Rob, Hello, I'm Rob Hirschfeld, CEO and co founder of RackN and your host for the cloud 2030 podcast. In this episode, we revisit where we are with crypto, or more specifically, where we are with distributed ledger technology or DLT, and there are some really fascinating, pertinent, real examples of places where the core technology behind crypto is thriving, doing great, making a big difference. And crypto also, at least has regained its value, if not, it's allure, and we will discuss the human impacts of that and what went wrong. Also, I know you'll find this conversation exciting. We really do talk through the tech and the human parts, as we always like to do. Enjoy you.
We've got our quorum into the last fad that everybody was trying to embed into their messaging, which is our topic for today, which is crypto and where, where the hype and the excitement for a technology ran ahead of the applications. Although I think the challenges with crypto were not crypto and the mix into currencies, and then the fraud, there's no other word for it, that was perpetrated. On top of that, that enthusiasm really, really caused the technology to become a challenge. Challenged. I'm parsing my words very carefully, but I I'm interested to know if, if y'all think there's a crypto comeback, if blockchain can break apart from crypto and be interesting if we're starting to get smart about their actual applications.
Oh, I can. I'm sorry. Go ahead. Go ahead, Rich. What
I was going to say is, I keep track of DLT distributed ledger technology, except when it has to do with cryptocurrencies and in point of fact, there is a lot being done with Blockchain underpinning smart contracts. Contracts, smart contracts, maybe you see a couple of indicators, and that's the fact that all of the big consulting firms are continuing to look for and hire people with serious blockchain chops. Okay? The World Bank is continuing to add a lot of applications based on DLT to basically kind of lubricate and make a good deal more efficient, less prone to the kinds of risks that often happen in bank transfers within global, kind of global, global money transfers and settlement, for example, right? And they're expanding a lot. There are a number of companies that are spending time look, once again, looking at smart contracts. Tokenization is another area where people are making use of blockchain as an underpinning, and I think there's something to be said there for the way in which folks are making use of that, including its incorporation into some of the larger digital twin projects that are multi vendor, multi organizational. I'll stop there, ecosystem
like.
I see a lot of the similar patterns, and on my end, yes, so that the the major activity that I see currently happening on the crypto side is not so much about implementing new blockchains, but about implementing things like safe and secure swaps, like atomic swaps. So that that's on, on, on the crypto side, on their blockchain, inside in particular, yes, so like global trades, like not just currency, but also securities, that's still going strong in particular, because the big advantage, one of the big advantages that you haven't mentioned yet is that transactions via DLT, but whether it's blockchain or something equivalent, They are transparent. Yeah. So having a public ledger where you can say, like, Okay, this is the petrogy of the Securities and trade trading that raises a lot of confidence in the buyer. And this is why, again, like the large financial organizations are are examining that. Yeah, there's also another field where blockchain has been used for quite a while, and it and it's still going really strong, but it doesn't receive a lot of a lot of public mention, and that is supply chain, because again, like, it makes them public, it makes them transparent, makes them efficient. And, like, really, like Amazon, like, you can use the Amazon's managed blockchain for this so that you have off the shelf solutions for it already. Right
now, the public leisure piece makes a ton of sense for for multi party transactions that you want. Actually, they don't all have to be publicly transparent. They can just be, you know, within the parties, yeah,
they have it make them traceable like you, unless you know what the token represents, it may mean nothing to you, but when you have the the history of the of the token, then you have the history of the supply chain.
Yeah, Joanne, you've got, you've got the lineage of it. You might not have all of the kind of external factors. You might not have all of the all of the information you'd want for a pedigree, but it is necessary. It's not, let's put it this way, it's necessary information that you wanted to be able to depend on, when you're thinking about the pedigree, it's not completely sufficient, yeah, but that's, that's what leaves open the whole notion of building applications on top of distributed ledger and when, in fact, pedigree is important when various kinds of transparency, and again, I think your point, Rob was doesn't have to be public and private. You know, private permissioned blockchains are still getting built.
Yeah, I have a question on it, but Joanne has her hand up, so I would defer.
