a port number. IP, internet protocols, are used to direct packets to a specific computer or server.

And our router connects, just to complete the alphabet soup here, to an ISP, an internet service provider, which gives us access to the internet through a range of technologies. A single device is assigned an address when it connects to the internet, an internet protocol address, an IP address. This address distinguishes our device in the network from all the other devices. It's kind of like your router's unique address.

Meltem Demirors
Yeah, and I think that's important because people throw around TCP/IP all the time. And I think a lot of us don't know really what it means. But basically the reason that we can communicate, can message, TCP/IP are these standards that dictate how all this data flows through the network. So it's pretty important. However, what's really interesting is ISPs, or these internet service providers, which consist of fiber optic cable and hard wire, these networks were really local. In the '90s, ISPs were limited to local areas. And over the last 20 years, a bunch of local, smaller ISPs have gotten cobbled together into national ISPs and this global network called the internet.

But our national ISP can only connect us directly to servers located in our country. If I want to connect to a server located in a different country, then I would need country's ISP or an ISP located in that country to connect us to those countries' servers. And so again, what I think's really cool here, to us the internet feels seamless. When I go to a site that has .co.uk at the end, like my CoinShares website, it's seamless. But in reality, what's happening behind the scenes is all of my data requests and the data packets coming back are routed via different local ISPs back to my local ISP so that I get the data.

So this large global network is really cobbled together using these internet protocols and standards like IP addresses and transmission control protocols. Pretty cool.

Jill Carlson
It's fractal. It's fractal, right, Meltem?

Meltem Demirors
Totally, Jill, totally.

Jill Carlson
So in the digital world, every connection requires our device to send these packets of data that we keep talking about, which is you can think of it as similar to sending letters by post or by mail. In both cases, we need an address, a system which handles our letters, and a letter box. On the internet, an example of an address we might want to quote/unquote send letters to is let's say www.MyShadow.org. Now, similar to a post office, our IP address helps direct our letters, while the TCP, the transmission control protocol, disassembles and reassembles our letters into a single port. And this, again, in this analogy that I'm trying to draw can be compared with a post office box.

So here's a really simple and simplistic example. If I'm in San Francisco and I want to send Meltem a message on Twitter while she's in, say, London, because that's where she works a lot of the time, let's run through what would happen.

Meltem Demirors
Let's do it, Jill. I'm in London. I'm actually going to be in London next week, so this example's not so far-fetched, is it?

Jill Carlson
There we go. Perfect. So my phone would translate the text into a radio wave and relay it to my San Francisco office router. My router will receive that message, again, via radio wave and decode it to turn it into a data packet. And that routes it to my local U.S. ISP. That U.S. ISP then directs it out of the U.S. network and through its gateway to a U.K. ISP, where it connects to Twitter's U.K. servers.

Meltem Demirors
And then my system repeats the process in reverse, so the packet gets routed from my U.K. ISP to my office router using the IP address of that router as the identifier. And then the router will take and receive that data packet, turn that into data expressed as a radio wave, and remit it back to my phone via the radio in my phone.

So one simple interaction, which to us feels seamless and instant, actually has a lot of different moving parts and a lot of layers. We have our phones, which are pieces of hardware. And they have their own software and protocols they communicate on. We have our local networks, or the Wi-Fis we use that are controlled by hardware, again, these routers and their local networks. We have our regional ISPs. So Jill has her ISP in San Francisco. She's on the West Coast. And I have my London or U.K. ISP. And then there's Twitter's servers.

Now, these different layers are often referred to as part of the network topology. And we're going to get into that a little later.

Jill Carlson
And so our device's data, such as our IP address and any browsing cookies, metadata, etc., they travel through so many nodes and layers in this network and the networks of networks, which means that we can be tracked all along the way. In other words, when we access a website, all the intermediary parties included in the network are aware of it.

Meltem Demirors
So let's talk about why this matters. This architecture we just described makes our communications and online behaviors susceptible to a number of different vulnerabilities. First of all, we're relying on a multitude of different types of hardware and physical computational networks, phones, routers, ISPs, servers, switches. So there's a risk of hardware failure. Second, there is a risk of network failure. So if one network breaks down, how do you reconstitute it? There are man-in-the-middle attacks, so if someone's spying on part of our message, they could intercept it, potentially read it, but also change it, send it onward, reroute it.

There's a loss of privacy. So if my IP address ties me to a physical location and I have location services on, Twitter and everyone else in that communication knows where I am. And then throttling also happens based on consumption patterns. This happened a lot when I started streaming content back in 2012. If you were consuming too much data on your local ISP, you would actually get your service throttled back because your IP address was consuming too much network bandwidth. This also, by the way, happened when I plugged in my Bitseed, which was the first sort of Bitcoin node you could buy off the shelf. In 2015, I plug it into the wall, and I would immediately get throttled and get angry emails from my ISP.

And again, because all of this is controlled by centralized entities, they can apply their own routing hierarchies and decide what messages get priority, how they get routed, and many, many more.

Jill Carlson
Yeah. And so now, the reason things developed this way and in all of these layers is because centralization, let's be honest with ourselves, has a lot of benefits. It enables efficiency and planning where to put towers, where to lay wire, where to build servers, etc. And private companies can finance it and charge for access to it. The idea of mesh networking or changing the topology of networks is to flatten the architecture. And this idea's been around for a long time. It just hasn't been as fast or as cheap. And the right tools haven't existed to incentivize people to contribute resources to this kind of a decentralized network, until maybe now.

Meltem Demirors
And here's my take on this. And this might be controversial. But in my view, Bitcoin is, first and foremost, a privately owned and privately operated telecommunications network where network operators are incentivized to engage in proof-of-work mining in exchange for Bitcoin, an incentive and a fee for service provision in terms of mining fees. At times when demand for network capacity is high, meaning when the mempool's full, sending transactions becomes more costly. And we see this because transaction fees on the Bitcoin network go up. Same thing with Ethereum.

My firm, CoinShares, actually has done a lot of extensive studies on the mining landscape to explore some of the economic drivers driving this behavior in miner responses to this. And what's really interesting about this is in the Bitcoin network, your public wallet address is kind of like your IP address or your identity on this new telecommunication network.

Jill Carlson
And so, Meltem, you and I talk a lot about privacy. We like to say girls just want to have privacy. And that's super relevant to this topic, right, this issue of can you hide what you're doing from your ISP? And the reality is, is it's tough. They still know your IP. Even if you're using a VPN, there are easy ways for the ISP to know what you're doing online. And it's very difficult to hide your activities from your service provider. And a lot of the details are buried of course in your T&C, your terms and conditions of service.

