a16z Podcast - State of Crypto 2024: Builder Energy, U.S. Election, Stablecoins, AI, More

Episode Date: October 29, 2024

We take you behind the scenes of our newly released, annual State of Crypto Report — a16z crypto's analysis of the latest data and trends that have defined the industry in 2024. This year's report features some brand new insights, from estimating the number of real crypto users globally, to understanding how much interest in crypto swing states may have ahead of the U.S. election. We also dig into infrastructure improvements to blockchains and key applications — including stablecoins, AI, and so-called DePIN. Be sure to visit a16zcrypto.com for all this and more including a new “Builder Energy” dashboard, which we’ll discuss on the show.Joining me to talk about the findings are lead data scientist and report author Daren Matsuoka and CTO Eddy Lazzarin. The first voice you'll hear after mine is Daren's, then Eddy's. a16z crypto resources:-State of Crypto Report 2024-Builder Energy dashboard-Estimating the number of real crypto users by Daren Matsuoka and Eddy Lazzarin Stay Updated: Let us know what you think: https://ratethispodcast.com/a16zFind a16z on Twitter: https://twitter.com/a16zFind a16z on LinkedIn: https://www.linkedin.com/company/a16zSubscribe on your favorite podcast app: https://a16z.simplecast.com/Follow our host: https://twitter.com/stephsmithioPlease note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. 

October 31st, 2008. That was the date that the Bitcoin white paper was released. And in the 16 years since, well, so much has changed. So where are we really today in the arc of crypto adoption? Not only at the infrastructure level, but also the applications built on top. In today's episode, you'll get to hear A16Z's crypto team dive into their new state of crypto report. So how many real crypto users are there?

Have gas prices come down enough for real applications to be built on top? And what is Crypto's relationship to the national conversation? Listen in to find out. Oh, and this episode was originally published by our sister podcast Web3 with A16Z. So if you are excited about the next generation of the internet, find Web3 with A16Z wherever you get your podcasts. And of course, you can find a link to the state of crypto report in our show notes. Now over to host Robert Hackett for more. Welcome to Web 3 with A16Z.

I'm Robert Hackett, and today we've got a special episode. We're taking you behind the scenes of our newly released annual state of crypto report. A16Z Crypto's analysis of the latest data and trends that have defined the industry in 2024. This year's report features some brand new insights, from estimating the number of real crypto users globally, to understanding how much interest in crypto swing states may have ahead of the U.S. election. We also dig into infrastructure improvements to blockchains and key applications, including stablecoins, AI, and so-called D-PIN. Be sure to visit A16ZCrypto.com for all this and more, including a new builder energy dashboard,

which we'll discuss on today's show. Joining me to talk about the findings are our lead data scientist and report author, Darren Matsuoka, and CTO Eddie Lazarin. The first voice you'll hear after mine is Darren's, then Eddie's. As always, none of the content should be taken as tax, business, legal, or investment advice. See A16Z.com slash disclosures for more important information, including a link to a list of our investments. This is our third year doing this. We've put together three of these state of crypto reports.

Why do we do it? I think we'd like to use this report as an opportunity each year to really take a step back, kind of look at the industry holistically, analyze all of the data that we have access to, we consult with various experts across the industry, and ultimately bring it all together into a single perspective. Yeah, I think that's right. And I don't want to get too lofty with this, but crypto as an industry was born in the social media age. and that means that information is constantly being streamed in real time. So I like to see the state of crypto report as a way to pause for a second and sort of appreciate what's happened in the stream.

You know, when you're in the mix of it and things are getting blasted at you multiple times a day, it can be hard to get a sense of the cumulative progress, let alone to remember what was getting streamed to you at full speed. speed a year or two years ago. So I think the state of crypto report is a chance for us to take a step out of the fire hose and to take a look at what's actually transpired. And to that point, our goal is not to just say that the crypto industry is great and get that message out regardless of what we're seeing.

We really try to be intellectually honest based on data and what we're seeing out there. Another thing I'll say is that crypto has a very overdeveloped. in-group language, not just because of the memes and the jokes and the personalities, but also because of the technical complexity. All those things together create this really strong in-group language. And I think of our report as a way to try to mainstream some of those big things. We try to avoid in-group language and make things very accessible through charts, through very simple explanations, and it's an opportunity to do that too.

Also, it's a way to shift the way people talk about crypto, because crypto is just filled with like really difficult concepts and terms and to be able to kind of peel back the layers and talk about what is actually happening in a way that people can understand is a unique challenge. Let's talk about how the report opens. So we're trying to summarize the state of crypto in the past year. So first up is size in the market. Because if we're talking about crypto, the first question you might have is, what are we talking about here? How big is the industry? How big is the market? And so that's the first question that we tackle here. Darren, you want to hit us off? Sure. For many years in the past, we've been mostly focused on

building the infrastructure to make all of this work. And so there really wasn't that big of a market for users of these crypto-based applications. But I think in 2024, we really did hit a tipping point with infrastructure improvements. These blockchains are becoming more and more accessible. They're cheaper to use. And as a result, we're starting to see applications come online. And now that we have these applications, I think it's really starting to come to life.

