Skip to content
Innovation Capital

Episode 9 of PatSnap's Innovation Capital podcast

Helium: The People’s Network and the Decentralized Internet of Things with Mark Phillips

Listen Now

About Innovation Capital

Inspired by the words of U.S. inventor Charles Kettering, “if you have always done it that way, it’s probably wrong,” Innovation Capital, presented by PatSnap, was born out of a desire to go where no other innovation podcast has gone. Just as the world’s top innovators have pushed the boundaries of what’s familiar and accepted, host Ray Chohan takes a completely fresh and unfiltered look at some of the biggest topics shaping innovation today. From the key drivers of innovation, to its role in the economic value chain and groundbreaking outputs, Innovation Capital leaves no question unanswered. When it comes to innovation, we are your capital; your mecca for daring discussion and the fuel for your growth and scalability. Welcome to Innovation Capital.


Subscribe to Innovation Capital:

  • Innovation Capital on Apple Podcasts
  • Innovation Capital on Breaker Audio
  • Innovation Capital on Google Podcasts
  • Innovation Capital on Overcast
  • Innovation Capital on Pocket Casts
  • Innovation Capital on RadioPublic
  • Innovation Capital on Spotify

In This Episode of Frontier3

Ray is joined by Mark Phillips, the VP of Business Development at Helium. Together, they explore the massive growth of Helium’s network to become the largest operating IoT blockchain in the world. Mark discusses some of the most exciting uses of Helium technology and their future Web3 plans.

Episode Highlights

  • How Helium became the largest IoT blockchain operating in the world
  • The global benefits of taking a Web3 approach to IoT for sushi restaurant owners, rat catchers, and more!
  • How organizations like Helium are partnering with well-known Web2 companies to shape the future of networks
  • Want curated insights into innovation across deep tech, IP and more, straight to your inbox? Sign up to the Connected Innovation Intelligence Newsletter.

The Experts

  • Episode Guest:

    Mark Phillips

    Vice President of Business Development, Helium

    Mark Phillips

    Mark Phillips is Helium’s Vice President of Business Development and one of the company’s earliest employees.

    Powered by the Helium Blockchain, #ThePeoplesNetwork represents a paradigm shift for decentralized wireless infrastructure.

    Helium was founded in 2013 by Shawn Fanning of Napster, Amir Haleem, and Sean Carey. The company is backed by Khosla Ventures, GV (formerly Google Ventures), FirstMark Capital, Marc Benioff, HSB/MunichRe Ventures Slow Ventures, and others.

    Connect with Mark Phillips on LinkedIn

  • Host:

    Ray Chohan

    Co-Founder & VP New Ventures at PatSnap

    Ray Chohan

    Ray started PatSnap’s western operation from his apartment in 2012, helping grow the team to 400+ by 2020. PatSnap now serves 15,000 companies worldwide – supporting R&D, Innovation & IP teams with market and technology intelligence research. Their SaaS-based platform helps Deep Tech innovators connect the dots between technology, markets, and people. PatSnap officially became a Unicorn in 2021 and now has over 1,100 employees and 15,000 companies using its software across EMEA, North America & Asia. The ultimate mission is to provide intelligence that improves R&D and Innovation productivity to help innovators bring their ideas successfully to market. Ray also works closely with Blockchain & Web3 community and is a passionate angel investor in this space.

Episode Transcript

Ray Chohan: Mark welcome to Frontier3, really excited to have you on the show today. And it’ll be great to kick off with your background story because I can see you graduated from Syracuse University back in ’07. So it’ll be great to kick off with your career arc.

Mark Phillips Yeah. Happy to do that. Thanks for having me on the podcast here, Ray. My career arc is a little bit non-typical for what I’m doing, but I guess that’s not too surprising. So as you mentioned, I went to Syracuse University in New York state here in the United States studied political science with a minor in French and managed to apply pretty much none of that, quite frankly, after leaving university. And then moved to the Boston area. I’m from Massachusetts and moved to the Boston area, looking for something to do professionally and ended up working for a company called Basho Technologies, which at the time was making this sales optimization platform. And I joined them as a writer. So basically I was writing sales pitches for sales people, which is a hilariously meta thing to do.

And then worked in a few different roles there until the company actually shuttered due to lack of funding. But I had the ability to work amongst a group of remarkably brilliant engineers and learned a lot about how technology and engineering worked in that role. And then what happened with Basho was most of the company was unfortunately laid off because we couldn’t raise money and that’s a longer story, which we can get into over drinks. But the core engineering teams stuck around to basically take this database that the company had built something called Riak, R-I-A-K, and you can still find people using on the internet. And it was an open source distributed key value store. And so the company paired down to about seven engineers, the CEO, the CTO, the COO and myself. I was lucky enough to be kept around.

Accidentally managed to keep the job by calling up my boss saying, “Hey, I’d like to buy my computer because I don’t have one.” He said, “Well come in and work a few days and you can have it.” And then from there we went off and tried to build a database with the storage layer for the sales platform into something interesting. And so I grew into a sales engineering and community development role there at Basho. And then after about six years, one of my colleagues at the company, a guy named Sean Carey ended up co-founding Helium along with Shawn Fanning and Amir Haleem, our CEO. And he asked me to come over and join Helium to run business development about eight years ago. I think my eight year anniversary is next week, which is pretty wild. So that’s how I ended up here at Helium.

Ray: Congratulations. I really appreciated that on your LinkedIn profile you’ve got this really meaningful tenure at Basho, and congrats eight years at actually a business which is in Web3. That’s rare.

Mark Terrifying, but thank you very much.

Ray: You’re at the 0.01% for sure when it comes to tenure and time in market. So you walk through the door at Helium in February, 2014. What was the business like back then and how have you guys evolved to a Web3 business model? Because correct me if I’m wrong, the business didn’t really start off with a Web3 thesis.

Mark Yeah. I will not correct you because you’re not wrong. No. So when the company was founded, so Amir and Shawn Fanning had met through some mutual friends, primarily in the space of gaming. And so through those conversations, they started to talk about the idea of building something for connectivity. They actually had a few peers who were building companies that needed to connect devices to the internet. And a ton of the options, they’d be things like cellular and wifi and crazy things like Zigbee, didn’t really worked very well. So they said, well, we’re not doing anything. Let’s give this a shot. And so that was the genesis for starting Helium. Fun fact, the company was actually called Skynet when it was founded. If you look far enough back, you can find some press releases about this mysterious company called Skynet raising money.

And so Helium was very much a commercial organization that was building public wireless networks. And there was not any blockchain component to it at that point. And so when I joined in 2014, the mission was to build a public open wireless network, but it was through somewhat traditional mechanisms. Which was you’d go out and you would sign customers who had a use case for connecting you a foot traffic sensor to the network, you would sell them some hardware to deploy in their building or on their building or inside of their warehouse. And then they would procure the sensors and they would put these on the network. And then the cloud piece was very basic, data from sensor to the network, back to a cloud endpoint. At that point the cloud of course was much less mature than it is now so there was still a lot of people running this in private data centers. And then they would build applications on top of this stuff.