So centralized communications networks, decentralized, two of our favorite buzz words. We are of course going to talk about and unpack what a decentralized network looks like. And that brings us to mesh networks. Do you want to dive in, Meltem?

Meltem Demirors
I love talking about this. The word "decentralized" still grinds my gears. But I'm going to allow it today, Jill. We're going to go with it.

Jill Carlson
Okay. So first, let's just talk about the environments in which you might want a decentralized communications network, a mesh network, because as I mentioned, centralized networks actually have a lot of benefits. But there are two situations in which you really might need a decentralized network here. And the first of those two is if you're in a high-density, low-connectivity environment, so for example, if you're in a city with lots of people or at a massive festival with too many people competing for a signal and lots of devices everywhere. I know I've certainly been in the situation where you're trying get a hold of your friends and you said to meet up at this stage at this time. You can't find each other, right?

And that's a problem because you're all fighting for the same resource. In this case, you're all trying to connect to the same single cell tower in the area, or you're all at one coffee shop if you leave the festival, right, and you're at a popular coffee shop. You're all trying to use one Wi-Fi connection, whatever it might be. So that's an example of high-density, low-connectivity.

The other example of an area in which you might want a decentralized communications network is low-density, low-connectivity. So this is my about remote, far-flung locations, if you're out hiking, if you're in the playa, although I guess the playa would be high-density, low-connectivity…still or no connectivity. But so then…companies like Filament, and we'll get into this in a little bit, creating hardware devices to create mesh networks for these types of locations as well.

Meltem Demirors
Yep. And the field I used to work in before this, oil and gas, we had tons of pipelines that we’re in the middle of nowhere. How are you going to monitor a pipeline? It's really expensive to monitor thousands of miles of infrastructure that runs through the middle of nowhere. Well, Filament, again, is a great example of trying to figure out how to use new types of networking models to send data in places where there's not a lot of connectivity.

So first, though, we're going to go back to something we mentioned earlier. To understand how meshes work, we really need to talk very quickly about network topology. So, Jill, have you ever looked at a topology map?

Jill Carlson
Yeah, sure. Well, okay, I'll clarify. I've looked at physical topology maps.

Meltem Demirors
Right, like a hiking map, right, where it has little circles and it shows you the elevation. You can kind of-

Jill Carlson
For back-country skiing, this is critical, right? Because you need to know what elevation gain, what the steepness angle is going to be, things like this.

Meltem Demirors
So when we talk about topology, really in the physical world, that's the layout of terrain. But in networks, there are two types of topology. And this really pertains to how the network is constructed. So physical topology, as the word implies, it pertains to the various components of the network that are hardware, so devices, their location, cable installation, etc. And logical topology illustrates how data flows within that physical network.

So for my example, the physical topology of Jill and my interaction that we described earlier is determined by the capabilities of the network, our access devices, the media type we're communicating, the level of control or fault tolerance of the systems we're using, the cost associated with cabling and the telecommunication circuits we're using. There's a transmission medium that's used to link our devices to the physical topology of the network. In some cases, in my apartment right now, I have fiber optical, so fiber in a lot of cases, it could be plain-old Ethernet cables. It could be something else.

And then-

Jill Carlson
A different type of ether.

Meltem Demirors
Yes, a different type of ether. That again, in a lot of these physical topologies, you have centralized parties who plan the location of these nodes, the links between them, the cabling, etc. So the physical topology's going to be really important because it dictates some of the hardware requirements.

In sharp contrast, the logical topology, pardon, is the way that signals act on this physical network or the way that data's going to flow from one device to the other without regard to the physical interconnection of the devices. And what's important here is while physical topology may dictate a certain type of design, the logical topology doesn't necessarily have to follow. And this becomes really interesting because some networks and in particular meshes, can actually dynamically change how the logical topology or routing works through configuration changes. But they don't necessarily require physical changes to the network.

So it's important to just separate there's a physical component through hardware and a logic-driven component, which is really the protocols that dictate how routing works.

Jill Carlson
Pretty wild. And just layers upon layers. I'm going to keep using this word here, but feels fractal to me. So a mesh network, it's important to remember, this is a local network topology that we're talking about. And it's a local network topology in which the physical infrastructure, like the nodes, the bridges, the switches, and other devices, they connect directly, dynamically, and non-hierarchically, so it's a flat structure, to as many other nodes as possible and cooperate with one another to efficiently route data to and from clients.

This lack of dependency on one single centralized node, this is what allows every node in a network to participate in the relay of information. Mesh networks dynamically self-organize and self-configure, which can reduce installation overhead. So we love to talk about how centralization is always more efficient, blah, blah, blah. Well, it depends on what kind of efficiency you're actually talking about in a lot of ways. Here is one example in which a mesh network will actually have less overhead.

Now, the ability to self-configure, this enables dynamic distribution of workloads as well, particularly in the event that a few nodes should fail. So you can think of a mesh in two key components, right? You have the hardware, again, as you laid out, Meltem, the physical topology, and then, two, the logical topology or the software that dictates how relaying of packets between nodes will work.

Meltem Demirors
I just love this. I think it's so fun. If you're interested in topologies, information theory is a really interesting area of study. And one topic I like to read a lot about is hyperspheres, which is multidimensional spheres. They have an important impact on how messages are relayed and how communications over really long ranges might work. That's a nerd-out topic. Wolfram Alpha, which is a math website, yes, there are math websites, has a lot of great stuff on that topic, information theory, hyperspheres, and graph theory-

Jill Carlson
We get it, Meltem. You're a nerd.

Meltem Demirors
I know. I'll stop. But in future episodes, I hope we'll also get to discussion fog computing, which is kind of like mesh networking in a way but a little different. There are a lot of really cool extensions of cloud computing that are focused on enabling infrastructureless compute, which means compute without having all the servers and other cool innovations and hardware. There's also serverless through innovations like Amazon Lambda. But anyways, we're going to go back-

Jill Carlson
Oh, God. I'm going to have to dust off my textbooks before that episode. I feel like I'm out of my depth here.

Meltem Demirors
We'll see. We'll see if people like our meshiness. If you guys, as our listeners, don't like it, maybe we won't. Anyways, let's get back to the mesh. I think this is really interesting. I want to delve a little bit into the history of mesh because like all things, what is old is new and what is new is old.

Jill Carlson
There's nothing new under the sun.

Meltem Demirors
That is right.

Jill Carlson
Now, like many innovations, mesh networks are by no means new. Mesh already exists on a localized basis as Wi-Fi networks.