And I think in the years to come, this will be an increasingly important question. Yeah, I think that's right, because beforehand, the technology is still coming together. And in many cases, it is still coming together. But we're actually beginning to see some applications. And that's something that if you haven't been paying attention to crypto, you might not realize. Totally. And I think the results speak for themselves, right?

Active addresses across all of the blockchains that we track hit an all-time high this last month of 220 million mobile wallet users, so data that we get from mobile app stores that also recently hit an all-time high of 27 million. And so, you know, the numbers do kind of speak to this idea that we are now seeing some of these breakout applications that are bringing real on-chain transactions and usage. Let's go through each of those. You mentioned that active addresses are at an all-time high. Now, this is a kind of controversial metric. Not everybody loves talking about active addresses. One of the things they have going for them is that they are easy to measure.

It's something that you can very easily quantify. I think to really understand why active addresses has become such a hot topic recently. I think we need to look at some of the historical context. I think in traditional software, most people tend to understand the basic concept of a user. Of course, there's ways to measure the quality of the user. There's a whole field of growth analytics around this topic. But in its most simple form, I think people can grok the idea of a daily active user or a monthly active user. It's just, you know, the number of people that are using your product and service on a daily or monthly basis. It's fairly

straightforward. But in crypto, things are a lot trickier. And the reason is because on blockchain's user identities are pseudonymous. This means that it's very easy for one person to create and control a group of public addresses or identities. This is called a Sybil. And up until recently, a lot of the most popular blockchains had very limited capacity, which resulted in high transaction fees, which created this natural barrier for spinning up and using thousands or hundreds of different addresses, because it would come with significant monetary costs. But recently, we've seen that crypto infrastructure has become more scalable via L2s and new high throughput layer ones, which has reduced the cost of transactions on many of these

blockchains to nearly zero. And maybe you might ask, isn't it also near zero? costs to create multiple identities on traditional internet applications. And I'd say, yes, that's true for the most part. You know, I can create multiple email addresses pretty easily. But the key difference here is that in the crypto world, we have very strong incentives for this type of behavior. Yeah, Google doesn't give you money for opening a new Gmail account, unfortunately. Exactly. And a lot of new tokens today actually bootstrap their circulating supply by doing something called anirdrop where you're rewarding retroactively users of the protocol based on a predefined set of addresses. And so as long as people believe that these tokens could be valuable,

there is a strong incentive for them to try to game the system by creating and transacting under many different identities. This is commonly known as air drop farming. Yeah, there's no reason that one person couldn't spin up hundreds of millions of addresses and just go wild and engage in totally useless and inauthentic activity in order to try to maximize their likelihood of extracting as many rewards as possible. And what we've seen over the last four or five years that we've had airdrops and yield farms and so on,

there's been a profound professionalization of the efforts to pull out those rewards. People have gotten really, really, really good. And I want to make it really clear to anybody who thinks that it is a gigabrain take to say monthly active addresses is simbable. Yes, we know that. And if you have a better way to measure using public data, you know, that's verifiable by everyone or that is easy to trust, I'd love to hear it.

Tell me a better way. I'm thirsty for it. Tell me, let's work together and let's figure it out because the space deserves really high quality transparent metrics, and that's what we're trying to do. Yeah. And so given what we know about these behaviors, how many users are there, right? Is it 10 million? Is it 50 million? Is it 100 million? This is the question that we put a lot of work in and we're going to publish some of our thoughts around estimating them. And we'll include that blog post in the show notes, but maybe you could walk us through how you think about actually getting

from this big number, this 220 million number. If that's the number of addresses, how do we get down to humans? How do we reduce that to the number of users, the actual flesh and bones on the other end of a screen? I think one approach sounds kind of obvious, which is, you know, we've got these 220 million monthly active addresses. What if we could filter out what we suspect are bots and sibbles? And the way that we can do this is some techniques using on-chain analytics and forensics. One method is to filter out addresses that received funds from something called a dispersion contract. This is a smart contract whose sole purpose is to take in funds and automatically distribute them across many different addresses. Of course, there could be

false positives here, but the activity implies that the destination addresses all received funds from a single source, and therefore they're connected in some way. So there's certain types of filtering you can do around dispersion contracts specifically that can eliminate a lot of the bots that you might see. The other approach involves looking at addresses that have a near zero balance at both the beginning and end of the period that you're aggregating over. For example, if you're trying to find the real number of monthly active users in September, you can look at addresses that had a zero balance on both September 1st and September 30th. This criteria implies that the addresses were transient in nature.

