Investing in DePIN: From Protocols to Power Plants

Transcript

Super excited to be here with you today. My name is Guy Wlette. I'm one of the investors for the crypto team and I'm going to tell you a little bit about from protocols to power plants investing in Deepen. First, I think it's helpful to ground the discussion in what the crypto team focuses on, our investing thesis and how we spend our time. We invest in blockchain projects that build decentralized networks. And we've always believed that decentralization is one of the key components that enables our projects to succeed. But when I say decentralization, that's a bit of an abstract concept. So to make it more tangible for you, decentralized networks are networks that are owned and operated by their users. And decentralized networks matter because they help to align incentives between users of a

network and the owners of a network. Many of the most successful businesses on the internet have been network effects based businesses. We heard Chris and David talk a little bit about this yesterday. And in the beginning of a network, the network itself is small and often not particularly valuable. So the owners of a network provide a subsidy to help make that network more valuable. Sometimes this is called doing things that don't scale like Airbnb taking pictures of the initial apartments onboarded to the platform or in the instance of something like Uber providing a subsidy to drivers paying them more than users compensate them for or providing a subsidy to uh riders and giving them cheaper rides. This is also very true for how a network relates with its compliments other businesses or

developers that interact with it. And one of the challenges is that over time as the network grows, it begins to compete more with the developers that use it and the businesses that interact with it. We've seen some examples of this and my favorite is when Zingo was essentially kicked off of Facebook marketplace. Zinga built games and those games were so popular that it became a strategic issue for Facebook. So Facebook decided to deplatform Zinga. We believe decentralized networks that are owned and operated by their users can help to solve this misalignment of incentives.

Throughout crypto, we've seen a couple of different eras where decentralization can be applied differently. The best opportunities have changed and grown over time. And crypto began as the uh fund in the industry is titled with decentralization applied to financial systems. The original project being Bitcoin or trying to apply the idea of decentralized networks to money itself. Then we saw decentralization applied to big tech to social media and to marketplaces and that was often called web 3 juxtaposed against web 2 or big tech today. And more recently we've begun to see a lot of founders and developers who are focused on building physical infrastructure that is itself decentralized. Those projects are deepen projects or decentralized physical

infrastructure networks. Crypto began trying to challenge financial services and many of the people that became interested in this time period were interested from an ideological perspective in the aftermath of the great financial crisis or they were interested from the uh technical innovations of Bitcoin focused around uh a decentralized network that was fault tolerant globally. We began to see more people entering the space in 2020 who are interested in the concept of applying decentralized networks to existing big tech companies or to marketplaces particularly around things like social media and the creator economy itself. And then we've seen a third additional wave of founders come to us more recently focused on

decentralized physical infrastructure networks or deepen projects literally building in the physical world brick by brick. Deepin protocols replace physical monopolies using decentralized networks. And you've heard me say decentralization or physical a lot now. So I'm going to hope to make this more tangible for you and tell you what problems deepin networks are tackling today. First, we've seen a lot of deepin projects focused on the telecom space, building new 5G cellular networks or delivering home internet to users faster and cheaper. Second, we've seen a number of founders focused on modernizing the American energy grid, building virtual power plants or micro grids. In

transportation, we've seen founders focused on replacing ride sharing apps like Uber or food delivery apps like uh Door Dash. We've also seen a lot of deepen protocols focused on crowdsourcing data about the physical world. And my favorite in in this instance is a project that encourages you to strap a camera to the dash of your car and film the world around you. Uh helping to collect and aggregate data that can be used to train autonomous driving models or make existing autonomous cars more aware of the recent state of the world around them. I've seen many of you uh this week who also have a Whoop or an Aura Ring uh fitness device helping to track your physical state every day. And that data is very valuable but lives siloed for you the

individual customer or simply with the company that builds your device. We've seen a lot of deepin projects helping to aggregate that crowdsource patient data and making it available permissionlessly for others to build with. In the robotic space, we're talking a little bit about embodied AI earlier and the idea that you can build robots or models that navigate the physical world. Unfortunately, the data to train these models is much less available than the large corpus of text or video on the internet. And so we've seen a number of deepin projects create an incentive for people in the real world to wear an XR headset and perform daily tasks that would one day be automated by a robot or to help crowdsource data of people operating robots directly for the purpose of training. And then finally, you're probably wondering why AI, an

inherently digital service, is on a slide about physical infrastructure. But as we've talked through, one of the challenges for many AI companies is getting access to compute. that's simply solved if you are able to raise a lot of money or have a direct relationship with one of the hyperscalers. But we believe that more people should be able to participate in these markets and economies. And so there are a number of projects that are focused on building GPU aggregators or the idea that if you have latent compute sitting in your data center or even perhaps on a gaming GPU that you use uh on weekends, can you make that compute available to anyone permissionlessly? And so there are a number of AI projects focused on building aggregated GPU marketplaces.

We'll talk through telecom and energy a bit more as specific examples, but first I think it's helpful to walk through explicitly why founders are choosing to build deepen protocols and the superpowers that you get when you decide to build a decentralized physical infrastructure network. The first is that you're able to connect latent capacity. You can aggregate unused resources and make them available. Uh decentralized networks make coordination much simpler than centralized networks. And so as we're talking about in the instance of a GPU marketplace, we can aggregate unused resources and make them available permissionlessly to anyone.

The second is that deepen protocols help to significantly reduce the amount of capital you need to spend to build a new physical infrastructure network. We'll see an example of this in just a second. And then finally, many of these different protocols are like all decentralized networks focused on eliminating platform risk and aligning incentives between the users, owners, and operators of a network. I got interested in the space because a couple of years ago, I was building a decentralized telecom project as part of our first crypto startup school program. This is a picture of of me at the first day where Chris was giving the opening talk. Uh unfortunately my project did not succeed but I learned a lot about the telecom industry that I'd like to share with you here today.

