S4:E2 Elias Simos of Rated - Creating Reputation for Machines
Until now, we have had almost no way of evaluating Ethereum validator performance. This lack of accessible data around nodes and their operators has been keeping users, builders, investors, and other validators in the dark about the very health of the Ethereum network. But Elias Simos and the Rated team have created a solution.
Elias joins us today to shed light on the state of Ethereum infrastructure. He explains the importance of accessible blockchain infrastructure data, how Rated turns that data into usable performance ratings, and how the idea came about in the first place. We also discuss the challenges that come along with building a product like Rated in a multichain world, and Elias sends us off with some advice for staying sane while working in crypto.
📬 To keep up with the latest from Archebyte and receive articles and research from the rest of the Archetype team, subscribe to our newsletter: http://eepurl.com/iCApL2
- - - - - - - -
2:39 Actors in proof of stake
9:52 Rating onchain performance
17:21 Node, validator, and operator reputation
23:08 Multichain reputation
29:18 Health of the network
32:08 Staying sane as a crypto founder
👋 FOLLOW US
Rated Explorer: https://www.rated.network
Elias’ Personal Site: https://eliasimos.xyz/about
Hello, everyone, and welcome back to Archebyte, a bi-weekly podcast focused on the builders and innovators in crypto. Today we are diving into the world of blockchain infrastructure data. While it might not seem like the most glamorous topic, it is very important to the health and security of each blockchain network.
Our guest today is Elias Simos, Co-founder of Rated Labs. I have had the pleasure of knowing Elias since his days as a protocol specialist at Bison Trails, and then our paths crossed again at Coinbase. We at Archetype are now so incredibly honored to support his latest venture as one of the founders of Rated Labs. With years of experience in blockchain infra/data, Elias is with no doubt the perfect person to help us unravel the mysteries of this topic.
By the end of this episode, I hope to collectively grasp the significance of a reputation system for machines. With that - hi, Elias. Welcome to the show.
Thanks for having me, Katherine. It's awesome to be here. It's awesome to be at another crossroads with our intertwining paths and super, super excited to have you guys on board for the journey.
Yeah, we're very excited. And also a fun fact, this isn't the first podcast we’ve recorded together, so it feels very full circle. On the topic of blockchain infra, before we really dive into everything about your building and thinking about, I want to first level set and map out all of the players that are critical to the security of a proof of stake blockchain like Ethereum in a post-Merge world. So paint the picture for us. Who are all the important players?
So maybe it's worth just talking about proof of work and proof of stake very briefly. They’re both security algorithms. They both sort of secure state transitions in distributed systems so that the truth is recorded with high fidelity and very high guarantees so that you can have things like not like a platform where you can exchange money digitally, but you can’t also double spend it, right? It's not like I send you a dollar and then I copy paste the file and then I also keep the dollar in my wallet. So they're both security algorithms, but they have very stark differences, right?
Proof of work is the algorithm that powers Bitcoin to this date. It's basically the algorithm that gave birth to this whole industry and the inputs to proof of work in order to produce the security is electricity and computing power.
So you run all these computations and you try to find the hash, and that's how security gets produced. This is a very electricity and energy intensive process. And part of the transition to proof of stake, at least for Ethereum and a large part of the industry, was saving on the energy footprint of these blockchains.
Also because capital, which is the input into proof of stake — so, you know, proof of work, you have electricity, computing power — proof of stake you have capital, value at risk, and operational excellence as the inputs to this security algorithm. And then it produces security. You know, large CapEx expenditures. It’s really hard to bootstrap a network when you have to have like all these miners and machines consuming actual energy. So it's much easier to actually bootstrap a network, but at the same time, it's a lot more energy efficient. Right?
So what we're looking at is Ethereum validators, at this stage at least. And so given that these are the two most important inputs into this algorithm, capital (i.e. value at risk) and operational excellence, you have people that allocate the capital that actually put that value at risk and they have different shapes and sizes and we can get into that. And then people that operate the actual machines and make sure that, you know, they're running the correct version, they are doing upgrades promptly and with low risk, they are operating in a way that ensures that the validator won't actually get slashed, which is basically the carrot that the protocol offers.