I think bots and sibyls typically want to clean up their balances after taking actions on chain, whereas real human users usually like to keep some balance in their wallets to cover future transaction fees. So this is like an address popped up and then within like some period of time, people dumped into another thing and the address never was revived after that. Exactly. This is the type of behavior where it received a source of funds, it took some action, and then it pushed all of those funds outside. So it's a very transient type of activity.

That's often associated with bots and civils. The other thing that we've looked into is just like the frequency of transactions. And so you can analyze these distributions to look at, did the address transact one time, two times, three times, four times, five or more times during the period. Addresses that have just one or two transactions, you know, at best, they're a low quality user. at worst, they're a bot or a sybil, and especially over longer periods of time, this type of analysis can help you in some of that filtering. You can also specifically look at addresses that transacted very, very frequently over very short periods of time. Humans can only reasonably through a wallet or app interface process a certain number of transactions, whereas bots can do

things much faster. And so you can do some filtering there. And then on the flip side, you can optimistically include addresses that are tied to some sort of identity protocol that requires some sort of set up cost. For instance, addresses that are tied to an ENS name or a Farcaster ID or some other linked social identity off-chain. Those are likely to be real human users. So obviously, it's very complicated, requires a lot of on-chain data and analytics and forensics. But we've done an internal analysis taking this approach. And I think we've come up with some decent estimates. Okay, so the first method is basically you're starting with the number of active addresses

and you're trying to whittle it down based on culling the wheat from the chaff, culling out the addresses that seem transient, bot-like, or low quality. Yeah, so this is reducing the ceiling. We know what the ceiling is. If there's 220 million monthly active addresses, then we're trying to lower the ceiling. And then we can use other methods to try to get at a reasonable estimate of the floor. Okay, so let's walk through method number two then. So method number two says, rather than looking at active addresses,

let's look at some other sources of data that we might have access to, particularly off-chain data. And I think an obvious place to start here is wallet users. For example, in February, Metamask, which is a popular Ethereum wallet, reported that they had 30 million monthly active users. And specifically, they define an active user as someone who either loads a page within the Metamask extension or opens the mobile app at least once during any 30-day period. And assuming that we're looking to estimate transacting users,

we need to make some sort of assumption on what percentage of Metamask users actually end up transacting. And that requires some research. I don't think we're going to find the perfect number. We do have some sources that we can cite that say that 30% of users of MetaMask, at least on a daily basis, actually end up transacting. And so let's say 30 to 50 million, you can use some sort of assumption there. And based on that, you have 10 to 15 million users that we believe are transacting through the MetaMask wallet products. And then all we need to do is try to understand what Metamask's market share in the wallet space is. Of course, the exact data here is not going to be readily available, but we can make some

educated guesses based on what we know. For example, we do get pretty good estimates of MetaMasks market share on the mobile wallet side based on some data that we get from Censor Tower. And so once we are able to estimate what Metamask market share is, we can simply extrapolate an estimate for the total number of crypto users from that 10 to 15 million that we discussed earlier, and then compare the results to approach number one, the filtering from active addresses to sort of triangulate on at least a range that we feel, in the right ballpark. And by the way, it doesn't just have to be metamask. We can also run this

exact analysis for other wallet providers who have either reported their numbers publicly or are willing to share their proprietary data. And then, as I mentioned, just kind of triangulate on a range based on all of the information that we have. So we have done this. We have done this internal analysis. It is definitely not perfect. I'm not claiming at all that we are super, super scientific about this, but we feel comfortable saying that we think there are 30 to 60 million real monthly transacting crypto users today. Of course, this is still a small percentage of the monthly active addresses. It's 14 to 27 percent. And an even smaller percentage, actually, of the total number of people who own crypto but do not necessarily transact with

crypto. It's actually only 5 to 10 percent of the 617 million reported by cryptocurrency. crypto.com, which means there's a huge opportunity here, right, to convert existing crypto owners into active crypto users by bringing them on chain. So the point is to say is not that, well, the Sybil analysis is perfect and we can end up at a, you know, flawless classification of which addresses are real humans and not. The point here is to get sort of an order of magnitude range so that we can look at growth. That's the goal. The goal is to be able to look back retrospectively and say we're bigger than we were last year, we're bigger than we were the year before that, and see approximately how things are developing. I also have this kind of like funny way

of thinking about this too, where I'm like, well, maybe we're entering this new internet world where actually AI and bots and these things that are farming right now for rewards and trying to snatch up incentives that certain projects are offering. Maybe we're entering an internet where like bots and AI are just infused everywhere? Like, why are humans the major metric? Yeah, look, I don't want to get to sci-fi, but everything is many to many now. Everything in a networked world is many to many. By that I mean, like, does one person have a credit card? No, you have many. Is a credit card only used by one person? No, like there are many. Internet accounts, cars, just everything is totally intertwined and entangled. And as we add,