Telecom is named for telefan or cell phone service or not cell phone service uh wired phone service. The the original version of the telecom network in America and globally was operated by switchboard operators who built dedicated circuit switch links by plugging in and unplugging physical cables. Today that infrastructure is largely run on the internet and you can see on the right a data center and an internet exchange point helping to connect people across the country. Even in the last couple of years, telecom architecture has changed drastically. On the left you can see a very large 3G cell tower which in many cases was up to 50 meters tall, cost several million dollars to deploy and required that

centralized telecom companies would rent land to build new capacity. On the right, you can see a 5G hotspot which costs somewhere between a few hundred and a few thousand dollar providing coverage in a specific area. It's deployed on someone's balcony and helping provide support to the people around it. Because the devices became smaller and more modular, as well as a regulatory change to make unlicensed spectrum available publicly, it's possible now for the first time to build a userowned and operated telecom network. And that's precisely what Helium did. Helium was the first scale decentralized telecom network and really their superpower was eliminating the capital expenditure required to deploy a 5G network across the country. Existing

telecom companies like AT&T and Verizon spent an estimated 250 billion to deploy their network. But instead, Helium sold devices directly to owner operators who provided coverage to the people that lived around them. We're here this week in Las Vegas and you can see a map of live Helium nodes that are perhaps providing coverage to you right now. You might ask yourself, why is someone interested in owning and operating a telecom network? Well, like with many decentralized networks, the first users happen to be hackers and hobbyists. But there's a real financial incentive to participate. Very quickly, you can understand why people are excited to participate at the earliest stages of these networks. And recently, Helium partnered with AT&T to do mobile offload. Meaning that if you're an AT&T

customer, you might actually be using Helium's network under the hood for cell service every day. Next, I'd like to talk a bit about the energy sector and how we see this intersecting with decentralized networks. Energy in America is a complicated mish mash of standards that has evolved over the last hundred years. And today, the grid is comprised of three sectors: East, West, and Urkott, which is essentially Texas. Energy demand in the US grew pretty steadily for decades, roughly 7% year-over-year until the 70s when energy prices spiked. Oil prices specifically spiked, and it became much harder for the large centralized energy companies to predict how much power their customers would demand. Energy utilities

had traditionally built large centralized power plants, but there was a shift happening in the American energy grid. Energy consumption per capita plummeted as a result of this inability to forecast when to build new capacity and uncertainty about demand. Our energy grid has gone from on the left large centralized power plants to on the right distributed energy resources, things like solar panels or batteries that exist at the edge of the grid and are bought, purchased, and operated by users themselves as opposed to by energy companies. This shift has had pretty significant economic impacts, but it's also made it much harder for existing energy companies to manage and

even really understand their own energy grids because they're not operating all of the devices on their grid. It makes it harder for them to predict and deliver power to customers. This is also particularly salient for today's point in time when we're interested in building new power capacity for the purpose of training and inferencing AI models. Large centralized power plants can take 5 to seven years to connect to the broader energy grid because of bottlenecks in transmission. But solar panels or batteries exist at the edge of the grid and are able to sell power back to the grid itself.

We'd been interested in the space for a couple of years, but became more seriously focused on it when it was clear the energy industry itself was interested in decentralization. One of my partners on the American Dynamism team, Ryan McIntosh, wrote a blog post called Decentralizing the Energy Grid that did not talk about or focus on crypto at all. One of the issues he emphasized was how hard it was for centralized energy companies to manage information about their own grid because they didn't own all of the connected devices anymore. We met with dozens of decentralized energy companies and ultimately invested in a project called Daylight, which is helping to build a decentralized power grid in two stages.

First, daylight is selling a small device that will connect to your power meter and make information about your solar panel or battery about your home's energy consumption available to existing energy company in energy companies in seconds. Usually, a physical truck has to drive down your street to collect information from your energy meter with a delay of a week or sometimes a month, making that data practically useless for managing the real-time state of the grid. Over time, helium, not helium, daylight will move from selling information about your energy grid to selling control over your energy grid.

And so it will build a virtual power plant or an aggregation of solar panels and batteries that can sell energy back to the grid with bargaining power on pricing. Deepen protocols are massive opportunities with weak incumbents. Energy is a $6.5 trillion market a year in the US. Telecom, a trillion dollar market, and transportation, an $8.5 trillion dollar market. These are large incumbents that have often not changed for decades and have little financial incentive to modernize or to innovate. We also believe the deepin represents crypto's billion user moment or the opportunity for a billion people to interact directly with the decentralized network every day without even knowing it. Many of you, if you're using AT&T

self- service, are probably taking advantage of mobile offload with Helium. And yet, I imagine many of you were totally unaware. We've seen how permissionless access to a network can increase innovation and drive higher returns in other sectors, namely DeFi. The idea that combining many Lego bricks together will result in a better uh a better combination. We believe this idea will be applied not just to the digital world or to financial services but to the physical world and infrastructure that runs our daily lives. And finally, like many crypto projects, we get much faster times liquidity here than we would with a traditional company. Crypto projects often launch a decentralized network and a token within 3 to 5 years of inception as opposed to more

traditional projects. Further, the structure of these projects is that we often have the opportunity to double dip investing in both equity in a company commercializing the network and tokens for the network itself. This was the case in uh Helium where we invested in the team early on while they were building the Helium decentralized network which is incentivized by a token. That same team then continued to work on the project building Nova Labs, a traditional cell phone carrier that's competing directly against incumbents. We believe that deepen is eating the world and are very hopeful that even more of you will happen to interact with the decentralized physical infrastructure network in your daily lives. We're hopeful that deepen will

come to everyone across America and across the world very soon. Thank you.