Also worth noting that in proof of stake, it's largely a game of truth telling. A bunch of machines kind of observe the state of the world, and then they say I saw this, I saw this, I saw this, I saw this. They commit their versions of reality to a common pool of versions of reality. And then there's someone tasked at every round that the game is played — these rounds are effectively slots or blocks or whatever they might be called in different systems — and collates all of these opinions, effectively these views of reality, and then says well, you know, the majority thinks that this is reality, this is space and time. And so this is what we take to be the canonical space and time.
So in that game then you have allocators, which can be capital pools, can be high networth individuals or whatnot, could be exchanges, it could be decentralized protocols like the likes of Lido or Liquid Collective and so on. They basically pool capital to distribute it to node operators.
And then the second most important agent in the system is the node operator. And you have like a pretty wide net, not that wide, but they don't come in one box, right? You have enterprises that are involved in that. You have exchanges that have in-house teams that run nodes, you have companies that their sole purpose is to run nodes in these networks. Then you have some sort of, you know, early cypherpunks that do it from home or have smaller operations, but they're not taking effectively delegated capital from other folks. And then you have people that are doing it from home when the requirements for the machine that actually runs the validator allows for such a thing to be run at home, right? Because different networks have different resource requirements.
Now, thinking about then these systems more broadly with Ethereum transitioning to proof of stake, then we actually started seeing the role of the proposer, which is basically the agent in consensus that collates all these versions of these attestations to space time, these opinions effectively, and then creates “reality.”
You saw that role actually be decoupled from the actual validator. The validator is still the proposer, but actually crafting the contents of the blocks - what the validator actually commits as sort of an attachment to their view of space time, that's kind of what happens is the consensus layer and all these messages. But then at the same time, this is how security gets produced.
And then there's another job that traditionally that role needed to perform, which is, okay, let's attach all these transitions, all these valuable things to that security that's produced. And that's transactions effectively. We write everything that's happening, specifically on Ethereum, on the execution layer of Ethereum — trading on DEXes, interacting with lending protocols, trading NFTs, just sending money, peer to peer, all these kinds of things, right, that make a whole large transaction space.
So you now have different agent types that sit outside of the core protocol that are specialists in actually building those blocks that then get effectively attached to the security that the validators produce. Right? And these are called builders and they're optimized agents that order transactions in ways that are more valuable than sort of the vanilla way that the protocol might prescribe.
And they communicate with proposers through sort of a neutral agent in the middle, which is called the relay, which is effectively a load balancer that kind of protects the validators and the core consensus of the protocol from getting flooded, for example, with packets of information and then having sort of exposure to denial of service attacks, which would then kind of be a network threat or different types of economic attacks. Right? So they provide this load balancing buffer in the middle so that this exchange of value from the builder entity to the proposer can happen in an orderly fashion.
Okay. That was like a very broad lay of the land of all the players that really make up the lifecycle of a transaction. And they're all like they all played their very important parts. And so obviously I think it's very important that you can sort of gauge their performance, although I think it maybe sounds easier than it actually is. Has there ever been a historical method to really gauge the performance of each player and why or why not?
The answer is no. There have been attempts to do that. But from where I stand, and this was part of the reason why I went and founded Rated and we're doing the work that we're doing, is that I saw a lack of that throughout and across different networks while I was at Bison Trails. For those folks that don't know the pre-history, it used to be a company that did infrastructure as a service. We had built a very large platform where people could run validators. And I think at the height of the market it was 20-25+ billion dollars on the platform operating across 30+ networks. So we got to see a lot in that.
But the one thing that was so consistent was that across every network, I couldn't really tell how we're doing versus our network peers. I couldn't tell how we're doing compared to the network. I couldn't even easily tell how the network is doing at any given point in time, let alone its history. And then also, because everything was so new and largely still is, there wasn't even a common language to actually approach what does it mean to be doing in the most general sense?
So when I try to answer the question as an operator, how am I doing versus the rest of the network, well, what do I need to look at? How do I define what this health or effectiveness or whatever that might be? And so very early on, we saw a lot of dissenting opinions. And also because the industry was growing so fast, not a lot of attention into that particular problem. When growth hits you really fast this is not the first thing that you're actually going to go and take care of.
So that experience, together with a data hackathon that the Ethereum Foundation put up on the last testnet before Beacon Chain mainnet, and I did that with my friend Sid Shekhar. We came in second or third, we won silver prize, and we did a whole analysis of all the data that the testnet produced and that's where we started sort of grasping that issue, that there is no common language, there are no standards, right?