Thanks. Three things, there's an article in data center frontier today about using AI to create composable infrastructure. There was also a post on LinkedIn earlier today that I'm trying to locate while you guys were talking, and as soon as I find it, I will send it about the use of platforming, just like we talked about for DevOps or internal development teams, something for operators around that as well. There's a suggestion of AI replacing IAC, where infrastructure as code is actually driven by AI and. I'm not saying I agree with it or disagree with it. I'm just telling you, in the last three days, there's been like a slurry of information on the tokenization at the protocol point, meaning the DLTs, because I follow it the way rich does, not as crypto, but as DLT and distributed computing and whatever. There was a very good paper, and again, I'll have to go through ResearchGate and find it again, because I lost some of my bookmarks with the latest stupid browser update. But the article was basically, or research paper was basically talking about using distributed ledger technology to trap, track and trace the use of AI tokens in regenerative AI or other capabilities like that, so that you could actually trim down the number of tokens being used by from the individual user to the large corporation, through the cloud, through the edge, etc, where I'm seeing a lot of that you're saying you gave me you have a weird
I'm thinking tokens. When you're saying tokens, are you talking about tokens on the blockchain, or tokens from an AI, API
tokens,
API tokens, yes,
got it okay. There's a lot of different ways tokens are used here. So I was, I wanted to make sure I was hearing it right, right, yeah.
But like, Can you, can you take the tokens that are used for AI and reformat those, this is one aspect in such a way that they are tokens on a blockchain, right? That, to me, sort of makes a certain amount of sense. I'm not sure technically, how you would actually accomplish that. But that's, you know, my brain is not going in that direction. I can understand the rationale or the logic for doing it. So there's a use a business case and a use case to the other to what Klaus was saying with regard to supply chain. I'm seeing a lot of companies that are trying to invoke two things at the same time on the under the guise of we want to do a small language model for our industry, or our segment of an industry. We want to also, at the same time, create the capability of passing data back and forth along a protocol versus along a DLT, right? So we can start using smart contracts more effectively. So we can start using AI to write those smart contracts.
I mean, why? Why? Why do you use, as opposed to just a online database or something like, like, what's the DLT adding, versus just, you know, a place to store records,
ability, tamper, tamper evidence, right? I mean, basically it solves, it solves a lot of the regulatory we especially in blah, in, in supply chain, when you're when you're crossing any kind of jurisdictional border,
right, moving from, you know, moving goods from one country to another.
So, so let me, let me take a I'll fabricate an example a little bit. But you're dealing you're Boeing, you're dealing with suppliers through their for their for their airplanes, and they could have a database that says this supplier shipped us this, or this supplier ships that. But when the federal regulators show up to and to ask, you know how the doors were, were there? Your the ability to have those things in a in a DLT, means that the regulatory you've actually now have some more verifiable conformance, because you can never it's very hard to prove that somebody didn't erase something from a database. So it's, it's,
it reduces the cost of audit, both on the part of all the participants and the part of the audit.
So the interesting challenge with all this that I consistently forget with DLP is that most, most enterprises actually can't trust or don't want to trust their own data sources. Here they have, they have motivation to have a an externally auditable or externally verifiable proof point for their internal or even for internal record keeping. Okay,
in Mark, in multi sided. Mark. It's one of the roles of a trading pit or a, you know, somebody that manages the transactions. Is exactly that, if there is a if there's a conflict, if somebody says you didn't hit your SLA, or it didn't I my order was not completely fulfilled, or, you know, Pick anything you know, and there's an argument about it, then the question is, all right, how does it get resolved? Think of this as you know, the easy, the easy path to kind of having a trusted intermediary, you know, kind of a source of of of truth that says, you know, this is what got put on the this is what got put on the ledger. This is when it happened. This is what got what got done. And if there is still a problem with that? Let's go and find that the organization responsible for putting it on that ledger.
It gives you a data governance
precisely,
and again, like in many cases, with, at least with multi jurisdictional DLTs, the organization ends up being not a single entity, but a conglomerate, A conglomerate. So it ends up being done by consensus.