Meltem Demirors
And there are also a lot of devices that allow you to run meshes locally on your Wi-Fi network. And the use case here is say you live in a really large house or say you have big property or big office. Now, I don't know what any of these things are like because I live in New York, where that is not a luxury we have. But say you have a really big house, you have one central router, and the reach of that router, that radio is not wide enough to cover all of the space in your home or on your property. So you end up with a lot of dead spots.

And so what these local meshes do is they allow you to have one central hub and a bunch of little mesh devices that connect to that central hub so that you can cover and have connectivity from that central hub extend through a local network of connected devices without needing cables connecting to the Ethernet or needing multiple core routers. So let's run through them. These are all VC darlings.

The most famous one is Eero. Eero raise ton of venture capital money, super hot. They consist of a home port that you plug in that's like your router and then these Eero beacons, which kind of look like Glade air fresheners, and you plug them into the walls around your house. And they help you build a faster, bigger home Wi-Fi network. It's kind of like Apple did this actually. Their AirPort Express was basically this but 10 years before its time.

So Eero was acquired by Amazon last year, $97 million. And these things are still really popular. Now, Google of course had to do their own. Google Wi-Fi is basically Eero's well-capitalized competitor. It's very similar. Another one you may have heard of is Plume. Plume is another router device that has local nodes for your house. What Plume really does well and what they're known for is they have this really cool software app on your phone where you can actually connect and interact directly with your router. And you can apply parental controls, manage device access like for guests who come to your house, and monitor the use of your network in real time.

So, for example, if all of a sudden, you're working away and your Wi-Fi's really slow, you can see, "Oh, my gosh, my husband's streaming a ton of content in the other room," you can kind of try to balance loads internally. And then Disney, the content company Disney, even got into the game. And they rolled out a device called Circle that lets parents control their home networks, apply parental controls. But what Circle did that I think's just really funny and also speaks to the dystopian future of networks is you can actually with this device freeze Wi-Fi for certain hours.

So, for example, if you have a teenager in your house and you know that they're going online and doing weird stuff from midnight to 2:00 a.m. every night, or if you don't want them to be online and you need them to do your homework, you can actually freeze your Wi-Fi.

Jill Carlson
So, okay, I have a little rabbit hole I want to go down on this. Is that okay, Meltem?

Meltem Demirors
I love rabbit holes. I thought you'd never say that.

Jill Carlson
Okay. So are you familiar with the phrase "gaslighting"?

Meltem Demirors
Yes.

Jill Carlson
Okay, so gaslighting is this idea that someone is manipulating you and making you feel like you're the crazy one when in reality, it's them who's the psychopath, right? And this phrase comes from this. It started out as a play, and then it was remade into a movie-

Meltem Demirors
I was going to say, did it not come from Crypto Twitter?

Jill Carlson
It did not come from Crypto Twitter, no, although it is definitely easily applied to Crypto Twitter. But it comes from this play in which a husband gaslights his wife. He manipulates her into thinking that she's crazy. And one of the ways that he does this is he turns down or he goes up in the attic and he turns on the gas lamps in the attic, which then turns down the lights in the rest of the house. And the wife will say to the husband, "The lights keep flickering in here. What's going on?" And he says, "You must be crazy. Nothing is happening."

And I have this Black Mirror episode in my head in which someone, the 2019 version of gaslighting, right? Someone comes along and freezes your Wi-Fi and then convinces you that you must be crazy. It's just you to whom this is happening, or starts censoring what you can do. And I've had this idea in my head now for a few months. And so the fact that this is coming up in this show, this has got to become a Black Mirror episode at some point.

Meltem Demirors
This is the point, Jill, is all of this already exists, and people-

Jill Carlson
It's totally possible.

Meltem Demirors
... are using it. As I started, again, researching this episode, as I started thinking more and more about what goes on in my home and just how many devices are connected to my Wi-Fi and how many trackers are on my network, it really started freaking me out.

Jill Carlson
100%, 100%. Okay, so we'll return to what we were talking about here-

Meltem Demirors
No more gas lamps. There are, by the way, gas lamps in my neighborhood in Brooklyn. It's like a historic thing. They have old-school gas lamps. So we'll close it out, the full circle.

Jill Carlson
Yeah. Well, watch to see if they ever flicker or get turned down when you walk by, Meltem.

Meltem Demirors
I'll call you, Jill, so that you can commit me to the crazy, to the loony bin, crazy house.

Jill Carlson
Okay. So back to mesh networks, these meshes rely, the meshes that you're talking about here, so Eero, the Google Fi, Plume, Circle, they rely on a connection to a single master node, which is the router, which is plugged into the Ethernet or the fiber optic cable you have in your house. But so the problem with these is that you're still at the mercy of an ISP, right? There is actually a central party that could potentially gaslight you. Probably not going to happen, but I like to live in dystopia. And-

Meltem Demirors
Or more importantly, Jill, they could also man-in-the-middle attack you, meaning they could intercept your communications and alter them, right?

Jill Carlson
Again, probably not likely, given the ISPs that we have in the United States, but a possibility. But so despite the many benefits of these types of mesh networks, they're still problematic. And mesh networks as a whole, they're still niche. And so this is partly because connecting to a mesh network is still far more difficult than just signing up for internet service by an ISP and paying a monthly internet bill.

So, okay, let's bring it back to the main question of this episode, or one of the main questions. How do you build a true mesh network that will connect devices directly to each other without relying on the central master node? Because the goal for a mesh is to keep all the components equal, decentralized, non-hierarchical. So the only way to shut down or disrupt the network is to shut down every node in the network, which makes it of course much more resilient to interference or other disturbances.

Meltem Demirors
And so in more practical terms, what ends up happening in a mesh is people set up specially configured routers that are known as nodes that connect to other configure routers, or nodes, and then allows users to locally create a network that is physically distinct from the internet. And what they're relying on in many cases ... You know how I talked about the example of how my phone talks to my router in my house using radio waves instead of traveling through cabling?

Jill Carlson
Mm-hmm (affirmative).

Meltem Demirors
That's what they're doing. You're relying on radio waves and radio bandwidth to relay packets of information. And although you can connect to the internet through mesh, you can also create it as a local network independent of anything else. This is actually how Bluetooth works, right? It's just radio waves. You're not relying on any cabling.

Jill Carlson
So I remember when my dad first came home, I want to say this was probably in 1999 or 2000, and he was like, "You guys, Bluetooth, Bluetooth is going to be the next big thing." And I remember my mom rolling her eyes, my brother and I being like, "What is a blue tooth? What is he talking about?" And for the longest time, this was actually a running joke in my family about Bluetooth because it wasn't until really seven, eight years later that we saw consumer products that really actually worked using…

Meltem Demirors
And the first one, by the way, was Apple's AirPort. If you remember this, I had an Apple AirPort right after college. And it was amazing because you could send stuff back and forth without being on the internet. So again, these things have existed before, but when you don't have all this physical cabling that's been invested and all of this capital expenditure that companies have to lay out to build this physical topology, all of a sudden, a lot of new things become possible.