And this is as much data and lack of data issue as it is a coordination issue. And there are multiple ways to address the coordination problem. Like you can you can address coordination with sort of explicit coordination in the way that DAOs do. Lido, for example, has been really great at actually coordinating operators and slowly introducing some standards and so on.
But I think there's a lot of power in interfaces as well, right? There's power in interfaces and there's power in products to actually push for coordination, which happens in an implicit way.
So what we've done specifically for the Ethereum infrastructure space, which was really like how the company started, is we created this effectiveness rating for Ethereum validators, which exclusively looks at onchain data, and then it basically strips down all sort of the noise that rewards might create because a lot of sort of the rewards that validators earn are stochastic, they're random effectively, and just looks at the jobs that a validator is called to do and just like keeps a point system effectively and says did you do your job or did you not do your job? And how fast, how promptly did you do your job? Because that also has a say into how well an operator is actually operating, right?
So that was kind of like the start of everything. We put up a dashboard that I had designed on Figma, which is laughable because my skills are just not, not all that great, but, you know, ship fast, right? Like if you're not embarrassed of your first version, then you're probably too late, as the old adage goes.
And then we really saw people gradually actually like, pick it up and find value in what we were offering. That also came coupled with very open and transparent methodologies about why we're composing that specific metric, how exactly we're composing it. We've had operators over time go and reproduce our results just with their own data and arrive at the same outcomes, which is always great to see that people go through the trouble. And then also in the process actually poke holes in our robustness and how well we represent basically the things that were set up and we seek to represent.
And this has brought us to today basically where I think besides the network explorer that we host, having grown in popularity and so on, and actually mediating a lot of important decisions in terms of allocators allocating to node operators, operators benchmarking themselves and improving themselves over time to things like SLA setting between capital allocators and node operators. I know of various cases where organizations are basically indexing on the effectiveness rating that we published to actually set SLAs for someone delegates capital to a node operator, the node operator will operate validators on their behalf, usually in a non-custodial way, but then, you know, there is a performance guarantee that they must provide, especially in enterprise settings this is actually quite common.
So I'm aware of multiple instances where these SLAs are mediated by the effectiveness rating that we publish, which goes to show you the power of the interface, right? There was no real explicit coordination that we pursued. We're not a DAO, we're not set up as one, but we still have a public Discord full of node operators, people are more than welcome to come and ask questions and solve their queries and poke holes in methodology.
And through this process we've gotten to this point, which a really sort of interesting thing to look back and reflect on, but that said, there is still so much work to do in terms of actually contextualizing all the useful data that kind of make up a node operator’s useful life in a network context.
There's a couple of things that you've said so far that I really, really like and just to reframe it — honestly, when it really comes down to like all of this, it's a game of truth telling. I know earlier we were chatting and I was like, would you say this is more a data issue? Is it more of a coordination-like dilemma? And I do think it is a bit of both, but so much of the emphasis is on the latter. And maybe that's not like no one's ever done it, but having a place where so much of the data is already parsed for you, already sits there, things like slashing or uptime or whatever, like they're all in one place can be so powerful.
And it's one of those things where it's like you don't even think about how important it is to look all of this up until you actually see it laid out in front of you. And I think the tagline for Rated - reputation for machines - it's very catchy, but it also is exactly what it says it is, which is just like maybe you haven't seen such an aggregate rating for an effectiveness on all these players, but, here it is. And once you see it out in the open, that's where you can actually compare it. There's some sort of standard that starts to get set.
That's right. This tagline just helps give us a sense of direction, but without being overly narrow and just limits us of optionality. So once we sort of stumbled upon it — and I even forget how we stumbled upon it — it just rung very true because, as you said, it does capture all of the things that we kind of stand for and that instantiate through product, right?
So another element of reputation is identity. And what identity is and how it gets defined and how it gets composed and how it gets refreshed over time and so on are incredibly complex subjects to even define or to even put words in a sentence and get meaning out of it when it relates to this. But it's something that we aspire to work towards and to build and to instantiate.
So when we started on the explorer, the goal was to show that like, okay, well if you just look at specifically Ethereum and the way it works, there are digital instances of consensus units which are called validators, but one validator doesn't equal one node operator. In fact, one validator can be a group of many, and that group of many when they are operated by the same entity, that makes up an operator. At least in the way we understand the world. And that node operator could be part of a larger set, which is a pool of node operators that sort of adhere to certain standards, way of operating, they just receive capital from the same sort of pipeline, if you will.