And so we built various versions of this product for about five years. And then leading into the most recent funding round. So the seed C raid… raised rather. So seed A, B, and then C, which was led by Multicoin and Union Square Ventures. That was back in May, 2019. Up until that point, we were still building traditional… it wasn’t traditional in the sense because it was public and open, but there was no Web3 blockchain component to it whatsoever. The shift for us happened out of a confluence of necessity and the available tools on the market. Right around when we started talking about this crazy idea of incentivizing people to build coverage, bitcoin and ethereum were becoming, I think they were having their second surge popularity.

We used to walk around the San Francisco office talking about how we might put the bitcoin blockchain onto hardware and have people operate this in their house. But ultimately we decided to build something brand new from scratch and have a go at this wild experiment of incentivizing people to create the coverage that was required for IOT or internet of things use cases. Because what ended up happening in most of these enterprise deployments previously, let’s say middle of 2018, was that people would want to use the infrastructure but no one wants to maintain a wireless network if that’s not their core business. So we kept hitting that wall. And so we basically went back to the drawing board and said, let’s just try something absolutely crazy here. This thing called blockchain is getting somewhat popular. Can we integrate this into the hardware that we’re selling and distributing and make the network coverage go a lot faster? And that has exploded in such a wild way that no one even imagined could be possible at dealing.

Ray: Just pausing there, I just want to unpack one key point for the audience because a lot of our audience are folks on LinkedIn. So folks who’ve spent most of their career in Web2 are probably exploring Web3, but let’s face it Mark, probably at scale don’t really get it. They kind of get it from a distance. So just put pin in. So initially the business is trying to scale a wireless network and the original thesis was we could now go to say, mom and pop shops, warehouse owners, any business location and in essence, ask them, would you like to be part of the network? So have some picks and shovels and hardware to be part of this network. But at that time there actually wasn’t no real incentive in terms of feedback loop to those different nodes on the network. Is that right? So you were going to them in a kind of evangelical fashion, but there was no real what’s in it for me doing this for Helium. Is that correct?

Mark Yeah, that’s mostly correct. Right. Basically at that point we were selling something specifically catered to the IOT use case. So one example is we used to actually have a product I think it was called Pulse, so Helium Pulse. And this was a pre-packaged cold chain monitoring product. And we would go to quick service restaurants, McDonald’s, Cava grill, that sort of company and say, listen, you’ve got a bunch of refrigerators, you probably don’t know what the temperature is. If you do know what the temperature is, someone’s doing it by hand on a clipboard once every 10 hours just to make sure that they’re checking boxes. But in some cases, sushi restaurants have a quarter million dollars worth of inventory that might just go bad in the middle of the night because somebody unplugs something.

So as an example, as a sushi restaurant, we know this from market research and talking to customers at that point, you might have a quarter million dollars worth of inventory at a sushi restaurant in a refrigerator. And a lot of the times these things just fail in the middle of the night to a compressor issue or someone unplugging something accidentally. And so we would sell a solution that would monitor that. And it was great. There was a dashboard that would show you real time data, you get alerts on your phone. But to your earlier question, there was no real incentive for the people deploying the network outside of the fact that they did get some value from it in that use case, which of course is extremely valuable and that’s ultimately what the network is about. But at that point we were selling specifically to a business need versus building this community of people who were incentivized to build a network and use it.

Ray: So, use case one was, for example, let’s look at that, refrigerator monitoring, IOT capability. So, that was probably more of a classic subscription model, like enterprise software and hardware, where if I am a sushi restaurant, like Yo Sushi for example, which is a brand here in the UK, I might actually would love to have the front end software because yeah, we do have inventory which parishes due to things which happened with refrigeration. So at that time prior to adopting the Web3 model, was it like a classic subscription based model then back then?

Mark It was, yeah. I’m forgetting the exact pricing, but it was some combination of the cost of hardware, plus a monthly subscription to the service that tailed down over time as you pay down the hardware or the CapEx, I guess. And the buyers for that product are facilities managers, heads of operations, if it’s a smaller outfits, the CEO or the CTO, but yeah, that’s exactly what the model is.

Ray: And what was the network model in that model then? Of the more you have on the network. What was the wider value to the community, the more customers you have? Was that part of the story and part of the long term value prop?

Mark We planned for that to be part of the story. The issue is that it just takes so much coverage for it to be useful in a public sense. Meaning if you look at the Helium network now where we’re well over 500,000 gateways, we have extremely meaningful coverage in many countries throughout the world. Deploying with the previous model, we may have gotten there, but there was really no incentive to make people go faster. So, I remember that we did use this in our pitches and as part of our strategy, but we just hoped to get there over time, by massively deploying this wireless network through our customers.

Ray: Okay.

Mark We certainly would not have gotten to where we are now in terms of the amount of coverage if we had stuck with that model.

Ray: Okay, great. So, I want to be really pointed here because your story is fascinating. What was the specific month, moment, business conversation where the business said, let’s go Web3, let’s go on the blockchain, this is the way to build that ultimate incentive structure, stroke feedback loop. What exactly happened?

Mark It’d be hard to pinpoint it to a date. A time range would be sometimes toward maybe the middle of 2018. From a company perspective we had tried, I mentioned that we built this vertical solution for temperature monitoring. The company’s specialty is building distributed systems. And we never really wanted to build a vertical solution for IOT. The plan was we would try a few verticals and prove that they were useful. If they got successful enough, we sell off the business unit to somebody who was interested in running it. But the company, Amir and the engineering team… and a lot of the engineering team came from Basho, the company that I came from, at least a lot of the early engineers. And so the desire was always to build resilient, scalable distributed systems.

And so I would say sometime around the middle of 2018, I believe what happened was one of the engineers, a guy who goes by the name of Bone said… Bones isn’t his actual name. That’s what we call him at the company. His name is Andrew. He said, “I’m going to write a white paper. Everyone who does these blockchain projects has a white paper. And I'm going to take a pass at one that would describe how we would build a decentralized open IOT network using a blockchain incentive model. And that’s really what started it. And it was we’d been sitting around having a bunch of drinks, kicking around ideas to basically save the company, because at that point although we had raised $37-$38 million US worth of funding, none of our models were working. And so we pretty much only had one more shot at it. Yeah. So I would say it was in the back half of 2018 where we really decided to make the switch. And it was just based on some crazy exploration. At some point, Amir said, “I mean, this might just work, so let’s go for it.”

Ray: Awesome. So someone internally picks up the bag and said, Hey, I’m going to spin up a white paper. The founder’s leadership, like yourself, are like, there’s actually legs to this. Let’s just do it. Was that literally the moment where let’s just execute and run an experiment?

Mark Yeah. That’s pretty much what it was. And our engineering team is so hungry and talented that it’s likely that as that white paper was being produced, somebody else was off writing code for a proof of concept, right? So we probably had something working very quickly. And we determined that the existing technology in the blockchain space, primarily Ethereum and bitcoin, those blockchains were not going to work for us due to the requirements of how the Helium blockchain operates. And maybe we’ll get into that. But it’s a ridiculous proposition to say, well we’ll save the business by building a blockchain and layering the wireless network onto it. Nobody wants to build a wireless protocol. Nobody wants to build a blockchain. I would still generally advise people not do that. I think if you talk to our engineering team, nobody would want to do it again. Yeah.