AI agents and other such things, it's going to get even crazier. In a way, what we really want is we want to measure the attentional capacity that things command, right? The total amount of like cognitive bandwidth that they are commanding from the environment. But that's a really like sci-fi way out there getting kind of kooky conversation to have, Robert. Next to your state of crypto report? Can we measure cognitive bandwidth? All right. So if we're sizing up the market. When we're looking at actual users and trying to think about the people who are using, interacting with crypto on a monthly basis, actual humans, we're looking at around 30 to 60 million, which is, as you said, 5 to 10 percent of the people who own crypto globally

and also a fraction of the active addresses too. Why such a wide range? Why 30 to 60? Like, can you not get any more precise about it? I think if anybody has a problem after they hear the methodology. They might complain that it's not wide enough, right? And maybe it should be wider. Maybe it's 20 to 70 or 20 to 80 or something. It's hard to say exactly, very open to that. We're not pretending that this is a totally flawless analysis. The reason for the range is just contained in Darren's reasoning, right? Like, we had to make some estimates. There's some things you have to kind of guess. Like the ratio of monthly active users is reported by MetaMask to the proportion of them that actually do transact per month, thereby become.

in active address, right? Like, those are just things that we have to estimate. And I think it helps even just to think in orders of magnitude, like what we're looking at here, the fact that active monthly users of crypto is about tens of millions, and global owners of crypto is hundreds of millions. There's a gap. There's a gulf there. But we're beginning to see the sort of outline of the market that we're trying to size up

here. That's right. one thing that will be apparent to anybody who looks at this report is that it's not just that active addresses are at an all-time high, it's that there are a few blockchains that are driving this. One of them is Solana. Salana is far and away got the most active addresses of any blockchain network that we looked at. It's got 100 million or so. What's going on there? Well, I think the obvious answer here is that the blockchains that have the most active users are the ones that it's cheapest to transact on, right? And so, you know,

blockchains like Solana are very, very cheap, which means it's very low cost to set up bots and Sibbles. We also got some interesting data from the Artemis team, and we have found that on blockchains like Solana, a large portion of the activity is simply one-time use or two-time use, which indicates that there is a larger percentage of bots and sibbles on these low-cost chains. And so I would take all of the face value monthly active addresses by chain data with a big grain of salt because there, of course, is lots of nuance associated with this. But with all that being said, base and Solana were some of the top destinations for crypto users in 2024. I don't want to take away from the great growth and development that those ecosystems,

have seen this year. Yeah, base has the most active addresses of any Ethereum chain that we looked at at 22 million exceeding even Ethereum, which has 6 million. So clearly, they've struck a chord. Totally. I was also interested to see Bitcoin up there. Like, Bitcoin's got 11 million by our account. Totally. And to keep going on this point, like the cost to transact on Bitcoin is not near zero, right? Which means there is likely to be less bought. and Sibbles on the Bitcoin blockchain. And so, you know, you can kind of take that number more at face value. Certainly, Bitcoin has been an ecosystem that we've seen.

More builders actually try to build new products and applications on the Bitcoin blockchain. And I think it's gotten very mainstream attention with a lot of top politicians talking about it. So I think definitely a good year for Bitcoin. Definitely. I mean like the strategic Bitcoin Reserve once a niche idea is becoming more mainstream. Yeah, and to add to Darren's point, all in all, I think it's a lot more difficult to look at Bitcoin activity and say that it's largely driven by Sybil. I think that's just a lot more difficult to believe. So if activity is positive on Bitcoin, the implication is that there's probably an increase in real activity across the rest of the space.

I mean, take with a grant of salt, but I think that that's likely to be true. So we're talking about some of these blockchain networks that have had a lot of activity and interest over the past year. We mentioned Solano, base, even Bitcoin. One of the most exciting things I think about this year's State of Crypto Report is this segment that we've dedicated to builder energy. This idea of looking at everything that founders are telling us about where they're building what they're interested in and where their activities and energies are directed toward. I think that that is going to be a mainstay feature of future reports. The big context is that I think the heartbeat of the crypto industry is, of course,

all of the builders who are building the products and infrastructure to make this all work. And we meet with and engage with thousands of companies every year as part of our course of business through investment team research, our crypto. dope startup accelerator. We do some external deal tracking. And for the first time, we now have a lot of tools internally that allow us to normalize a lot of that data, aggregate it in a way that we can share some of that data externally so that people could see the bigger trends here across blockchains, geographies, technologies, and categories. And we've done that. We've made actually a dashboard that we call the Builder Energy dashboard. It has some very cool

interactive data visualizations around all this data. And you can check it out at builderenergy.a6crypto.com. One thing I'll point out is we were talking about blockchains that have the most active addresses. And it's interesting to me that it sort of lines up with where builders are interested in building on and what they are building on. We've seen more and more builders interested in Solana and actually building projects on Solana. That was the biggest jump in interest that any blockchain had seen between this year and last year.