But initially Ethereum doesn't have an identity system baked into the way the infrastructure works, like other protocols do. For example, in the Tendermint world, it's only like, you know, a fixed cap of validators normally. There isn't one in Ethereum. You have like 100 or 150 in Cosmos, Tendermint-land validators, and there is a naming system so you can go and like actually, like reveal your identity.
But also the consensus unit is the operator really. It's not like a validator is also an operator and it's, I think, rare that you see in these types of networks, multiple validators run by the same operator because there are certain seats and the number of seats are really relatively limited.
So what we went out and did is first kind of pulled from onchain registries like the Lido schema, for example, and the operators as they're mapped from the view of Lido. Or open source research that people have done to go and tag different addresses, deposit addresses or withdraw addresses that map to these validators.
And because after a point we started getting a lot of requests for people to list their sets on the network explorer, like oh I'd like my set to actually be part of the views that we offer, we released what we call the Self-Reports API, which is basically a way for people to just report their pub keys to our explorer.
So we released that in the summer, we have about 20% or so of, of the beacon chain reporting through the Self-report API and then together with like all the other work that we've done and so on, I think we've gotten to around 70% plus of the keys that make up the beacon chain and there's like a longer tail of that, that is solo stakers that are like really unidentifiable, right? They're not entities and they shouldn't really be identified.
So that's kind of like a really interesting aspect of reputation because you have to have some identity to start building reputation, right? And I see kind of like that work that we're doing with the Self-Report API is yet another lever that we're pulling to get ourselves closer to just really instantiating this concept of reputation.
So you have, you know, performance. We're going to probably launch features relating around risk, but all of that is just blobs in the information space if you don't have identity. And that's another part of reputation.
So really what we're trying to do is to kind of weave all of these concepts and data — some live onchain, some live offchain, some you can get permissionlessly, others you need permission to get — and really put it all together so we can have a spherical view of what's happening at the core of these networks. Because we really think that these are amazing pieces of technology that we can really build a lot of alternatives and better alternatives to the current way that things operate in the world in terms of finance, in terms of trade and commerce, in terms of a lot of things. So we take our mission very seriously, perhaps a little bit more seriously than we should. But we're really excited to be on this journey and to have such a broad remit of things and big, important challenges to tackle.
Well, the other thing about reputation is that reputation can only be established in a relative way. It's actually somewhat subjective, but it's sort of based on an objective set of a larger data, right? So it is relative and it is important to have that wide set of data. Since you mentioned product earlier, you guys have built a number of dashboards. Just as we established earlier, validators aren't just like one set, there’s actually a lot of diversity even within validators. But you guys have so far built various dashboards for pool operators, node operators, as you're building currently, how do you think about the balance between going the multichain route, which is like incorporating more proof of stake blockchain networks and onboarding them to your dashboard versus expanding infinitely with the Ethereum ecosystem.
It's a great question. It's one we periodically grapple with and we've landed in going multichain because we think that in the future we're heading towards, Ethereum is by far the most consequential network that we've got at the moment, right? At least insofar as networks that you can do useful things on. You can build application logic. They're more expressive and so on.
Bitcoin is, you know, important, great, but also like, it’s a pet rock. It doesn't do much. It's a meme. There's a lot of good work that's happening in Lightning and people building ordinals and adding value to Bitcoin. But Bitcoin has been incredibly successful because of that non-sovereign money idea and now it's transcended sort of the world of technology. And it's really kind of like a cultural phenomenon, right? The idea is much bigger than the technology itself.
So with Ethereum and then all the other proof of stake networks, they're all about application logic, right? And there are really, really valuable properties that they have. But I'll be very hard pressed to say that, you know, one network design with all its past dependent architecture, some of which might be desirable and some of which might be not, because we're all, to some extent shooting in the dark. Like there were things that, you know, Ethereum was hoping to achieve and the designers of the protocol thought that like, well, our design decisions today will lead to certain outcomes that are not true. And then looking back, you're sitting there and saying like, well, that was obvious, but hindsight is always 20/20, right?
So if we then take that to be true for every single network, then if we hope this industry becomes as big as we think it might be and has potential to, I can't see how it's only one network that dominates the spectrum of applications that run on blockchains. So that's one thing. And I personally have a very long term view in the things that we're doing. We as a company have a very long term view in the things that we're doing. And so we're building with that principle very close to heart.