But the problem with these things is, as you and I know, radio waves are not that strong. This is why you get dead zones in your house when you don't have a strong-enough router. And also in the U.S., air waves are regulated by the FCC, right? And so retail communication devices are not allowed to exceed a certain bandwidth or certain strength. Only government devices can have really high reach, which is really interesting. So then what you need is, if you have all these local devices and they're all interacting using these radio waves, you need antennas, right? They're installed on the outside of buildings typically.

And these antennas have stronger signal. And so what they can do is take what's happening in a localized area. So say, for example, my neighborhood in Park Slope, here in Brooklyn, we can have a local mesh network. And then we can beam and connect to other networks using these stronger antennas. And effectively we can connect our little localized network to another localized network, forming a larger mesh. It's pretty cool.

Jill Carlson
Okay. So most users on mesh internet, they still are depending on a traditional ISP to connect to the web either by their own subscriptions or a connection that is shared by another node. But decentralization, which makes mesh more resilient to interference, this also means that connecting to the internet through many hops of mesh is slow. And the real solution is combining mesh technology with direct access to…

Meltem Demirors
And so, Jill, this is actually my biggest criticism of all of these things, Bitcoin, these…

Jill Carlson
Does it grind your gears?

Meltem Demirors
It grinds the F out of my gears. And I don't want to curse, but it grinds my gears so damn hard because people sit and look in my face and be like, "What we're doing is decentralized, and power back to the people." And I'm like, "You are connecting to an ISP. Oh, my goodness. What you're doing, you're basically taking all of this F'ing data and putting it on a network that is deeply, deeply insecure." And so all of this talk about how amazing and resilient Bitcoin is, if you're relying on ISPs to get access, you are screwed, my friend.

Jill Carlson
100%.

Meltem Demirors
It grinds my gears so hard. And, look, that's not to say that right now this isn't a huge constraint. But if we don't design for this, then we will never fix it, and the system we have will be totally vulnerable. And it grinds me that nobody thinks about this or talks about it.

Jill Carlson
Well, some people do. And we'll get into them in a few minutes. But I totally agree with you that this is completely underestimated. And you know I've done all of this research on Venezuela-

Meltem Demirors
Yeah.

Jill Carlson
... and the closed economies and cases like it. And there you see it in action, right? All it is ... No, I mean forget the fact that in general, people are accessing Bitcoin via centralized service providers and all the local government has to do is just IP-block that website. But people don't think of the other layers at which things get centralized and, therefore, censorship can happen because they're so distracted by the censorship resistance with the blockchain. There are many other layers of infrastructure we have to think about here.

So, okay-

Meltem Demirors
I know…

Jill Carlson
...get back to mesh networking…

Meltem Demirors
So many rabbit holes today.

Jill Carlson
…It's good. Okay, so mesh wireless, there's this guy Dan Phiffer.

Meltem Demirors
Phiffer.

Jill Carlson
Phiffer, right? Phiffer. Oh, my gosh. Okay. I've only ever seen it written. So he's a prominent coder and a free internet advocate. He built a darknet for Occupy Wall Street. He's done a whole bunch of other things. And he talks about mesh wireless as occupying a place in the public imagination that may not always sync up with the boring reality. And I think that is so true. And he adds that there are several components to mesh that get conflated, but that mesh technology, which allows for devices of course to connect to each other peer-to-peer, and the peer-to-peer community possibilities can be understood separately.

Additionally, it's not either of these capabilities alone that allows for users to connect to the internet. And so we can run through a bunch of examples of existing mesh networks and see how all of these play out.

Meltem Demirors
And I think what's really being pointed out there is that there are a lot of ... Again, when we talk through examples earlier in the episode of how communication works and all of the layers and architecture and interacting pieces, I think what he's really trying to emphasize is there are a lot of different components and layers, whether it's hardware or software protocol standards that go into enabling networks to connect. And so let's talk about it.

So existing mesh networks, I'm a big fan of NYC Mesh here in New York. It's a community-run internet network that has supernodes, which are these large antennas that are installed on buildings in Manhattan and in Brooklyn. I've actually been looking for a while at running a supernode myself. Unfortunately they cost about $5,000 to set up and $1,000 a month to run. And I don't-

Jill Carlson
Oh, Meltem, you can sell some of your shitcoins and afford..

Meltem Demirors
But we'll get into this. But there is no way to monetize that. It's really just a contribution back to the NYC Mesh community. But what these antennas do is they connect directly to internet exchange points, (singing), IXPs, not ISPs but IXPs, through fiber optic. And thereby, they bypass these ISPs. But supernodes basically are interspersed throughout the city to provide internet access to these localized meshes. And they also have community-installed routers. So if you're roaming around, you can connect to NYC Mesh.

Now, one thing I'll note that goes back to why meshes in reality are much harder to build is there are a number of different competing meshes in New York. And none of them interoperate. So you have NYC Mesh. There's a pretty big goTenna mesh community in New York. There are a few other localized meshes. And it's kind of interesting to even see at a local level some of the challenges around coordination when things aren't really heavily centralized. So I'll just point that out because I think it's..

Jill Carlson
Right. But so it's not just sort of hobbyist communities that are toying around with this. In January 2009, in Jalalabad, which is a major city in Afghanistan, they wanted to bring high-speed internet to a village, hospital, university, and an NGO in the city. And so they implemented something that they called Fab Fi. Jalalabad now has 45 remote Fab Fi nodes with the longest length spanning about six kilometers. And the system is extensible by anyone without getting central permission, which is so cool, simply by adding a node and pointing it at the right direction.

The materials needed to make a new endpoint link cost only about 60 U.S. dollars. And they're available locally in Afghanistan, in Jalalabad. Now, an endpoint node that provides 360-degree coverage of the area near it, that might use a bit more hardware. But it still only costs about…

Meltem Demirors
Yeah, and I think the example of Jalalabad is a really cool one because it was built with a bunch of research institutions. And really what they've done after they got the system up and running or seeded it, and we'll talk a little bit about the challenges of seeding meshes or doing what's called a zero launch basically, what was cool was once they seeded the network and set up the first few nodes, the city actually, this initiative Fab Fi, they actually ran workshops to train local people on how to manage the network and set up nodes.