Precisely. It's a community that trusts quick places, realistically Places Trust in the in the authenticity of the data available to all of them. And that's that's a big deal that reduces any number of, you know, exchanges of faxes, rubber stamps, you know, if you've ever tried to to move anything across an international border, especially perishables,
stuff. Which is also why these implementations tend to gravitate towards proof of stake blockchains, because then your stake is essentially your membership in the consortium,
right? It's like having a seat on a trading pit. You know that those are, those are extremely valuable. Without that, you don't have the kind of access and you have to go through a the Edit somebody else.
We're not, we're not burning area, yeah, but we're not running data centers full more just to compute sums with proof of work at the same time, which, which
is a totally different
problem. But
the big, the big challenge here of using DLT is, you know, they are not, they're not the place to put, actually the data. It's, it's, it's, basically, it's a, it's a notarization that you know, the data that you will find at this location in the address space
has this checksum, or this, this, this signature has
has a check. Yeah, right, because using Blockchain as an actual database is got to be one of the most idiotic things that I've ever heard. And every time I hear somebody say that, it's kind of like, okay, you know, I'll write you off, because you'll never, you'll never get there. Yeah, I
guess that's a different because when I, when I think of a distributed ledger, I think of the ledger entries as being in the ledger. But what you're saying is, no, the ledger can be much more sophisticated than the distributed part of it. If so,
if you, if you you, if you put data into some sort of data space with whatever encryption security you want, as long as you're using the DLT to kind of say, This is what got put into it, and this is what I know. This is whom I know that actually took has to take certain responsibility for it, or accountability for placing it
on there. So I mean jumping all the way into fraud in Bitcoin, which I know is a jump, but I think it's a reasonable jump. Here, a lot of the challenges that I heard from a Bitcoin fraud perspective, where somebody got access to your wallet and you lost all your Bitcoin and, you know, they had all these scams and things like that. Some of that strikes me as lack of controls in a new, you know, in a new monetary market, right? You know, rather than a fault of
the underlying tech, it wasn't the other underlying technology. Was access to it in and out. I mean, basically it was, you know, it wasn't the bank. It was standing, you know, bunch of guys standing outside the door of the bank, you know, ready to to, you know, bug everybody that came in or came out. It's a it and, and then, in point of fact, trying to simplify it for the general public, made it a, you know, a perfect place for the artists to sit there and and just rip folks off and the exchanges had no particularly good reason other than jail to, you know, improve the situation. So it's, it's, it's the ingress and egress from these systems. That's really the place where there's a lot of there's there's too much opportunity for
it shows both the benefits and downsize of a libertarian system.
Yeah, and I think there's, there's places where settlement delays are that the financial system has built into it. Have some, there's, there's some value in settlement delay, potentially,
yeah, well, but you can, you can build that in, if you like, to
point, yeah, is it? You could actually, you know, have some type of system or trigger, which is what ends up right there? Actually, there's, this would be a whole conversation of systems in which what had been human delays that allowed some propagation motion are getting eaten up by technology. And so we're doing things faster than human than you know, we're removing buffers from things. But there's fat.
It's not just faster, it's it's uniform. And number one, you can count on the timing, and it's a whole boatload cheaper than having and, of course, more dependable. But it's it. You know, the point being you, you can count on transactions having a certain kind of of cadence.
I am not convinced about the cheaper side, but at the very least not at scale like blockchains have a problem with with long term scale. Like that's true, keeping that like the ever growing history of the chain, and
the fact that you have to distribute it, that ever growing history,
yes, yeah. The The other issue, which has not been ubiquitously solved with the blockchain either, which is particularly a big problem with with trading blockchains like crypto currencies and securities Trading ones is that there are few recovery mechanisms. If I say, for example, if I have a million dollars in Bitcoin and I lose my wallet, or, let's say, I die, yeah, that stuff is locked in that there, unless I specifically have a back of my wallet that I hand off to someone else that, yeah, that resource is currently locked away.