The second thing is that if we hope to serve node operators and capital allocators, pools, and so on better than we do at the moment, then these are organizations that operate across multiple networks very often. An allocator wants to allocate resources not only to Ethereum, but maybe they want to do it in Solana, or maybe they want to do it in Filecoin. And really at the highest level, the more we apply ourselves to this problem and the more we think about it and the more we think about how our product is going to instantiate it and so on, there are these master abstractions of rewards, of performance, of risk, of metadata that make up this concept of reputation that are actually pretty easy to translate — easy again, in air quotes. It's not actually easy, but if you develop the expertise and the design, it actually becomes a lot easier than then you'd think to abstract sort of protocol rules to these archetypical categories that make up reputation. Right? And then it's both a hunch, but also like we're getting some pretty interesting sort of primary data as to whether that hunches is going to pan out to be true or not.
But there are interesting things that might happen at the intersection of many networks and their defined abstractions and so on and the value that this might create. Again, for node operators, for capital allocators, for applications that reference these networks, in particular the infrastructure sets of these networks and of course the networks themselves so that, you know, even networks might be able as node operators on Ethereum today look at the networks for the benchmark themselves. They learn from the information that we curate. I have a sense that we also might be able to do that for networks down the line.
Yeah, and since you're going after other proof of stake networks, can you apply essentially the same-ish methodology to other types of networks? Like are there fine tuning that is specific to — I mean, presumably, the players are probably largely similar profiles, like you mentioned.
That's right. It works in two ways, right? So we can't fit what we've done on Ethereum bottom up. We can't go in and take the same sort of first principles or being-true-to-protocol approach and then apply it to something like Solana or something like Polygon or something like Avalanche. But what we can do is work both top down and bottom up, or at least work bottom up, but like net new, but with the goal of we know where you're going to get to, we know that there are three or four abstractions that we need to instantiate, so let's take all of the data that these network produce, let's attach semantic labels to the different data points originating from protocol rules, just really being true to the protocol and representing what the protocol is about — because otherwise our work has no value — and then just make everything more interpretable. Because that's like really the reason why we're striving to get into these abstractions and instantiate work that we do in other networks, into these abstractions, it's because we need to make it a lot more interpretable and consistent with what we've discovered in Ethereum.
Yeah, and taking it back to the top of the episode, you're sort of getting ahead of the coordination dilemma. For the other networks as eventually, as you gain momentum, people start to ask, Hey, what's the health of our network? Who are these different players? How are they doing? So in a way, you're sort of also getting ahead of that.
We certainly hope so, and I think it will pan out like that. But it's not as straightforward. The value is not as straightforward to people before the thing exists. Also as it were in Ethereum when we started putting sort of the early sketches together and so on and socialized the idea to different folks around the ecosystem. We got a lot of no’s, we got a lot of this is a shitty idea and that's the rite of passage I suppose whenever you start something new. Like obviously not everybody is going to get it and they have every right to not get it. So I don't hold any grudges to anybody. But equally I think it's not obvious to folks.
We have a very strong idea about it. But you know, as a founder, you always grapple with these things. Yeah have a strong idea, and you've seen similar patterns play out in the past, but then at the same time you're like, well, what if it doesn't work? And the feedback loops that you're exposed to are not like, What if I put my finger in hot water? Well, you know what's going to happen, or if you don't know, you're going to know pretty soon. But that one is like a multi month, multi quarter feedback loop. And so, you know, that's just like what you need to grapple with sometimes. But we think that kind of fundamentally it makes sense and we're excited to see what sort of flourishes on top of that sort of intermediary state, Base Camp 2, as we call it, when we have these, you know, multiple network footprint, common abstractions.
We're very, very excited to see what the path to Base Camp 3 looks like. And then, you know, what getting to Base Camp 2 actually enables.
I want to switch gears entirely and sort of ask the last question of the episode, but I want to move it away from product and move away from blockchain and let's talk about being a founder in general. The experience of being a founder of a crypto company, especially in Q4 of 2023. Things are all fun and good when the markets are hot and everything's great, but when things slow down a little bit, there are days where I'm sure where it feels like a slog. And I think it would be really important to talk about just the idea of burnout and especially just feeling sane in this current market.
So, any advice? How do you stay sane? How do you avoid burnout? Any advice for other founders out there?