And so it's kind of a cool way to think about it as a mix of kind of public and private sector where a lot of it was also focused on really enabling locals to manage the network and not requiring the central entity who would go around and set up nodes. So I think that's a really, really great example.

The last one I'll bring up is actually a really interesting one because it has some political implications. And this is Guifi. Guifi is a countrywide mesh network in Spain. But it's especially…in Catalan villages. And if you know anything about Spain, Catalan is an area that historically has been really politically active, and they tend to not always agree with what's happening in the rest of Spain.

What's really cool about Guifi is they have 34,000, pardon, active nodes and over 50,000 kilometers of wireless linkage. And so I think this is a pretty cool example of a countrywide network that has reached sort of adoption and scale, and it's also slightly politically motivated. So it's just interesting to see how these things fit together.

Before we delve more into meshing, I want to take a quick moment in this episode to thank someone. Dan Onggunhao, his Twitter is Onggunhao, and we'll put it in the show notes, he was so awesome. He reached out to me in Twitter and said, "I love the show. How can I help you with research?" So the sections you're about to hear he helped us with a lot of this research and putting it together. So, Dan, major shout-out. We'll give him a shout-out again at the end. And again, his name will be in the show notes. He's on sabbatical right now. He's getting really into crypto things. So reach out to him on Twitter. Ping him. Give him a big thanks. Dan, thank you so much.

Jill Carlson
Heck yeah. Couldn't do it without you. So, okay, enter Bitcoin. Let's talk about incentives in mesh networks. You ready, Meltem?

Meltem Demirors
Jill, I've never been more ready for anything.

Jill Carlson
I know. You've literally been talking about doing this show I think for a year since we started.

Meltem Demirors
I know. I'm obsessed with this topic. And so I finally had time, I spent all weekend reading all this weird stuff. Dan helped a bunch. I gave him a long list of things I was ... I basically dropped my entire Evernote on meshes into an email to Dan. And he just went and put it all together and was like, "Here's all of the cool facts." I was like, "Yes!" It was great.

Jill Carlson
God bless, God bless.

Meltem Demirors
So good, so good.

Jill Carlson
All right, so mesh networks they face a number of challenges that have historically been really difficult to resolve, right? So let's run through a few of these. The first one that comes to mind for me is identities. Traditional identities don't work on a mesh. This is a problem that recurs throughout all kinds of decentralized systems. And so we need a new way to identify people. How do I find Meltem in this network?

Now, one proposal that has come to light around this is, well, what if we used cryptographic wallets as an ID, like a public-private key pair? And so-

Meltem Demirors
Right. That's like a replacement for an IP address, right?

Jill Carlson
Exactly. And so here we start to see the intersection of something like Bitcoin with mesh networks. Now, another issue is if I relay for other people with my node, how do I manage who is consuming resources and how much? So mesh has been around for a long time, as you've said a few times, but it hasn't taken off because there's been no payment system around it. You would have to rely on some PayPal or credit card, and then you're back into the realm of centralized entities.

And with Bitcoin, again, we can start to imagine a world in which we can have new forms of incentive occur and decentralized digital money that works on this network. Peter Van Valkenburgh actually wrote an article on this that I think was published in Wired four or five years ago. So that's worth going and rereading if you're interested in this area of incentives. And we'll get into it a bit more.

Now, another area, again, in the realm of incentives is the tragedy of the commons. So how do you basically throttle people who are consuming too much or interfering with the network. You need to make it prohibitively expensive for them to do so. And then finally the last problem that comes to mind for me anyway is how do you enable handshakes and handoffs between these localized mesh networks?

Meltem Demirors
That's right. And that's one we see often. And so then you really get into the needs of the network. So as we talked about with topologies, there's the physical topology and the logical topology of a network. In terms of physical topology, you need compatible hardware. As we've talked about repeatedly in this episode, the radios we have today don't really work that well. And device bandwidth is pretty limited. So unless you have a lot of money to set up antennas and other broadcast-worthy sort of infrastructure, it's going to be pretty expensive.

And this is something that's called the zero-start problem in mesh. How do you get enough people on the network to make it useful? How do you get enough devices connected? And then how do you build some of that core infrastructure if you don't have a way to bootstrap it? You can't have people PayPal you. If I want to set up this giant antenna on my house to support NYC Mesh, I have no way of paying the $1,000 a month other than just paying it out of pocket. So that's interesting.

And then the second piece is compatible software and open standards. And this is really about some of the logical topology, some of the routing. So how do you build tools and standards so that developers who are going to build stuff on top of these networks ... If I want to have a messaging app that's highly secure and leverages a mesh, I need a way for people to build on top of it. And the only way to do that is to have widely accepted standards. This is what made a lot of innovation on the internet possible. It was standards like TCP, IP, FTP, or file transfer protocol, and other things that allowed different email clients to interact with one another regardless of who developed and built them.

It allowed different routers and devices to communicate with one another. So in absence of sort of a centralized entity that's going to coordinate those standards, it's kind of hard to see how that's going to get built by a bunch of these localized kind of smaller groups.

Jill Carlson
Right, exactly. And so let's bring it back to Bitcoin here. So there are two main advantages of Bitcoin that kind of interact or parallel with mesh networks, the first of which we touched on already, the censorship resistance of the blockchain itself. Again, that's not to say anything else about the infrastructure around it, but censorship resistance, and then the permissionless nature of my ability to join. So if you're anywhere in the world, as long as you can run a node and connect to the global network, you're on board.

And that, again, that parallels a lot of what we see in mesh networks. And we'll start to see the interplay here as we go through this conversation.

Meltem Demirors
So then let's talk about censorship resistance, right? These networks, blockchain networks, like the Bitcoin network, they're not just used to store data about transactions. They're also used to store text and other information that can't be deleted or changed or edited. And in fact one of the things that's so valuable about the Bitcoin network and the Bitcoin blockchain or this ledger is the fact that you can have a timestamped record of data. And over time, that data gets crystallized. Nick Szabo has a great example of a fly getting trapped in amber. And as the years go on, that amber gets hardened and crystallized, and that fly is preserved with more and more certainty.

And similarly, something interesting happened in April of 2018. There was a student at Peking University who actually ended up submitting a transaction not to the Bitcoin network but to the Ethereum blockchain where she detailed intimidation and threats that had been made against her by the school and her attempts to investigate claims of sexual assault. And so she'd earlier submitted this information on WeChat, where it was widely shared. And then censors in China actually purged it from the platform.

And so she decided to use-

Jill Carlson
God, this is so messed up.