So unless you've made, unless you've made allowances for the fact that you're unavailable, you've you've figured out a way of, you know, with the dead man switch, whatever you know, at a certain point, somebody, somebody else, or someone, something else, gets access to those codes. Yeah,
and it. Interestingly enough, most of the way systems are structured today, there are intermediaries who can unlock, if you not, not in blood, not in crypto, but if in a financial service, right, or a need or trust or something like that. You know, we have mechanisms that have evolved to address, you know, absent owners and things or theft and right
on their laws around the state on the handling of it. Yeah.
So did
you want to add something? By the way, your hand is still up. I didn't know if that was a new raise or a
All right, okay, you're still muted.
These are, I mean, these are human problems versus on, on how to make systems function. The the tech, the tech, from that perspective, you know, isn't, isn't, isn't caught up yet would be sort of my thinking on it, but there's nothing inherent in the tech that allows you to do this right? There isn't a back door. There shouldn't be a back door to decrypt it.
No, the whole Well, the other thing is the mistake that so many people made about anonymity. You know, it's like, no, it's exact opposite. It is not anonymous, okay?
Well, kind of so. It is anonymous in the sense that your identity is not directly tied to your transaction, but it's not private Exactly.
I think that's that's a good, good distinction to make. Now,
I heard some really interesting reporting about people who were sleuthing these crypto transaction. You know, who is behind crypto transactions, and it's very traceable, but the people who are, you know, trying to obfuscate their identities are very good at bouncing through a whole bunch of accounts, I guess, which you should be able to just untie. But it's
harder, yeah, takes a lot of work. There are lots of different ways of doing it and
then laundering the stuff, like, like, using the like stolen assets to to interact to the as payment for some of the the edge users who may not have the checklist, okay, is this a is the source From a largely stolen set of of cryptos? Yeah. And in the end, the like, once there is a sufficient number of transfers between the legitimately acquired assets on the final one, at some point you can't say, like, No, I'm not going to accept something from this address because its supply chain came from for a question of source, because at that point you have to block 99% of the addresses on the blockchain.
Oh, goodness, right? And well, and then I we have a malicious actors challenge on the internet in general, fueled by some of this stuff. But you do you know is that it's like standing cockroaches. You could identify a bad wallet, and there's easy enough to create that there's no that that doesn't really protect you, protect your resources for your transactions. Yeah, I would. I mean, none of this strikes me that the general populace is ready to Actually Trade in in crypto, especially not the ones they have that are proof of work, that are very expensive, ultimately transactionally.
And that's, that's one of the other things you know, the anything that's still making use of proof of work is an expensive ledger. Great. I mean, you, you running a ledger there, and you're the the amount of the amount of power that goes into, you know, all of that, all of the proof of work stuff is just you. Um, enormous and it's wasteful. It's which
is a lot, it's pretty sad, which parallels what's happening with AI as well. Yeah, the power consumption is on. AI is also enormous and wasteful, yeah,
and very difficult for people to even get their head around when they're doing one on one actions, right,
right? Well, this going back to what Joanne was talking about earlier. One of the reasons you want to track
AI API calls is each one of those API calls is has a cost as a price and and, you know, if for no other reason than reducing the number of API calls in order to, you Know, reduce the cost of running these, some of these applications, is, it's, it's something that you would absolutely want to try and make happen, try to do. And the fact of the matter is, you know, right now, AI based
requests, searches, by most accounts, are running about a full order of magnitude more costly than a straight on Google search or Bing search.
Ah, okay. Oh, yeah. So,
I mean, and the knock on effects are great for public utilities, because their stock price has gone up. If you look at all of these, you know, take a look at what's happened this year to the public utility stocks. You know, sleepy as can be right in, you know, this is the place where, you know, Mom and Pop put their put their money in ETS. They've had growth this year that is faster than Nvidia and most of the others. Because, yeah, take a look at their take a look at the stock prices of Duke and some of these other speaking of your T shirt there,
it's 32% year over year growth
per utility. Yeah, that's as crazy, but
it's a combo.