Ray: So, you make that change in 2018. Is that when Multicoin come in or did you guys start building momentum with number of customers / revenue, where they came and what was the story behind Multicoin jumping in?

Mark Yeah, so Multicoin along with Union Square Ventures ended up leading the C round of funding that we did in May, 2019. They were not involved to my recollection in the very early conversations around integrating and writing what became the Helium blockchain. But we did talk to them a while before they ended up leading the funding round. Those guys, Tochar and Kyle and the entire team, they were extremely instrumental from the beginning. Outside of the fact that they have deep knowledge in the technology and the economics of blockchain, they were instrumental, for example, in suggesting this data credit mechanism that we have, which basically enables us to build real utility around the token and also has some really nice properties when it comes to making the Helium network usable to people who don’t want to actually interact with the cryptocurrency at all. Yeah. So, they were a huge part of the success for sure.

Ray: Makes sense. So, unpacking now the core value prop. Now you’re fully all in on Web3. Like what is Helium today? Like what is the vision and mission, value prop of the organization?

Mark Yeah, that’s a great question. So today, in the easiest and most straightforward terms, the Helium network is the largest contiguous wireless network in the world. It’s also probably the largest blockchain in the world. We don’t talk about that too much. We probably will more in the future. But we’ve got over 5,000 full nodes running on the Helium blockchain. But again, more importantly, it's the largest wireless network in the world. It specifically is made, at this point, to send data from IOT devices, small devices. And we believe that we have a shot at being the largest wireless network, the largest decentralized wireless network in the world within the next five to 10 years. So, that’s where we are currently.

And so the value proposition is two-sided. For people who want to have a stake in building something brand new and transformational, it’s as easy as acquiring a $400 piece of hardware and spending 10 minutes to set it up and then becoming part of the decentralized network. For users of the network… and again, all this really drives towards is utility, because that’s all we care about in the end. We hear day in and day out from large companies and small companies, but mostly large companies, that have use cases for an IOT network, but were not able to do it because the network just didn’t exist. And so that’s really the important takeaway here is that we now have given people the ability from a utility perspective to send data from devices back to a cloud application that powers GPS location trackers, and mission critical supply chain and logistics pieces on this open decentralized wireless network.

Ray: It makes sense. So to simplify it further, is there a future state where it’s retail, local grocery shops who participate in the Helium network? So they buy a 400 pound device and they’re now a node on the network. Is that in essence how it would look like now and in the future?

Mark I mean, that’s a current state, right? So, what’s fascinating about how this gets deployed is that the typical journey for someone who wants to create coverage, there’s not so many early adopters now. Of course, because we’re a few years into this. But one learns about Helium on a message board, they’re on Reddit, they’re on Twitter. They see this thing about a blockchain for IOT. They do some research. They buy a piece of hardware and they deploy. It’s called a hotspot. So they deploy a hotspot in their home. And then they get very addicted to the mechanism of deploying it because the community, aside from being attracted by the incentive mechanism, is actually very committed to changing the way that wireless networks are deployed. So they tell their parents, they tell their sister, they tell their neighbor, they tell the person running the grocery store.

And what’s typically the limiting factor to how much someone can contribute to the community and create in terms of coverage is locations to deploy hardware, right? You probably have an apartment. You might have an office where they’ll let you deploy something like this, or maybe you own your own office and you put it there. And for most people that’s pretty much it. We’re now starting to see the emergence of very well funded and very sophisticated coverage deployers that might own towers, they might lease tower space from traditional companies like Crown Castle in the United States. They’re talking to billboard companies.

So people are doing fascinating things to actually deploy coverage. And then to your specific suggestion about grocery stores, what happens is after someone learns about how the economics work and the actual use of the network, they then actually go about forcing utility. What I mean by that is they say, well, great. I live in, I don’t know, Topeka, Kansas. I’ve got 20 hot spots running. I want to deploy another 50. What is it that I can do to convince people to put this in their location? And sometimes what they do is they share a percentage of the mining, right? So they’ll say I’ll give you 20% or 30% of HNT. But more and more what we see is people saying, Hey, listen, you’re a grocery store operator. You have some refrigerators, there’s a bunch of pre-built sensors that I can take and put into your refrigerators and I can deliver you this very nice dashboard and I’ll do it all for free, or I’ll do it all for $5 a month per sensor. And that gets the network deployed extremely quick.

And what’s amazing about that is that if you recall the story before about that vertical application, we didn’t want to build that. We just felt that we had to prove that the network had utility and now the community is actually forcing that utility for us. It’s quite fascinating.

Ray: Wow. This is fascinating. This model, this naturally incentivized model, you in essence build the world’s largest decentralized SalesForce in essence, because A, they’re part of the network themselves. Say someone in an apartment will have a node, but then to your point, Mark, at Thanksgiving, tell family and friends, and then continue evangelizing at their local grocery store. And the list goes on and on. So what does that look like now? And obviously then you mentioned the B2B side, which is interesting where you’re getting people who might have access to specific towers or other types of localized infrastructure where they’re connecting at mass.

So, simplifying it at the more retail level. What does that look like for you guys now? Is it exponentially in terms of the uptake where you’re getting folks approaching Helium and wanting to be part of the network? Well, what does that look like? Do I send an email to you where you send me a device? What does that actually? Because I know a lot of our friends say, “Okay, Ray, how does that work? Do I have an app on my phone where Helium are constantly sending devices to people I proposed a network to?” What does that look like in practicality terms?

Mark Yeah, it’s gone through a few phases. So when we first started, so the first hotspot came online, I think it was July 29th, 2019, and so the strategy was to start with one city in the United States. We started with Austin, Texas for a variety of reasons. We determined that there was a lot of forward thinking technologists just down there. There were some crypto fans. And so we pre-sold, I think it was 150 Helium hotspots into the Austin market and we launched the network. We kicked off the blockchain, the first HNT was mined at the genesis block, roughly July 29th, I would say August 1st, 2019. And at that point, Helium Inc. was manufacturing and selling and supporting Helium hotspots directly. So we built the first version of this for a variety of reasons. The wireless technology that we use is known as LoRawan, which is a specification that's owned by a company called Semtech, which builds the silicon that does the wireless protocol, but it’s generally an open specification.

So we used that for the wireless. And we had to build a blockchain minor. There was 30 companies that were building LoRawan gateways at the time, because as a protocol, it was gaining some momentum. It didn’t have a ton of usage, but it showed some promise. And so we decided to use that. But none of the existing hardware could handle the Helium blockchain requirements, even though they’re quite minimal compared to other blockchains. And so we built the first version of the hotspot. We built and manufactured, I think about 17,000 of those. And so those were the first gateways on the network. And then we along with the community at that time, and that would’ve been 2020, made a very specific decision to open up the manufacturing pipeline for other companies to manufacture Helium enabled gateways. And so along with that, we have pretty much moved out of the business of instigating coverage creation. I used to spend a lot of my time working with people who wanted to deploy 10, 50, 100 gateways and help them acquire those and give them advice on locations and talk them through the whole process.