Meltem Demirors
But isn't that kind of crazy that this student was like, "Hey, I'm going to submit this data to a blockchain network because the Chinese government can't purge it"? In fact-

Jill Carlson
Yeah…

Meltem Demirors
... there's a really cool thing that some people do when their kids are born. And people use it for all types of stuff, but they'll actually timestamp that record to the Bitcoin network…looks kind of dorky. And there's actually service called OpenTimestamps I think that does this but with the Bitcoin network.

Jill Carlson
Yeah, yeah. And so, okay, it's not just about censorship resistance, though. There is also this notion of redundancy around Bitcoin permissionlessness. And so let's talk about the Bitcoin relay network for just a second. This is a high-speed block relay system. And it helps relay blocks around the world and acts as a fallback in case the public Bitcoin network encounters issues. And then we also have FIBRE, or the Fast Internet Bitcoin Relay Engine, which is a follow on to the BTC relay network, which relays blocks with no delay beyond the speed of light through FIBRE.

This allows the Bitcoin network to operate incredibly well even when faced with suboptimal internet conditions. And so anyone in any country can easily join the Bitcoin network and be able to participate in it on par with first-world country nodes with strong internet connections, sometimes actually even better.

Meltem Demirors
Yeah. And I think these are really cool examples. The third example is the Blockstream satellite. And I love this example. I'm such a loser for loving this. But basically there are a bunch of satellites orbiting around the Earth that assist with various telecommunications functions. In fact, your phone is constantly pinging with satellites to triangulate your location. Your cell phone is actually a GPS tracking device, FYI. So turn off location services, everyone. Have some privacy.

But censorship resistance with the satellite. So here's a weird idea. So say that governments want to shut down Bitcoin, and they try everything, packet inspection, throttling connections, shutting down connections. And some day, they're just like, "You know what, screw it. We're going to shut down all ISPs. No internet for anyone."

Well, guess what. There's a satellite dish in outer space that allows you to receive data blocks from the Bitcoin network, so you can get of data from outside the traditional ISP system. So Blockstream leased capacity on a satellite, and they call it the Blockstream satellite service. And they're currently broadcasting real-time Bitcoin blockchain data to two-thirds of the Earth.

They're currently using commercial satellites, as I mentioned, where they're leasing capacity. But in the future, who knows, maybe they'll launch their own satellite into space, or a consortium of companies will get together and launch one. As you may know, Jill, ConsenSys, the Ethereum development studio, did in fact acquire the assets of Planetary Resources after they went bankrupt, the space mining company. So you never know.

Jill Carlson
Oh, my God. This is all a lot for me. Even me with my tin foil hat collection, this is a lot.

Meltem Demirors
I know.

Jill Carlson
But, okay, all right. So in 2015, BlockSat, which is a test of the proof of news run by a pseudonymous Bitcoiner who goes by SafetyFirst, they started broadcasting bundles of news using the Blockstream satellite. So you can check out BlockSat as an example of how this gets used.

Okay, and then Push Transaction, or PushTX, is an SMS service. And so users in censored areas can push transactions to a Bitcoin node via SMS through transaction relayers who help to publish transactions onto the Bitcoin network. And this, again, is still leveraging the Blockstream satellite and is another example of this is me basically trying to play devil's advocate with myself that this is something that people actually want.

And then finally, there's the Blockstream satellite API, which allows users to pay with Bitcoin to broadcast data to the world through the Blockstream satellite. And it leverages of course, our favorite, the Lightning Network.

Meltem Demirors
And I don't recommend following the Blockstream satellite's Twitter feed. It's basically people paying to submit messages via that Twitter feed. So it's interesting. It's kind of like-

Jill Carlson
Yeah, again…is this something…

Meltem Demirors
Do you remember Satoshi's Place, which was super popular when it first came out?

Meltem Demirors
Yeah, so if you're interested, Satoshi's Place, you pay one Satoshi per pixel on this giant screen. And so people would make all sorts of weird art and paint over other people's stuff. Dorkiness. We'll move on.

Okay, the thing I think is really cool, so I actually there's a small satellites conference that happened recently. And I really wanted to go, and I couldn't because I had to do stupid Bitcoin stuff. But CubeSats. CubeSats are really, really small satellites that are really easy to launch. And you can launch a bunch of them. And together, they can form these networks that allow for satellite communication but at really low cost. And they're not necessarily all the way in space. They're kind of high up in Earth's atmosphere outside of where planes fly. But they don't require a rocket to launch.

So CubeSats they're small, but they can carry something like a Raspberry Pi device and run a Bitcoin node. So in 2015, Jeff Garzik actually worked on something called BitSats where he wanted to construct 24 nanosatellites that would be able to process and fully propagate the Bitcoin network, which is super cool. A lot of people actually end up pitching CubeSats as a platform for space-based cloud computing. So in last week's episode where we talked about banning Bitcoin, one of the things we mentioned was when Pirate Bay, which was a peer-to-peer file-sharing site, got shut down or was being investigated by the authorities, they actually talked about potentially putting some of their servers onto drones and flying them around.

So similar idea. Space doesn't belong to anyone. So if you could launch these little CubeSats into space, that would be pretty cool. And actually the radio uplink-downlink time for some of these is actually going to be sufficient for small transaction sizes on Bitcoin network. Problem with CubeSats, though, the technology is still really new. A lot of it was actually developed at MIT. And it's only a few months sometimes before these things fall out of orbit.

And so small satellite technology's still emerging and developing. But I think it's pretty exciting when you start to step out of just the world of what's happening in Bitcoin and you think about the broader world of hardware and connectivity and some of the really amazing innovation that's happening, also commercial space exploration and commercial space development.

There's just a lot of really interesting things that are happening. And it's cool to see that they're actually getting connected. We may not talk about them a lot. They may not get a lot of attention. But I'm pretty excited about it all. I think it's so fun. It's like living inside of a sci-fi novel.

Jill Carlson
That it is. All right, let's cut. Okay, great. So we've just run through a whole bunch of ways the Bitcoin network and mesh networks kind of parallel each other. Now, let's get into what we were talking about earlier about the incentivization, excuse me, of network development. And this, again, is this idea that you can use Bitcoin, or another decentralized currency if you like, to incentivize people to spin up and run mesh nodes. And so Bitcoin is an example of this of how you might be able to use incentives to bootstrap network development, the zero-start problem that you were talking about, because block rewards, as an example within Bitcoin, they incentivize people to provision hardware and network infrastructure to run the decentralized mesh of nodes that run Bitcoin.

And so now, we can start to see how these ideas bleed into the bootstrapping of telco infrastructure, right, as an alternative to centralized ISPs. You can think about combining public Wi-Fi access points with an app-based micropayment system. And this theoretically could incentivize network development, i.e., have a profit incentive to invest and share in the infrastructure.