Base as well, super high up there. And even Bitcoin in slot number three had a really big bump in people who want to build on it. Yep. And maybe just to add one clarifying point here, the way that we source this data comes from a question that we ask founders in building. We asked them simply what blockchain ecosystems would you be interested in building on? So it does not mean that there are this many builders specifically deploying contracts or that there are a certain number of protocols on these blockchains.

It simply just means kind of where, generally speaking, are these builders interested in building on. And I'd say still today, Ethereum and all of the layer twos, including base and Arbitrum and optimism and ZKSink and the others, still capture a majority of the builder mind share. So I think those were the key takeaways from that data. But you can check it out yourself by looking at the dashboard. Excellent. So this is an election year, and crypto has become a part of the national conversation. People are talking about it.

Politicians are talking about it. Candidates for the office of the president are talking about it. What have we found about how crypto has sort of changed from last year when we did the report to this year when it comes to a policy perspective. Yeah. So crypto has become political in this sense that the crypto industry believes that political engagements can help resolve some of its major challenges. And a big question is whether that has made it to the mainstream or not. Have the crypto industry's issues becomes something that some voters or enough voters care about.

Probably for me, the most interesting slide in the policy section of this report is this analysis of the crypto interest that we're seeing within various states of the union. Walk us through this analysis that you did here and what you found. I think everybody knows that this presidential election is shaping up to be a very tight race. I think it will likely be decided by a handful of counties in a handful of swing states. And so, you know, we wanted to at least understand is the interest based on some of the data that we have, is it growing or shrinking across these different states? And so we basically used Google Trends data. We looked at a basket of search terms that were relevant to crypto.

So we looked at Bitcoin, Ethereum, and crypto. and we looked specifically at the relative change in search interest as defined by Google Trends from 2020 to 2024, and we looked at which ones increased in rank and which states decreased in rank. And then we basically plotted that on a map and showed which ones had the highest growth, which ones had the highest declines, and then basically plotted all of that data. And I think if we look at the swing states specifically, I think the most interesting takeaway is that Pennsylvania and Wisconsin actually had top five, I think they were fourth and fifth in the ranks by change in this relative crypto-related search interest

since 2020. And these are some of the tightest races currently, as indicated by various prediction markets. And so I think there is many different factors at play, and you need to be careful in what you conclude from a very simple analysis like this alone. But I think there is evidence to suggest that crypto is relevant here and could have an impact. Yeah, I mean, it's not a sentiment analysis, so we don't know exactly what those people who are searching for these crypto terms actually think about crypto, but we do know that they are searching for crypto terms, which just kind of indicates general interest and is pretty interesting to note in and of itself. You mentioned Pennsylvania and Wisconsin in the top

five. Michigan is, I believe, number eight, so that's in the top. 10. And then there were some declines. Some states saw a little bit of a drop in interest. One of those is Nevada, a place that has the city of Las Vegas, which is often associated with gambling. So if there's less crypto interest in Nevada, I don't know that that's necessarily a bad thing. But it is really interesting to see that, yeah, this election could come down to a few counties in a few states. And these are places where there is increasing interest in crypto, especially since last election in 2020. And just to add clarifying detail, all of this data is by default from Google Trends population adjusted. So they are representing this data as a percentage of the total searches in the

region. We don't actually have visibility into the actual total numbers by state. That's right. And total numbers anyway would be sort of skewed just toward the states that have the biggest populations, which wouldn't be extremely informative. Another really big thing that happened this year from policy perspective was the Bitcoin and Ethereum exchange traded products were approved. And I'm deliberately calling them exchange traded products, ETPs, and not ETFs, because ETFs, while they are a form of ETP, the underlying portfolios of them would have security. whereas in this case, these crypto ETPs were registered with the SEC in a way that indicates the underlying portfolios or not securities.

Anyway, that's a really important point, but it is the first time we've ever had these market products accessible on major stock markets. It's kind of a big deal. I think it's a very big deal. I think it legitimizes crypto as an asset class. I think it puts us on a path to really having crypto be a part of every diversified portfolio then just makes it much more accessible, which I think is a good thing. You know, in the report, we do look at these kind of net inflow numbers because crypto is all

on chain. We can see when the ETFs are accumulating and distributing crypto out of these products. They haven't been all that high relative to expectations. Maybe it has underwhelmed a bit. But I have talked to some experts, and they do say that getting the distributors activated is just something that's going to take time. You need to go through a long process, you need to get it into the model portfolios, you need to discuss it with your clients, and I think that could be a multi-year process.