We generally don’t even touch that anymore only because the coverage engine is really just working now. We add, I think 3,000 or 3,500 gateways a day, which is pretty staggering. 3,000 miners to the network every single day. We’ve got about 40 different companies that are approved by what's called the Decentralized Wireless Alliance, which is the foundation that governs the network. And so these are companies like Seed and Bobcat and TEKTELIC and 35 others. And there’s another 30 in the queue that are manufacturing gateways that should be approved in the next two or three months. And so to that specific question, we don’t spend too much time actually on coverage anymore. We’re spending a little bit more time thinking about how to instigate different wireless coverage creation. So cellular wifi, for example. But when it comes to the core LoRawan gateways, that’s entirely in the hands of the community now. And the ecosystem is remarkable and accelerating that still.

Ray: Wow. So, this is really key. This is what I want folks on LinkedIn to really understand. Like what you’ve described there is, I mean, it sounds like utopia, Mark. But I know the execution must be so hard. It’s blood, sweat, and tears. What you and the team have done is exceptional. I know there’s lots of nitty, gritty and execution and effort behind this. But is it fair to say now you’ve approached a part of that exponential curve where gateway providers, be it retail, be it smaller, local businesses doing things at mass, and now you’ve got the Alliance on the hardware side. The incentive loop is built now, like the flywheel is going. Where be it you’re a hardware provider as a part of the Alliance family, or a gateway provider, be it retail, be it more B2B at mass, everyone’s just all in, because the incentive model is a creative to everyone who helps the network scale and operate effectively. Is that the lay of the land now?

Mark Yeah, that’s entirely right. That’s entirely right. I mean, we have, as I mentioned, there’s a half a million LoRawan gateways that are online now. We’ve been told through our manufacturing partners that there’s upwards of 3 million additional chip sets that have been procured to build hardware just for the LoRawan line. So yeah, that is operating as we could have hoped. We’ve certainly reached escape velocity when it comes to with the incentive mechanism aligning with creating coverage. So yes, that’s working now.

Ray: And unpacking some of the economics. So we’re going to have a lot of folks here from BTB enterprise software, marketing, all types of tech companies, a lot of our listeners work in R&D, actually some of the folks in the Alliance might be some customers. So we’ve got quite a broad audience. What do the economics and incentive structure… first let’s start with kind of mom and pop or retail. Say it’s a bright 22 year old who’s just graduated, all in on Web3, is a Helium token holder anyway because he or she just loves your mission. Like how does the economics work for that individual who decides to be a gateway operator at home or in their apartment in downtown New York? And then kind of scale that network? What do the economics and the incentive look like in the most simplistic format?

Mark Yeah. So I’ll say that they’re generally uniform across all demographics. So there’s two incentives to deploying the network. And I’m going to talk specifically about the LoRawan and IOT incentive, because for future wireless protocols that get added to the blockchain, and we do have those starting to come online now, the economics, of course, we’ll still be based on HNT, but we’ll be slightly different in the specifics. But for LoRawan and IOT, we started with incentivizing just raw coverage. So the blockchain at this point, when it was launched it was slightly different, but now about 67% of all HNT goes to hotspot operators. And within that 67% of HNT, most of it goes to people actually just deploying raw coverage.

So to your example, this person buys a hotspot, they deploy it in their home or business. They put it online and to get it online you have to just give it power and give it back haul. So usually wifi or ethernet, but increasingly cellular. And then to just from that simple act, you’re providing coverage and the hotspot will continue to provide coverage, but in the process do something called proof of coverage, which is to emit radio packets and also capture other radio packets from hotspots that are geographically approximate. So that’s the primary incentive mechanism, just simply providing coverage.

Over time the proof of coverage algorithm is weighted to push more HNT to people who are actually providing coverage for real devices. So real devices are the examples we’ve just talked about with refrigeration monitoring. One of my favorite devices is a Helium enabled rat trap from a company called Victor Mouse Traps in the United States. If I had my camera up, I’d show it to you. I show it to everybody that I can. I’ve got about five running in my house because we live in an older house and sometimes we catch mice. But every time this trap has a mouse wander into it, it sends a packet back over the Helium network that ends up in a phone app that says, “Hey go check trap seven. There might be something there.” And so over time, the network incentivizes more about data routing from devices versus raw coverage. But you earn as an operator on the network, or a host as they’re called, HNT for both of those activities.

And so the two things that dictate it are the quality and breadth of coverage that you can provide. And what I mean by that is, if you look at a city like New York, there is a tremendous amount of Helium coverage. I think there’s something like 3,500 or 4,000 hotspots in the five boroughs of New York. We probably don’t need that much coverage. And so what happens there is that using a mechanism called transmit scale, basically you’re rewarded less than you could be because there’s too much Helium coverage. Conversely, if you are a scrappy individual that wants to deploy 10 hotspots in an area where there’s none, you’ll be providing net new coverage to the network and you’ll likely earn at a higher rate versus someone who puts one in Los Angeles or London or Paris due the fact that there’s less coverage in that new location, and it’s not as saturated. So the network weights it more heavily, if that makes sense.

Ray: Makes perfect sense. And what does it look like? I mean, say you get someone who’s bright, young and enthusiastic, they’re already started developing coverage for your new network, does Helium enable them directly as well? So send them enablement content, support, allowing them to speak to say local businesses or a local factory, which is around the corner from their apartment and they should be using the Helium network? Like, what does that process actually look like? Because they’re not full time employees of the business, right? So to enable them with the Helium narrative and just that starter pack, do you guys provide that as a business for all these evangelical people who are part of the network?

Mark Yes. So we actually provide very little and that’s by design, and I can describe why that is. So, it’s worth drawing a distinction between two entities. One is the Helium blockchain, right? The protocol, HNT, all the code that makes up this wireless network. That’s all open source, it’s all very much decentralized. It operates on its own. It’s very much a decentralized open blockchain based on a utility token. And pretty much everything that happens in that realm happens on its own, just pushed and managed by the community, under the leadership of the foundation, the Decentralized Wireless Alliance. All that generally just happens thanks to the ecosystem. Helium, Inc, which, which of course founded the blockchain network, we do you very little when it comes to actually enabling people and that’s because the community is so powerful in doing so.

And this was always the plan, right? We thought we would have to enable this for quite frankly, a much longer period of time, but the ecosystem and the companies that have come into the ecosystem and the community that operates are just so efficient and so rabid, quite frankly, about enabling the growth that we actually have to do very little for that. I mean, at this point, I’m not sure if you’re in our Discord server, we’ve got about 150,000 people in our discord server. And so these people show up every day to do things as simple as optimize antenna heights, right? So there’s a channel for figuring out which antenna to use. If you want to put something outdoors, you go into a different channel and ask questions about I’m putting this on the side of a mountain, what solar panel do I need? Et cetera, et cetera.