Now, of course there are practical limitations to this. There's the physics of telecom networks. They depend on large supernodes, for the most part, that are approximate to those IXPs, the internet exchange points. It's difficult to incentivize any kind of large capex investments here. And then there's also the fact that mesh, I mean it's composed generally of a lot of small devices. And it can be technically unfeasible once you start thinking about the bandwidth decreasing by 50%, latency increasing by about 10 milliseconds on each hop, the routing table gets really big. And suddenly, you're getting into a significantly less performant infrastructure than the centralized alternatives.

And then finally, of course with Bitcoin, we think about proof of work. Well, proof of work on a router, that gets really hard because of CPU limitations. And so proof of transfer becomes really difficult without compromising privacy. And so we start to get into these questions of what options are we left with? So is this just theory? Is this just a fun, theoretical idea? Or do you think that there's something real here, Meltem?

Meltem Demirors
I mean I think there's something real here. But I think what you just pointed out, this point around physics and bandwidth, look, I think we can't ignore the laws of thermodynamics, right? They're the limitations on the simulation we're living in, in this iteration of reality, if I'm allowed to go down that rabbit hole briefly. And I think one of the things that's challenging is there are real limitations to how fast we can communicate. And at the end of the episode, we'll talk about some really farfetched, sort of out-there ideas.

But let's talk about Bitcoin-related meshes that are actually being built today, because they exist. So there are two meshes that are being built leveraging Bitcoin. The first is goTenna. So goTenna is a company that over the last three years has developed a compact, off-the-shelf mesh network using this interesting dongle, I guess I'll call it. The way I got introduced to goTenna was actually in 2017, when I went to Burning Man.

And one of my friends gave me one and said, "Hey, on the playa, it's really hard to find people. But we can message each other using goTenna," because instead of doing what you'd normally do when you communicate where my phone sends radio waves to try to find connectivity and tries to find a satellite connection and then beams it back down, so basically it's going from my device up and then back down to try to find the other user's device, what goTenna does is says, "Forget all of that. Let's completely flatten the architecture." And instead, I'm identified by my IMEI, which is a phone identifier. And you can find other people's IMEIs and connect locally. So I'm just relying on other people's phones to ping my message to my friend, who is out there on the playa.

So a network of these devices, these goTennas, creates a mesh network. They have their own app. But what they did last year is they partnered with Samourai Wallet to build something called TxTenna, or Transaction Tena, where basically you can use Bitcoin to push, sorry, you could use a goTenna to push Bitcoin transactions to the network even when traditional internet infrastructure is down. A bunch of people did cool experiments with it. I ran a couple of meetups where actually the goTenna team came in, and we tried to troubleshoot some of it.

In fact, Richard Myers, one of the engineers at goTenna, recently branched off. He's doing a summer residency at Chaincode Labs, where he's working on Lightning. And he started a new initiative in partnership with goTenna called Lot49, which is focused on enabling Bitcoin-incentivized mesh. And so basically what goTenna's trying to do is they're saying, "Look, if we're going to use bandwidth on people's phones, if we're going to use people's battery, we want to try to figure out a way to use Lightning and incentivize people to be relayers in the network and to contribute their devices to the network."

So that's one really cool idea.

Jill Carlson
And there are more instances of this, right, just past playa. So in Venezuela, for instance, you all know this is an area I've looked at pretty deeply. There's a Bitcoin Venezuela mesh network. Now, I won't lie. It's not used to any great degree. But it does exist, which is very cool. And it's been built in response to political censorship, IP blocking, this kind of thing.

And so what's going on here is inexpensive Turpial and Harpia mesh network devices have been used to create inexpensive mesh networks within Venezuela that relay transactions to internet gateways to then be transmitted to the Bitcoin network. This is in case, for example, local Bitcoins goes down. And then these nodes can then, again, theoretically, connect to the Blockstream satellite through a receiver, again, another great use case of the Blockstream satellite. I'm proving myself wrong here.

And this is a very cool instance where you see something that starts to feel a bit more practical and a bit more, frankly, just needed.

Meltem Demirors
Right. And I think in both of these cases, what's interesting is there's actually an incentive. So this zero-start problem we talk about, I bought goTennas, I think a four-pack of goTennas, the devices is $399. That's a lot of money. Even if you're using inexpensive physical hardware devices like the coin Venezuela does, it still costs money to start a mesh.

And so what I think's interesting is potentially using Lightning Network to enable people to pay a couple of to relay a message, for example. And then all of the people who participate in the routing of that message gets some portion of it, or maybe it's something where on the basis of monthly or some sort of timespan participation depending on how often your node is up and relaying, you get some sort of split of the fees.

There are a lot of different ways to think about it. But again, what I think is so unique is historically in meshes, the monetization and incentive component's been challenging. If I spend $5,000 to set up my supernode antenna and then I spend $1,000 a month to run it, there's no way for me to monetize that. We can actually potentially use Bitcoin and the really, really small transactions facilitated by the Lightning Network to potentially think about interesting ways to resolve some of this zero-start kind of incentive problem and try to figure out how to get people to contribute resources.

So I think it's cool.

Jill Carlson
Yeah. Very cool.

Meltem Demirors
All right. Well, of course when there is Bitcoin, what do people do, Jill?

Jill Carlson
They make new coins.

Meltem Demirors
That's right. Everybody loves coins. It's raining coins. So of course people look at these mesh networks, and they say, "Hey, let's make our own coin." Althea is one such coin. They let routers pay each other for bandwidth, allow people to set up decentralized ISPs in their communities. And they're using Ether, not Bitcoin. The network topology's divided into gateways that connect to wholesale internet backhauls, kind of like those IXPs. They have relayers and then end users who pay in Eth for bandwidth.

Jill Carlson
Which is pretty cool. And they're not the only ones doing something like this. There's another project called Helium. Slow Ventures is actually an investor in them, full. And they create about $500, you can acquire a physical hotspot that uses something called LongFi, which is a wireless protocol that's optimized for miles' worth of range. And so what they claim is that they can blanket an entire city with just 5,200 hotspots, which you have to understand is not being done by any other mesh network hardware creator.

And they of course are also introducing their own native currency called Helium that routers are rewarded with and data credits can be bought with.

Meltem Demirors
I actually had Helium in my office for a meetup. I'm not an investor. I'm not incentivized in any way. I'm actually thinking of buying a few hotspots. What I think's cool about Helium is they're really thinking about the future of connectivity with all of these IoT devices. And in fact, one of the interesting use cases they initially talked about when I met them two years ago was all of these scooters. So scooters need to connect to the internet. You want to connect your scooter to the internet? Helium's trying to figure out with all of these increasingly interconnected devices, even some of these home devices we mentioned, how do you-

Jill Carlson
Internet of things.