And so I'd keep an eye on those inflow numbers over the next few years, because now that we have them, I think tremendous tailwind for the industry. Yeah, I think it absolutely does broaden access to a whole new segment of investors that otherwise would not have dabbled with crypto beforehand. You know, I'm thinking about my dad, for one. He's like, where's the ETF? Where's the ETP, as it were? Because he's probably not going to set up a hardware wallet. All right. Okay. Anyway, so let's dig into some of the other things we found in this report. People always ask this question, where are the crypto applications? But there's one that's kind of been

staring people in the face for a long time. And that is stable coins. Stable coins are a product that people are using a lot. It's a crypto product. I think people discount that. Well, stable coins have been in crypto for a long time. In the earlier days, people used to think of them as something that would end up being for payments. And very interestingly, what we found in 2021, 2021, as the stable coin supplies really started to pop off, was that they were primarily being used as means of settlement between centralized exchanges and also as collateral for defy protocols, something people could supply and lend in a protocol

and then enable other people to borrow and create debts denominated in dollars, which is a lot easier to reason about than a debt denominated in Ethereum or some other volatile asset. Another reason why stable coins make a lot of sense as collateral is that if you're borrowing a fair amount on chain,

it may be totally reasonable to pay a $30, $50, or even $100 transaction fee to perform that borrow because it still may be meaningfully more affordable than other ways to engage in an over-collateralized loan. And that is really the type of use case that we saw for people in the 2021 through 2023 period for stable coins. But as a result of the infrastructural improvements in 2024, where lots has come online, not just 4844, but the proliferation of L2s in Ethereum world generally. 4844 being the Denkun upgrade that Ethereum went through in March of this year, also called Proto-Danksharding. Yeah, so as a result of those infrastructural improvements, as well as increasing adoption in newer generation, newer architecture, or blockchain.

like Solana, Sui, many others, now sending stable coins in a single transaction often costs a penny or less. So because of this, just this year, because of these improvements, I think a lot of people are talking about stable coins as a medium for payment again. It's really interesting to me because I can't emphasize enough how in 2022 this is just a non-starter. Everyone had given up on it. I don't think I heard the word payment in 2022, except to debunk the possibility that stable coins were being used for payments. Whereas now, it's something that I'm hearing about all the time. And it's not just in a conventional pay for your coffee with crypto setting, although that is actually a real possibility. Coinbase had to pay for a coffee with the USDC

day in New York not long ago. It was a lot of fun. But more concretely, now people can play with blockchains as a piece of infrastructure to potentially bring down interchange fees and merchant network fees, all the types of fees that go into settling an actual consumer payment for something. I just preordered a book from Stripe Press with the new Stripe Pay with Crypto feature. USTC on Solana was fantastic. It was just a really straightforward experience for me. Of course, I already have crypto in my wallet, so that made it. pretty convenient, but these are real things. Right. This is because you've got to still onboard to like a layer two network, say,

an L2, a roll-up, where the fees would be lower. But once you're there, it becomes quite seamless to transact and move around. Yeah, and there's still very meaningful UX challenges. But the point is to say, is like, now this is in the story, and there's a lot of crypto projects that are interested in trying to improve payments or aspects of payments or take advantage of the ease of settling payments internationally, computationally. There's all kinds of ways that payments really ought to be programmed and experimented with, and now just this year, that's a possibility. So just to set people's expectations, like the way I like to think about it is,

now that there is a glut of supply for being able to transact stable coin dollar-denominated payments, how long will it take for builders to consume that? extra supply in terms of interesting applications. It's not instant. People don't realize this. People kind of like see new technology. It sort of appears. But from my perspective, it takes at a minimum six months to start to see people really talk about it. And then it takes about two years for people to soak up all of the excess. This reminds me a lot of the midwit meme where it's like maybe it is as simple as just, you know, when the fees come down, Stable coins work well, right? When the fees are really high, stable coins kind of suck.

And so I think to all of the points that Eddie made in 2024, we made a ton of infrastructure improvements. It brought fees down just to share some data from our report on this. In 2021, to send USDC on Ethereum using the average gas price, it was about $12. That's like, that's a lot of money. That was like my lunch that I just had today. Yeah. And as people took for granted, like that's obviously totally unacceptable. for a payment setting, unless you're moving like tens of thousands of dollars, then it actually may be better. But there's a smaller number of people who are in that situation. Definitely. Today, if you use USDC on Ethereum post EIP 4844, it's like $1. But if you use USDC on base,

which is a layer two, it's less than a cent, right? And so that makes all the difference when it comes to Stablecoin adoption, and it's no coincidence that 2024 was the year that Stable Coins found product market fit. And it's, I think, because it's as simple as just the infrastructure got better, fees came down, and now it works. It's really amazing to wrap your head around this that just a few years ago, it cost $12 to transact. Exactly.