We also have a remarkable local grassroots community piece to it. So people form groups around specific states in the United States or specific countries, even down to the specific neighborhoods. There’s a fascinating set of tools from the community that lets you basically connect in the real world if you want to, with people who are deploying coverage in your same geography, so you can optimize deployments. So no, Helium Inc does very little to this. We’re not giving people materials, we’re not giving people collateral, they’re learning on their own, right? They’re going to Reddit and reading about it. They’re going to Discord. They’re reading blog posts. If you go to TikTok or YouTube, there’s an absurd amount of content on Helium and HNT, but also a ton about how to optimize antennas, which antennas to use, how to do a safe antenna deploy.

Because what’s happening is people who knew nothing about wireless and were just interested in the blockchain component and the incentive are now becoming very, very knowledgeable on all things wireless and signal optimization. It’s remarkable. So no, Helium Inc. does very little. The community manage pretty much the entire process.

Ray: Well, this is amazing. And how has the impact been internally back at base at Helium Inc.? I know you guys decided to make the move in 2018 with the funding from Multicoin and various other believers. But have you guys been surprised with the speed and the up take and just general adoption?

Mark Oh, yeah. We continue to be blown away daily by what the network does. I’m trying to remember the first hotspot sale, right? So we did a pre-sale for hotspots. We launched it when the company was at an offsite. I think it was somewhere early 2019. I think it was March, 2019 or April. And we were selling hotspots for delivery in July and August and so on. And a bunch of us had friendly wagers as to how many units we would sell on that first day. And I think I had the highest guess at something like 1,100 units. I think it was maybe 350 or 400 hot spots were sold on the first day. But we’re now deploying, as I said, about 3,000 or 3,500 of these hotspots every single day across the network. It’s still very hard to get your hands on a hotspot, even though there’s 40 approved manufacturers we’ve managed, or the community has managed to basically be bankrupt the entire LoRawan hardware supply chain.

And that’s getting better because large players like Semtech, for example, have built new FABs just to build chip sets for the Helium community. But now the company’s still quite small. We’re only, I think 45 or 50 people, and most of them are on the engineering side. The sales and marketing org where I fall, I think we’re about eight or 10 people. And we do have a fairly sophisticated usage adoption engine. And what I mean by that is we do have people that focus on getting usage on the network. And so if you look at our blog or you look at Twitter, you look at YouTube, we have tons of videos about people actually using the network and walking through specific use cases. But yeah, I think if you asked us where we would be when we launched this, nobody would tell you we would be this far in the mission.

Ray: And now looking at 100,000 foot context, right? A lot of your mission feeds off this exponential curve around IOT, right? And IOT has been talked about for a number of years. There was kind of a mini winter in the adoption curve. I think couple of years back, if you recall where people thought is this actually going to happen? So in a macro context, what does that look like now? I mean, for example, PatSnap, we have tons of customers who are hardware providers in that space, innovators in that space so we see some really interesting use cases now. But generally, how’s adoption curve been on IOT in manufacturing, in retail? Is there specific sectors which are moving quicker than others? What’s your macro view?

Mark Yeah. So it’s funny. Getting back to proving utility, we used to debate what verticals we should spend a lot of time on, right? I can remember numerous go to market exercises where we look across all the different verticals in IOT and try to figure out which verticals we should go after. What has ended up happening is due to, again, the power of the ecosystem and the LoRawan protocol, which had a very good base of technology before we got into it and now has just really exploded. All the verticals are happening in parallel.

So for example in LoRawan, if you wanted to build a supply chain and logistics tracking application for your fleet of 100 trucks, you can now do that with completely off the shelf technology. Meaning you can buy a sensor from one of five different companies. You can pair that with the Helium network coverage or deploy some of your own gateways, which tends to happen, if you’re looking at providing coverage in areas where there might not be coverage traditionally, or at least not yet. And then you can pick one of five different dashboarding companies, two of whom probably have a dashboard specifically built for supply chain and logistics, and you can get this application running in the course of a few weeks.

So we’re really seeing adoption across all of the verticals. What we’re seeing the most in, I think is probably things that look like both consumer and business location tracking. So for consumer think like, we just did a big deployment, I don’t think it’s public yet, but the company that makes basically LoJack for scooters, right? So I think Ray, maybe you’re familiar with LoJack, but if you’re not, they make location tracking hardware. Typically they use cellular and GPS for location. So with Helium, you can build that same sensor for much cheaper. And so there’s now that just did I think 1,000 sensor deploying in the Pacific Northwest with these very small hideable trackers that sit on the back of things like scooters that you might rent for a few hours. And those applications were impossible before because the cost to send that data on a cellular network was far too high, the hardware cost to build that sensor was far too high. And so that application is huge.

And then similarly, we’re seeing the same usage patterns but on the commercial side where someone wants to track fleet of 20 trucks that move between Boston and Washington, DC at all times, delivering linens that they might have just washed for a restaurant. Those are some really interesting ones.

Ray: And so obviously the transportation thematic, that’s looking exceptionally bullish at the moment. Is there any other verticals which have caught your imagination and are moving quick in terms of outside of logistics?

Mark So we’re seeing some interesting movement in… so consumer hardware, for example. So I’ll refer back to the rat trap example that I made before. So, Victor, for example, has always had a connected line of sensors. So they were a first mover in this. They started with wifi and people’s wifi networks weren’t working too well. And then they looked at cellular. And then they landed on LoRawan and the Helium network. And what’s really fascinating about that is it’s opening up new business models for them. So they traditionally would sell to a pest control company that would then go service a bunch of different restaurants. And then they would have to sell connectivity along with it. So they have to sell a gateway, for example.

In a lot of cases, now they don’t have to sell the gateway, right? So the cost of the product actually comes down significantly, which makes it a lot easier to sell. In pest control there is a very big resistance to paying more than $2.00 dollars for a snap trap, right? I think their Helium connected traps, one model sells for $50 US and one model sells for $100 and that’s a 25X increase over the standard wooden trap. But their cost of service is now coming down because they don’t have to deploy connectivity. When they do have to deploy connectivity, they actually don’t mind doing it because they’re deploying Helium coverage and they’re earning HNT for doing so.

And so, for someone who straddles both sides of the network, right? So they’re creating coverage, but also using it, it’s a remarkably compelling proposition because when you actually provide coverage, what you’re doing is you’re earning HNT, but you take that HNT and you use it to actually fund the cost of sending data over the network. And there’s without a doubt, always an overage there because it’s so cheap to send data over the Helium network. You can send 100,000 packets for effectively $1.00 US dollar. So you’re always mining more HNT that you need. And that actually, that accrues to the balance sheet for these organization. And it’s pretty wild.

Ray: So I’m just pausing there, because this is fall out of your chair crazy in terms of the business model. So, because I know you’re just so used to it, so you’re quite relaxed in the way you deliver that, because you’re living and breathing. You’re just like, yeah, this is how it works, Ray and cool. But to me, I’m literally falling out of my chair here. So even as a user of the network, say I’ve got a manufacturing, I’m a clothing manufacturer, for example and I’ve got a bunch of machines all using IOT capability, right?

Mark Mm-hmm (affirmative).

Ray: And I’m benefiting from the network in terms of monitoring and productivity, but also it doesn’t cost me anything because I pay in HNT but also I earn in HNT for using the network. So, and that balance always ends up being more. So in essence, for getting value for the network, I’m also getting a new revenue stream.