Yeah, it's tough out there. I'm not going to lie. How do I stay sane? Eat well, sleep well, be surrounded by positive energy so much as you can inside your company, outside your company. There's no room for negativity for its own sake. Criticism and negativity are two different things, right? Exercise a bunch just to just keep your brain focused and really a sense of forthrightness and rigor in terms of what experiments you're running. And that gives you a certain sense of structure, right? Even then, you know, you have doubts, you have uncertainty. You're exposed to a lot of uncertainty and you're like, well, what if I do everything right, but we end up being nuclear engineers And this is 1979. I once met someone who spent his career as a nuclear engineer, he basically studied to be one super hard. He was top of his class in high school, top of his class in university, a celebrated PhD student and then goes into nuclear engineering and it's 1979 or 1980 and then it's all downhill. So he worked so hard, like in a field that was celebrated, as I remember when I was a kid, like my grandma used to say when you grow up, you'll be a nuclear engineer.
And his whole career, he did everything right, even in his career. And then it had like a very, very hard ceiling that was imposed by the world's view on nuclear energy. So that parable is there to say that even if you do everything right, even if you sleep out, you eat well, try and stay sane, balance, be surrounded by positive people and really just be forthright about the experiments that you run, controlling, input, output, measuring, and just really kind of iterating, there is still a specter of things that you can't control. And I guess the only thing you can do you can do there is just observe it and let it pass. Because you made a decision. You believe in certain things and there's nothing you could do. There's some things that you absolutely can't do anything about. Or your probability of just tilting anything is just so minuscule that it just doesn't feel like a good way to be spending your time or energy.
One thing that I think about a lot is like when things feel really, really hard — and you know, this could be whether you're a founder or you just work at a company, like a crypto company right now — like I sometimes stop and I'm like, this sounds so silly but, when you're just focused on the hard things, things all feel hard.
Yeah, but on the other lens, sometimes I think I'm really lucky to even be in this seat, to be in a place where you're pursuing your passions, like, so hard, where you care about something so much like so many people chase their entire lives and never feel that.
It is hard, there are moments where you just are like, I'm tired and I don't feel well and whatever. But like, at the same time, how many people get to literally chase their dreams? You know, like as a founder, you're working after something you truly believe in day in and day out. You're solving a problem that you truly believe needs to be solved like that in itself is a feeling that like a lot of people never feel in their lives.
Unbeatable. I wouldn't do anything else. Right? Just to be clear, I’m where I need to be and I'm probably happier than I've ever been. But two things. Jensen Huang, the founder and CEO of Nvidia, did an interview recently, and then at the end of the interview he gets asked, So knowing what you know now and being at the seat that you're in, you're starting the company today, what company do you start? And there's a moment of silence and he goes, I wouldn't do it.
And then there's like this awkward silence. And then he goes again, I wouldn't do it. Yeah, because it is blood, sweat and tears. But he never quits at any single point in time. He never left the thing. So he got to the end of the road and then he said, I wouldn't do it. So it's this like two minds do track, like the reflective mind and then the actual like being in the heat of battle.
Well like you know what, your dream job is also har. I mean everything isn’t like flowers and rainbows and butterflies or whatever. Like your dream job is still effing hard.
Yeah, that's right. It's like, I guess like professional athletes as well. Like kids dreams to grow and be professional athletes, a tennis player, for example. And then even at the highest level, it's just so excruciatingly hard. You know, you're like in an airplane every week, you're training super hard. The mental game is just crushing. And yet this is your dream life.
So yeah, it's interesting how the balance of these two things like that interpolate across one another. It just stands out to manifest.
Yeah. Well I think this is a good note to end on. Last thing, just tell our listeners where to find you, where to find the work that Rated does.
Yeah so we're under rated.network and you can find us on Twitter at @ratedw3b, I also hang out on Twitter, although as I got into the founder life my time spent on Twitter actually diminished, but I'm on @eliasimos and you can also find me at eliasimos.xyz.
That's a green flag. May we all spend less time on Twitter. Okay, well, thank you for spending the time with me today. I know we started our talk about blockchain infra and we ended up with life lessons. So, you know what? I think that's a great place to end. And thank you so much.
I appreciate it, Katherine. Thanks a bunch.
- - - - - - - -
DISCLAIMER: The information in this video is the opinion of the speaker(s) only and is for informational purposes only. You should not construe it as investment advice, tax advice, or legal advice, and it does not represent any entity's opinion but those of the speaker(s). For investment or legal advice, please seek a duly licensed professional.