It totally changes the applications that are possible. And I think there's a number of examples that we can point to that are only now possible in a low-fee environment. And we do our best to highlight those emerging applications in the report. This doesn't just speak to staple coins. It also speaks to other sorts of applications that weren't possible before. I mean, I would say the most striking thing to me in this report compared to past reports was, look, NFTs were going crazy a few years ago. The secondary markets, you know, NFTs were trading on them for billions of dollars regularly. That activity has subsided. But now, because fees have come down, we're seeing totally different sorts of

consumer behaviors arise around NFTs. I think a lot of people might think that NFTs happened and then they went away and they're over and done. And your sort of mainstream listener here might not realize that they've actually stuck around and they're just gaining traction in a completely different context. Being honest, the NFT market has faced some pretty tremendous challenges. There was a lot of speculative activity in prior years, very high trading volume, high priced premium collections. We definitely got over our skis a bit as an industry. As a result, we've seen a 90 plus percent decrease in trading volume over the last few years. There's been a ton of drama around how creators can enforce royalties. And so we have faced a lot of challenges. And I think there are

some data points that you can look at that say, wow, the NFT market really is not what it once was. And I think that is very true in all senses. But I think what's also happened at the same time is to the point we were just discussing about the infrastructure getting better, these infrastructure improvements enabled NFTs to be exchanged and traded and interacted with at much lower costs, which really kind of shifted the behavior away from these high volume, high price speculative secondary markets into these newer, low-cost social collecting experiences. For example, if you look at the number of NFT collections with 50 plus unique mentors, the trend is actually very positive.

It means that, you know, people are still interacting and using these NFTs and collecting these NFTs, but they're doing it much different than they had in prior years. And this is only possible in a low-fee environment. You can look at products like Zora today. You can look at products like Rodeo today. And there are some really fascinating, cool, new emerging cultural behaviors around this social collecting that I think many people would find really interesting. And so I would say the NFT market is very much alive, but I think behavior has shifted.

And I think it's all driven by the infrastructure developments in 2024. Yeah, I don't know that there's a perfect way to communicate what this new behavior is other than like people like to collect stuff. And on a social network, you're probably familiar with liking hosts. And I've heard people describe this new sort of social collecting experience as a kind of super like behavior where you're just expressing that you're interested in something. And critically, it's on the order of magnitude of like one to two cents per collecting. In prior years, we were paying tens of thousands or hundreds of thousands of which, of course, only very few people can afford.

But now we're opening it up to a much broader audience, which I think is a net positive for the NFT market. We got to recall that the way crypto has developed is the first thing that you can do, and actually the simplest thing to do computationally, is just to buy and own the thing. That is the novel component is being able to directly own the underlying asset on this big shared state computer. doing more complicated things

like transacting or mutating the state of a program these require much more computational capacity in addition to creativity from application designers in addition to user experience improvements to make it easier to do those things we're seeing advances on all of those fronts and so you should expect the complexity of what people do with blockchains to increase over time

starting from the simplest which is just owning and moving to the more complex things, interacting with programs. Even just the mental shift that it takes to understand what is activity on a blockchain, it's affected by this evolution where it's not just a money payment system. It's actually a shared state computer, as you called it, Eddie. And these interactions can be interactions with applications, just like people are interacting with, like, websites or something. All these interactions can be much more expansive

than this sort of limited view of crypto just as like a money thing. Yeah, and I want to sketch out like there's a pattern to how these things unfold, right? Just buying an NFT when the costs are very high, that's the simplest thing you can do,

that's the best supported thing you can do, and that's probably the only thing that's reasonable to afford, but NFTs are programmable. Experimenting with what that looks like requires low gas fees. So what I expect, like the exact pattern I would

expect would be speculative boom with very simple uses where most of the interesting things that happen in the very, very earliest stages are actually the themes and the ideas that come through as people sort of imagine what's possible. You see glimmers of the very exciting things. Of course, then they run into the pragmatic reality that it's too expensive. It's developing too slow. There are certain pragmatic limits. As those things scale, then experimentation returns, and then as experimentation develops, more complex and interesting applications start to take place. We just describe that with stablecoins.

The exact same thing is going to happen with NFTs and with all these different settings. It's not a crypto-specific thing. I think that's how AI is developing as well. Because you mentioned AI, I think people listening will recognize that AI is one of the biggest trends of the past year. But what's interesting is that it's not just a trend. within tech. It's also a trend within crypto. Maybe it's unsurprising that crypto influencers are talking about AI. But what is more surprising is that the people likely as to visit

the websites of various AI tools like ChatGPT are also the likeliest to visit some of the top crypto websites out there. And that goes to this sort of inference that could possibly be drawn that there is strong overlap between the user bases of these kind of frontier technologies. Yeah, I mean, definitely it's a hot area. And AI was clearly the breakaway narrative in 2024, even among crypto social circles relative to other topics. And so clearly people are thinking about this. And a lot of the top crypto sites are highly, highly overlapping with chatGPt.com, which indicates that there may be an overlap in crypto and AI users. And the fact that there is, I do think reinforces this point that it's a hot area, and I think we should expect to see

more innovation in this pocket of crypto specifically. So let's cover some more emerging applications. The cost of transacting on blockchain networks is coming down. Like we said, it can cost as little as cents on the dollar now, or even less than a cent. Just a tremendous limiting of costs there. What other territory does this open up? Yeah, some areas that have popped up include D-PIN. There are new forming marketplaces for energy, for wireless bandwidth, for mapping, for weather, for food delivery.