Mark Yeah. I mean, it’s likely. I mean, we generally, so much like the community operates coverage on their own and people figure out how to use the network on their own, we don’t really get involved in suggesting to people that there might be some revenue number attached to this. But inevitably what happens is IOT sensors, they don’t send very much data, right? Even if you have 1,000 sensors sending 10 every minute, you’re not really going to use that much data. And again, the way that the network is constructed because we have the community that essentially takes on the cost of operating and acquiring the hardware to build the network, we can offer to enterprises or the blockchain can offer to enterprises, a remarkable low cost of connectivity, this data credit. Specifically what it is one data credit sends 24 bytes of LoRawan payload. And you can do a lot of stuff in 24 bytes.

And so when you do the modeling with these organizations, and I do this a lot as part of my role, you say, “You’re going to have 100,000 sensors and they’re each going to send 10 packets a day. The cost for your full year is going to be this amount of dollars.” And their jaw drops saying like, “Well, that’s remarkably cheap.” And then you say, “Well also listen, you’re going to use public coverage for 65% of this, but you are going to have to deploy some gateways. If we look at our coverage map, the cost of gateways is X, but you are providing something that the network or that other customers are using. And because of that, the blockchain incentivized you to do that.”

And so, yeah, it’s entirely different. It’s entirely new. We do have to do a tremendous amount of education because still a lot of these large organizations are very new to blockchain. There’s a lot of reluctance. But yeah, it’s a remarkably compelling model when you get what you want, which is a usable network for your sensors. And then on top of that, you’re incentivized to keep that coverage running and you’re incentivized with HNT and the token. It’s quite remarkable.

Ray: Okay, great. So just so I understand this, so I could be a manufacturer. I deployed the Helium network because I’ve got a bunch of machines. Great. And I’m paying for that, but it’s exceptionally good value for the return. But then on top of that, your team might approach me as the owner of the factory and say, “Look, Ray, you should all also run a bunch of gateways.” Is that in other locations outside my factory and just be part of the network, because this is a way for you to generate HNT to maybe offset what you are directly paying for it in your own factory? Is that in essence, the positioning of it?

Mark It is. And actually what ends up happening quite frequently is the person whose responsible for deploying the project at a given company will come back to us very quickly and say, “Well, where can I get another 15 of these? Because we have all these factories that might not have a use case for sensors today, but if I’m going to create the coverage and earn HNT for it, I might as well just start now.” One really good example I can give without naming the customer is a Fortune 500 company in the United States that does a lot of agricultural business. And one of their lines of business is they have this distribution network of fertilizer tanks that sit at these large locations, and then farmers will come from miles around and fill up their smaller tanks and take it back to their farm. And the tank sits on the farm for a month or something. And they use that fertilizer and they go back and refill it, and so on and so forth. And so they had never done any actual IOT monitoring of these tanks.

And it came to us looking for help for building this. And so we walk them through the technical side of solution. And then we started talking about the blockchain and HNT and providing public coverage and went down the path of helping them think about how they would deploy coverage on all their locations. And so now, they’re very quickly moving down that path and they’re now looking across their entire US infrastructure to figure out where they can put additional hotspots because they’ve done the modeling on the utility of the coverage, right? They can start deploying these applications, not just the tank monitoring application, but pretty much anything else where a sensor can come in handy. But there is also this added public network utility component where they’re earning HNT for do that in the process. And we at pretty much in all the enterprise deployments that we’re doing. Once people wrap their head around the economics and how the network gets deployed, they generally go all in and want to deploy as much coverage as they can.

Ray: That makes sense. And stepping back, this model that you guys have rolled out, what are you competing against? What are you displacing, replacing, or removing out of the conversation? Like, what is your competition?

Mark So in a lot of cases, and maybe you’ve seen this with your clients, I mean, some of these applications just weren’t possible before that people wanted to build. In the last month alone, I’ve heard from four CEOs saying, “Hey we couldn’t build this application before the Helium network came along, because we knew we needed something like LoRa, and we didn’t want to build the coverage and no one was doing it. And so you guys have really saved us in that sense”. We even heard the same thing from nationwide connectivity providers that rely on cellular that have always wanted to offer an IOT line of business and just could never do it.

From a competitive standpoint, for most of these sensors, you could look at something like Bluetooth or wifi. Typically the range doesn’t work, or the power doesn’t work. The device isn’t static typically, right? You want it to be able to wander 100 yards or a mile or something, and still have coverage. There are from a pure connectivity standpoint, cellular networks and all the major MNOs and carriers in the US have an IOT offering. And there’s a bunch of companies that basically use their network to productize IOT offerings. We see those most in the competitive sense. And there are some use cases where the cellular networks are better for a specific device, but generally, if you can constrain the amount of data that you have to send, and you can live with coverage that isn’t complete but is generally where you need it to be and is growing remarkably fast, then the Helium network is going to be a fit for what you’re trying to deploy.

From a crypto economic standpoint and the utility, there’s really nothing out there, right? There’s a few other wireless blockchain projects that have a little bit of momentum. They’re mostly garbage between you and me, but there’s no pure competitor that marries the demand side which is the sensors and supply side, which is the coverage.

Ray: So in essence, you guys have this compelling first mover advantage, a phenomenal team and momentum. And in essence, you’ve got network effects already, right? You can see the numbers in your Discord community and all the people adding gateways on a daily basis. And so just for context, in terms of the classic competition, it might be an MNO who have an IOT offering, but it might be a very pointed offering focusing on a specific use case and has to move around certain size of data. That’s why it’s meaningful, but it’s very niche and focused on specific use cases.

Mark That’s right. So in the enterprise sense, the Helium network is what’s referred to as a LoRawan network service, or like a public LoRawan network. And so in that space globally, there’s probably 100 companies that offer some commercial LoRawan service. There’s a few Telcos in Europe, for example, that have LoRawan service. In the United States you’ve got companies like Actility, which is based in… I forget which country they’re based in, but those are the two largest LoRawan commercial networks across the United States and Europe. And they’re now partners of Helium. And what that means is they plug directly into this public coverage that our community is built.

And so sensors that are deployed to their core network can do what’s called roam onto the Helium network and use our public coverage. So they get access to the largest LoRawan network in the world virtually overnight using basically an API integration. And then our operators love it because companies like Senet bring customers like… I forget. Oh, Volvo for example uses Actility, not Senet. So companies like Actility bring customers like Volvo, who actually put devices on these Helium network, drive utility through our coverage and operators of the hotspots that route those packets for Volvo get compensated by the blockchain for providing that coverage. And what happens on the backend is that Actility, after plugging into the network, they pay the blockchain directly for those packets. And then that ends up getting pushed down directly to the hotspot that routed the packet. That’s been one of the biggest developments for us when it comes to usage. I firmly believe that roaming partners for LoRawan will drive the most usage through the network over the next five to 10 years.

Ray: You mentioned, is it LoRawan network? So if you could kind of simplify it.

Mark Sure, sure. Yeah. LoRawan, so L-O-R-A. I believe it’s short for long range. It’s the naming and the ecosystem is quite dismal. But yeah, L-O-R-A, LoRawan is the protocol.