Meltem Demirors
That's right…

Jill Carlson
Or internet of shit, as we like to call it.

Meltem Demirors
I love the Internet of Shit Twitter handle.

Jill Carlson
It's so good.

Meltem Demirors
I highly recommend everyone follow. It terrifies me for the future. But it also makes me feel good in a way because I know that no corporation has got it together enough to actually do anything meaningful with my data. So makes me feel good in its incompetence.

Last one I'll mention is there's a project called…I don't know a lot about it. Network infrastructure peer-to-peer network connectivity. They don't express the hardware explicitly, but they have a tunneling overlay, kind of like what Tor provides, that's on top of TCP/IP, and a marketplace for selling excess bandwidth. Think that's interesting. If you go on Twitter, there's a number of different projects that are also talked about, RightMesh. There's Skycoin. There's a ton of them. They're all doing different things. We just mentioned a few.

So I think that kind of covers a lot of this. So, Jill, let's start to wrap it up. We've talked about a lot. I want to talk about the future. I know you don't want to talk about this, but I'm going to talk about it because I like it.

Jill Carlson
Go for it…

Meltem Demirors
Okay. So IPFS, or Filecoin, IPFS stands for InterPlanetary File System. Jill's laughing at me, and she thinks this is funny. The idea ... Look, regardless, I'm not saying anything on Filecoin or the project today. But when I met the founder, Juan Benet, who's working on this as a Ph.D. research project many years ago, the idea really is what I want to talk about. The idea was, okay, if humans start to travel through space, how are we going to communicate and send data across large distances?

And so this is where the idea of IPFS came in. They're building a distributed file system with a network that's resilient and redundant, etc. The other really cool thing about IPFS, and this is very personal for me, is when Turkey started really cracking down on political censorship back in 2017, IPFS actually took a copy of Turkish Wikipedia at that moment in time, and they housed it permanently. So there are a bunch of different IPFS nodes store a different part of that. But basically it's preserved that site.

And then the second thing I think's really interesting that IPFS does, there's a lot of research that requires really large data sets to be stored. And so it's really interesting to see how building kind of a distributed way to share computational capacity to store these files, analyze them, etc., could potentially play in here and maybe part of meshing is not just relaying messages. But it's also storage. It starts to get very quickly really interesting. I'll stop there.

Jill Carlson
For what it's worth, I'm not laughing at the idea. I think the idea is actually very, very cool. I also, though, think it's quite out there. And I try and stay as grounded as I can in what's going to be useful today and in the more immediate future. But, hey, the way things are going, maybe we'll all be living on Mars sooner than I think.

Meltem Demirors
And, look, again, I think we have to go back to what's realistic and practical today. Do I think the IPFS concept is practical? Absolutely not. Do I think it's an interesting idea? Absolutely, yes.

The other thing I really want to touch on is DARPA research on private telco networks and military operations. So as I mentioned before, there are different rules that apply to the military. And actually what a lot of people don't know is in the 1980s, late 1980s, when ARPANET became the internet, there was an offshoot that remained called MANET, which was the military internet. And so apparatus in this country and every country tends to have its own telecommunications infrastructure. And they actually when infantry goes to new areas, they actually carry around comms packs that double as nodes in their own mobile network.

And so when the U.S. military shows up places, they actually have this really cool mesh architecture they set up to build their own telecommunications infrastructure that communicates back to central control and the executive branch using satellites and really fast network connections. What's interesting is also, according to FCC rule, consumer devices can only be really low bandwidth, but military and rescue operations devices can be really high bandwidth. So it's just interesting to think about the fact there is a separate communication networks privately owned. It's for the use exclusively of the military. But there are not a lot of the same limitations, which makes a lot of really cool, really interesting things possible.

And as you can imagine, this is how the internet all started was this desire for military network, right?

Jill Carlson
Totally. And so, okay, Meltem, I have a feeling that you could go on about this all day, which I love. But I think that for the sake of our listeners and for the sake of the rest of my evening, we should start to wrap this up. And so what I want to propose is that as I start to think about what the future of this looks like, I can imagine a lot of good that comes out of this, of us collapsing down the hierarchies of us basically disintermediating the way that we access information, the way that we access even payments networks and our finances as we start to think about ways that we could integrate Bitcoin with mesh networks.

It's very exciting. But I've also been thinking a lot lately about the potential problems that come with decentralization. And I'm not just talking about it's less efficient, it's harder to plan things. But I can also think of a lot of problems that could come out of this in terms of ... We love to talk about libertarian ideals of free speech, all of these types of things. But you take that to the extreme suddenly, and, I don't know, that to me can also be a pretty dystopian future. What do you think?

Meltem Demirors
Look, I agree with you, Jill. I think there's always a happy medium, right? Anything taking it to an extreme becomes its own form of perversion. And so I think the point with meshes is not that ISPs and centralized service providers are bad. I think for 99.999% of people in use cases, it's totally fine. And if you want to have privacy, I do think we'll continue to see privacy-oriented products and services like Helm and Casa and other types of home hardware, other types of software continue to get built for people who want it. And maybe it'll become more popular.

But what I do think is important just like the reason why I find Bitcoin so important is it offers choice. I think these new types of networks and these new forms of communication and these new networks are important because they offer choice. And as you and I have talked about often and as you often mention with Venezuela and I mention with Turkey and some other events that have happened around the world, sometimes you don't have time to plan until it's too late. And it could be because of a natural crisis. It could be because of a political crisis. It could be because of military operations or a military crisis.

And so I do think it's really cool to see people working on this technology. Now, there are a lot of forces that are going to work against them. There are so many companies today that are highly dependent on keeping the existing architecture and the existing power structure we have. And if meshing starts to work at scale, I wonder if that entire world will start crashing down.

Jill Carlson
Well, time will tell. We shall see. But for today, I hope that all of our listeners at least learned a thing or two about mesh networks. Also hope that we got things right here. And I also want to give one more big shout-out to Dan Onggunhao, who, again, we'll put his name in the shows notes and his Twitter handle, but thank you so much for all of your help in putting this together.

Meltem Demirors
All right, I think that's it. Hopefully we didn't piss off all the super technical people by mangling some of the bits and pieces. This was like the beginner's guide to meshes and networking infrastructure.

Jill Carlson
But by all means, leave us feedback on the Twitters, on Mediums, on the LinkedIns, and tell us what we got wrong here. And we'll make good on it next time.

Meltem Demirors
All right. We'll see you next week.