This is decentralized physical infrastructure networks. So it's where the digital world meets the physical world, like using a digital overlay to manage resources in the physical world. Yeah, and the difference, you know, of course we have these things like with Uber and Lyft and so on. You could call those web 2 deep end networks. The key difference is that in a blockchain, there's a degree of credible neutrality about the underlying platform. That means that people who build on top of it, including not just the suppliers of the good and not just the people buying it,

but also people who build applications to make the whole thing work better, can take for granted what they're building on and not risk being rent collected or sherlocked by the underlying platform. And if you can make payments affordably, If you can update reputation systems, incorporate all kinds of interesting identity systems, all these things that these marketplaces need, then they can finally exist. So those have been emerging. Maybe a final category, which we give a little glimpse of in the report is decentralized social networks,

Farcaster, which is another portfolio company of ours. It's an incredible place for builders to experiment. with different types of tools, different types of toys to extend and improve the experience for all types of users. And I know for a fact, just having spent a bunch of time with the Farcaster team, that they are totally capacity limited. They're working really hard to allow the network to post more and more people while remaining totally decentralized. And you can see in their report a lot of projects where people have been experimenting with software on top of this network. Even in our Builder Energy dashboard, we're seeing that a lot of builders are interested in

social networks. I think 10% of the founders that we've talked to and looked at are building social-related projects. So I think people see this opportunity, and they expect that it's going to be a thing, given how big it was in Web 2. Yeah, then there's prediction markets. Of course, those have been huge this year. No surprise, given the incredible dynamics. and volatility related to the election? Yeah, prediction markets is an idea that people have been talking about for decades, but have been hamstrung,

and people are actually coming around to their value, an ability to source up knowledge from disparate arts of the world and kind of get a better, more informed picture of the state of the world and how things might shake out. It's interesting to see that it's actually finally getting some traction

when it's a very old idea that has been stalled for so long. All right, Eddie, we've been putting out reports for three years. What's been most surprising looking at this year's report compared to earlier reports? I don't know. Stepping back, I'd say when we're deep in the mix, like we are, it often feels like nothing is coming fast enough. I feel that all the time.

I want it right now. I want it right now. I just take a waiting. And there are some areas that have definitely lagged. But at the same time, it is a little bit unbelievable when you look back at the progress. You know, let's be very conservative in our estimates. And let's say we have 30 million monthly active users of blockchains. That's pretty damn good.

Like, I'm pretty happy with where that is given the historical infrastructural limits. Like, about a year ago, we were all debating how long it would take things. to start to heat up in terms of on-chain usage. And where I ended up was that six months after Denkoon, six months after 48-44, is when we'd start to see, like, new types of stories because that's about how long it takes. And we're just like six months after,

and I think that that's exactly what's happening. Like, I think the payments thing is one of those things. And I think some of these new networks, some of these like deep end networks are examples of that. So it feels very on track to me and I'm very pleased with that. That's kind of surprising, maybe in a certain way. That is surprising to me that it feels very kind of like we are thinking it's going to unfold. If you had to cast your prediction a year ahead, what do you think is going to be the theme of state of crypto 2025?

I hope that all the extra block space is filled with something, something productive. That's what I hope it's about. It's about what fills that block space. We just started creating all these L2s and new blockchains. And maybe something that we didn't put in the report, because it can be challenging to present in a very straightforward way. But bridges are great for people in crypto. You know that three years ago,

there were very few bridges and they were terrible and they were slow and they were expensive and they were error-prone. There's just a lot wrong with them. Now, bridges, there's myriad options, really high quality, fast, super affordable, great connectivity. That's like, no question. Anyone in Crypto would agree with that. Well, now that we have good bridges and now that we have good L2s,

we're getting tons of more L2s and tons more bridges and tons of more interoperability questions and fragmentation. So as always happens with technology, you solve one problem, you create another. I'm beginning to see signs that there are really, really great strategies to solve interoperability. And so I hope the combination of good interoperability plus good block space means finally great user experience for a family of applications. If that was the case at the end of 2025, I'd be really pleased.

That's great for the wish list. I really hope that in the coming years, our section on infrastructure gets smaller and our section on applications gets bigger. I think there's still a place to report on key infrastructure developments, which I think will continue to happen over the coming years. But I think it should be hopefully more about how do we just overall improve the user experience. You know, there's count abstraction protocols, interoperability, various developments that we can report on. But I hope that that is not the main story. I hope the main story of 2025 and the years following is that we now have a new set of applications that,

are proving to be real use cases for this technology that is attracting people. And so that is my hope for the years to come. We've made a lot of progress on infrastructure, but hopefully from here going forward, the improvements are more on the margin. Maybe in the years ahead, there won't even be a need for a state of crypto report. We will all just be using these applications and won't even know that there's crypto on the back end. Could be. We'll have to wait to know for sure.

Thank you both for joining. Thank you, Robert. Thank you, Robert.