Ray: Okay. And is that a specific protocol which is for a lower density of data packets, but for a specific range? So it’s kind of more of a lightweight entry point?

Mark Yeah. Compared to something like wifi, for example, it’s optimized for much longer range.

Ray: Yeah.

Mark So a gateway, a LoRawan gateway for example, if it’s deployed correctly, can provide upwards of hundreds of square miles worth of coverage. But the trade off you make there is that the transmitter, the sensor, can only send very small packets, right? So hundreds of bites at the most. So the data profile device is relatively constrained. That said, you can do a remarkable amount of things in just a few hundred packets. Yeah, so LoRawan is specifically crafted to enable coverage over huge swaths of land and huge areas, but for very, very small data.

Ray: How big is that TAM? Like forward looking, how big is that segment going to get, the LoRawan segment? In terms of like, how far are we in terms of number of use cases, number of diverse ways to use that type of data transfer? Out of a percentage out of a hundred, how early are we?

Mark Oh, I think we’re still sub 10%. I mean, it’s incredibly early. I’m trying to remember the exact usage that’s on the Helium network specifically from the last month. There’s hundreds of millions of packets right flying across the Helium network per month. And then we see probably 10X that number of packets from other networks that have devices that we just hear from a coverage perspective, but don’t actually route. But I still think we’re remarkably early. I mean, if you read the research reports from the last 10 years, they’re all projecting hundreds of billions of devices in the next 10 years. And we haven’t gotten there yet. And I think a lot of that is due to the fact that the low cost connectivity for these small computers that have to send data just hasn’t been there. But I think we’re remarkably early still.

Ray: Yeah. So looking out, moving forward, and we have so many customers in this space, the mind’s going to boggle, isn’t it, Mark, in terms of everything will have a sensor. I mean, the wall in your house will probably have a sensor for some specific use case probably.

Mark Oh, absolutely. Yeah. I mean we’re seeing some applications where people are putting sensors, there’s one company in that LoRawan and that does this already, they embed sensors in concrete, right? So when you pour concrete in a building, they put a sensor in there that can transmit and give you a state of the concrete over time. And as I understand it, when concrete hardens, it’s incredibly critical to the success of the building over time. And so, yeah, I mean, as the cost of microprocessors continues to go down and the ecosystem for off the shelf sensors increases, I don’t think I’ve seen a use case in the last year that didn’t have an off the shelf sensor that can satisfy. Because most of these companies shouldn’t be building their own sensors, but there’s very few situations where you need to actually do some innovation on the sensing side at this point. You can get them all off the shelf.

And so, yeah, it really does boggle the mind, right? You should be able to put a sensor in any location to do just about anything. I would argue that you don’t need them for everything, quite frankly/ some things are better left un-sensed, I guess. But it’s incomprehensible, or I guess it’s hard to enumerate all the use cases, given the potential.

Ray: Yeah. I mean, is there potential hurdles around more demand for larger data packets? Are you using the LoRawan network right now, which is great for distance, but has a relatively low trans capability? Is there any challenges around that where the market does evolve and they need a higher bandwidth of data transfer and so your technology will have to evolve? Do you think that’s around the corner? What’s your thoughts on that?

Mark Yeah. Once again, the community has already brought us to that point now. So if you look back and read the Helium white paper, which actually I’d recommend, it’s a pretty short read, a lot of this stuff isn’t present in the network as it’s described in terms of exact technology, but we always talk about a new blockchain for incentivizing the creation of wireless networks. And so the idea long term was always to build, basically give people the ability to build any type of wireless network using the Helium blockchain. We started with IOT because that’s what we knew, right? We had various lines of business that were IOT and sensors, but we ultimately, we didn’t care about that being the only wireless network incentivized by HNT in the blockchain.

And so about eight or nine months ago, a company proposed to the community that we add cellular connectivity, specifically, 4G and 5G in the United States to the Helium blockchain. We integrate capabilities that would enable handsets and anything with a cellular modem to use the Helium network. And the network has a governance mechanism called a Helium improvement proposal or a HIP for short. And so anybody can come in and propose changes to the economics or changes to how the technology works. And I think at this point, we’ve had I think 50 HIPs that have been proposed by the community. And so one of the HIPs was to produce a cellular gateway and integrate HNT earning economics for cellular gateways.

And so that work has been ongoing for about a year now, a little under a year. And we’re now starting to see the first 4G and five G capable hotspots deployed in the United States. There’s a Twitter bot that tracks them all. I think there’s something like 5,000 or 4,500 that have been deployed just in the last month and a half. And there’s one approved manufacturer now to make cellular enabled gateways with another four in the queue. And so, yeah, the community is moving very quickly to satisfy the use cases that require more data. So cellular in the United States and we’ll probably see cellular outside of the US. It’s a little bit tricky for a variety of reasons, which I’m happy to get into, but cellular is starting now and we’ll likely see wifi relatively soon.

Ray: So with the cellular use case, then you’re going after the elephants in the room, right? The classic incumbents. So eating into their market share, I mean, they’re going to be doing what they can right? To kind of respond. So-

Mark Yeah, I mean, yes and no. I mean, so to be completely transparent, it’s no surprise or people should not be surprised to learn that we’ve talked to most of these companies, right? We announced a pretty big partnership with Dish networks in the United States, about three or four months ago. Where Dish alongside Helium and a company called FreedomFi, which makes the first cellular enabled hardware in the United States, Dish has agreed to do what’s called roaming onto the Helium network, which basically says, okay, you guys are out there building all this coverage. It’s impressive and we find it useful. In the event that one of our subscribers phones uses the Helium coverage, we will purchase that handset data from the Helium blockchain. So there’s already some precedent there for MNOs in the US to find it useful.

We’ve talked to most of the MNOs in the US, a bunch of them in Europe, even companies that are ISPs, right? The Comcast and Cox and that caliber of company. They’re all generally very curious about Helium, right? Some of them are farther along in conversations than others and thinking through how they might actually use the network. I would say that on some timeline, I think most of them will be friendly to the network. Sure, some will be adversarial. But it’s just really hard to argue with the economics of how quickly the Helium community can deploy useful coverage, right? As an operator of a business who’s always looking to optimize and cut costs and look for more margin, you really just can’t argue against it, right? And so you’re better served to figure out how to make use of it and potentially integrate versus get aggressive and adversarial. So, yeah. It should not be too surprising to most people that we’ve talked to most of those companies.

Ray: Yeah. Because I’m guessing with the advent of self-driving progressing well, I know the data requirement for that is very different, right? It is an identity and rapid data requirement, especially if you’re asking a car to urgently stop in a dangerous situation. So it looks like things are all trending this way, right? You’ve got a lot of macro technological tailwinds and primitives, which should naturally in time in due course, you’d need expansion of the network, right? Like it seems very bullish on that front.

Mark That’s right. And one of the things that’s really driving… so 5G, which has a tremendous amount of promise, has to be deployed in a much different way than traditional cellular infrastructure. So it’s much more focused on deploying a lot of small radios or RANs they’re called, or N nodes essentially. Traditionally, Telcos have deployed on towers using huge base stations and blasting out as much power as they possibly could across their own spectrum that they’ve acquired, but 5G is much different, right? You need a lot more saturation.

And so what we’ve proven is possible is that people will deploy this infrastructure in their homes and in their businesses. I mean, if you’re AT&T and you call up a subscriber and say, Hey you want to put this piece of hardware in your house? It costs X amount of money or X amount of dollars, and we’re not going to give you anything for it. They’ll say, no, I’m not going to do that. Right? But with Helium where we’ve built in this incentive mechanism where you can buy a piece of hardware, plug it into your house, you create better cellular coverage for yourself, you get to use the coverage, your neighbors get to use the coverage. And then from a Telco perspective we’ve figured out how to get more coverage into locations where they couldn’t.

Ray: I mean, this is spectacular. I know this is at the core of Web3 values, right? It’s the first time in history where the customers get to own the network, but still the more and more you dig in, I mean, this is huge, Mark. Like what you guys are doing-

Mark We certainly think it is.

Ray: It’s unprecedented. The innovation, the business model innovation, the ramp. Congratulations, Mark. I mean, this is an amazing story.

Mark Yeah. Thank you very much for the kind words, Ray. That being said, we still think it’s remarkably early for this.

Ray: Yep.

Mark Right? We think in five to 10 year timelines, it certainly is not going to be a change overnight that upends how wireless infrastructure is deployed. But no, it is remarkable the force and impact you can build when you give people ownership, right? What’s really exciting, what we think will make the cellular infrastructure go a lot faster is that with LoRawan when you deploy it, you typically don’t have a use for it. Meaning you’re an operator of a hotspot, you’re providing coverage, you’re earning HNT, you can see packets flying through your gateway. You can’t actually see what they are, but the network tells you that you’ve been routing traffic and earning HNT for that. But from an end user utility perspective, this is changing, but traditionally, most hotspot operators don’t actually use the coverage.

With cellular and wifi, if we get there, you get that, right? You earn HNT, you earn the incentive mechanism, you build the public network, but your phone can use it, right? Your computer can use it. You can always FaceTime call with your mom over the Helium enabled infrastructure that you’re providing. And so that connection to the infrastructure is entirely unique and there’s really nothing else out there like it.

Ray: Yeah. And with the 5G network, technically it’s all systems go in terms of all that hardware risk and development, that’s not in the hands of Helium Inc., right? That’s down to the manufacturer is building out the hardware to enable that connectivity, right? And that’s progressing nicely. So you’ve got no issues on that front.

Mark That’s right. Yeah. I mean, that’s another example of how we leaned on existing ecosystems. And it really does lend to the decentralized nature of the network. There’s 50 different companies that build high end cellular hardware that can be deployed in certain different situations. There’s a little bit of work that has to be done to get them approved to operate on the Helium network. But at this point, most of that is in software. But no, we’re not building up new sets of hardware specifically for this connectivity link. It’s all relying on existing prior art.

Ray: Wow. I mean, so looking out, now getting to wacky land, Mark. More in the kind of blue sky we hear. And Michael Saylor, I love the way he describes digital energy, the oscillation of energy like a hotel where you can slice and dice it a million ways, send it across the world and get it back and it becomes an asset. So looking out long term, say, 2030, right? Or 2032. What are some of the wacky stuff? I know it’s so early. This is pure speculative, but just super innovative. A lot of our listeners are dreamers, innovators, they love looking into the future. That's at the core of what PatSnap do as a business. So what are some of the slightly wild value props or use cases which make you think, wow, that could actually happen. I’m quietly bullish on that.

Mark Yeah. So we firmly believe that there’s a point in, we’ll say the five to 15 year timeframe, where you can access one unified layer of connectivity all powered by the Helium network. That’s the ultimate vision, right? So your phone, your computer, any sensor, other things that require wireless, all this will be unified under one Helium umbrella. And access to that will be metered by the blockchain and enabled by applications, like so called layer two apps that allow access to that connectivity. That’s really where we think this is going. And on top of that, all sorts of wild things can be built.

Actually I saw a really fascinating pitch deck recently where someone was planning to build a network of transportation assets, targeted at consumers, call them scooters, call them bikes, call them small vehicles that the entire network was predicated on the idea that the purchase and maintenance of these assets would be funded by the community. And then people would be rewarded basically for the utility of those assets over time. It’s one of those ideas that seems ridiculous when you first read it, but then you say, I mean, this might just work, right? If you give people some very well built tools to acquire and deploy physical assets and then get compensated as those things are used, it might just work. It’s not too different from how the Helium network operates.

And so I think you’re going to see a lot of projects go in that direction. There’s an interesting one on the Helium network, well they’re integrating with Helium. They’re called Planet Watch, where basically you’re incentivized to put weather data on the blockchain. There’s a token that compensates you for providing weather data. There's another interesting project that does use Helium called DIMO, D-I-M-O. That was just announced a few weeks ago where they’re making what’s called an ODB2 sensor that plugs in to most cars made in the last 15 years, provides data back to the DIMO blockchain then rewards users for actually putting that data onto the blockchain. There’s so much absurd potential. It’s hard to argue against projects where users get actual value out of something and then get rewarded for it. I think that’s exactly where things are going.

Ray: Yeah. I think at a first principles level, this feedback loop where the community and the customers have participation and equity in the network, like it’s the ultimate model of good behavior economic-… well, behavioral economics, which are driving good incentives, right? Like it’s one of the core of things which are beautiful to Web3. But Mark, I mean, I’ve really enjoyed the exchange today. I mean, we could probably go on four hours on some of the use cases where the business is heading. But this has been an enthralling conversation, Mark. Just want to say congratulations to you and the team. I mean, A, you’ve been there for eight years. That’s what I loved about your LinkedIn profile. Like wow, Mark’s a builder at Basho for X number of years, now at Helium. So I want to say kudos to you and the team for doing what you’ve done. It’s super inspiring. And also if people want to learn more where can they learn more about Helium and also get in touch with you?

Mark Yeah. So Helium.com, best place for generally learning about Helium. I think if you would only go one place, I’d recommend to go to the Helium Discord. So if you just Google for Helium Discord, it should come up. It just discord.gg/Helium. And then there’s a whole Reddit hole to go down. You’ll end up on YouTube, you’ll end up on Reddit, you’ll join the DUI community calls and have opinions on where the network is going. But the Discord is really the best place to get started. I will caution that it’s a little bit noisy. And so you have to do some work to tune what channels you care about, but definitely recommend Discord. For me, you can email me, [email protected]. I might respond within a couple days, if not, it’ll be a week or so. So, I apologize. And then I'm on Twitter as Pharkmillups, P-H-A-R-K-M-I-L-L-U-P-S. I’m generally on all channels Pharkmillups but that’s me on Twitter.

Ray: Awesome, Mark. Well, it’ll be great connecting with then. And hopefully we can do part two maybe in Q4 and see how far you guys have come. But looking forward to maybe connecting later on the year.

Mark Yeah, absolutely, Ray. Love to do that.

Ray: Mark Thank you.