It's a combo of things, right? What is other things? Yeah,
go ahead. TV chargers, right, right? One is public like, like on the highways and things like that, where they are putting them in. So it's private homes. Walmart in certain states is actually putting EV chargers in parking lots. And the other part is data centers from from small colos all the way to the biggest ones it that's the other the biggest percentage of the growth, and it's projected to grow rich to 43% by 2025
Exactly. So why is that making a regulated industry the stock market darling? Well, it's because when they're building all of this additional power generation, they're buying a lot of goods. They're buying a lot of stuff. It changes their it changes their tax burden, and it changes their their profitability, because, you know, they have a they have a cap on what they can charge, right? But if they're, if they're getting the kind of demand and building like crazy because, you know, meta has open, it wants to open up yet another, you know, AI data center there. I mean, it's,
it's a captive it's a captive customer, yeah, it's
a captive cuff customer, and their their profitability. Well, just the, just the, the the's a small
little factoid that will drive this point home. Look at the cost of a roll of wire, raw mix that you would use for, you know, home, lighting, building, lighting, whatever, yeah, it went from $69 a roll of, I forget, the meter, 1000
feet, yeah,
to $189 a roll. If. Can get it, and it's in short supply, so that people now are going into homes that may be torn down, not, not that they're condemned. They're not they're just, you know, like someone dies in the houses being sold and whatever they're going in, and they're ripping out all of the wiring and the copper piping, because the price of copper per pound is that great. So everything that goes in, in terms of what Rich is talking about, materials, aluminum, steel, copper, PVC, all of the plastics, anything around utilities, for data centering or for just power generation at all. Short supply, double digit pricing increases, even in some cases, triple digits.
So interesting, it's the the sudden increase in demand is what's absolutely driving this. And Joanne is completely correct. Both, you know, EVs and data centers seem to be among the biggest ones. So, you know, I realize we've gotten a little bit off topic here, but it's the the point being, this can be crypto, and well, the underlying infrastructure for crypto, Blockchain, DLT, can be a very expensive proposition. And this goes back to Klaus's point you know about, about costs and and and price. And I would, even with that, I would still claim that if you have total cost of of operation in things like logistics, supply chain, and you introduce the cost of delay. You introduce the cost of kind of contingency plans that you have to put in place in the event that a transaction does not take place on the on the exact calendar date that it needs to. When you get rid of that kind of uniformity of transaction, I would say you're still doing a better. You're still doing better by way of price and cost with DLT, but it's changed who's getting the who's getting paid for what.
But you know, I would argue to that point, if you take companies that have pre configured servers right like Mark's company, where you can you're literally renting a piece of equipment in a data center wherever you want. It's all pre configured for you based on whatever protocol you want to support. If it's Bitcoin or Ethereum or Solana or Dogecoin or whatever the heck it is. Doesn't matter. It's all pre configured. So you can literally one click and have a validators set up available to you, right? If you look at that, and you look at the model for the infrastructure, you can actually come up with a sort of an argument for using DLT or or maybe the argument is against using DLT and going to cloud as an architecture versus using the distributed architectures from a cost perspective.
Okay, so, yeah, okay, you're all you're saying, is that there? Well, let me, let me play that back. What you're saying is that there are, there are situations in which you can misuse on the cost basis. TLT, yes, that in point of fact, you can have a more centralized, also secure, secure, reliable, all these other good things in a centralized, centralized data organization or cloud based data store, and as long as you can
convince all of the participants that it's as tamper resist. Isn't tamper evident and authenticable, then you're, you're doing just fine. And there are some, there are exactly some technologies that do that come pretty damn close, at any rate. So, yeah, okay,
I think we're look well, my point was from that cost control perspective, I think we're about to see a new hybrid emerge, which is the combo of I've got DLT and edge for much I've repatriated some stuff from public cloud to private cloud, and that's my central store that you're just discussing. And overall, would you use?
Would you use private cloud for what we're now looking at as permission block change?
Yes,
what you're saying? Okay,
yeah, me and my five biggest customers and my 10 biggest suppliers. That's my private cloud, and I can run it at lower cost than putting an instance up on AWS or whatever. Tyler actually had one of his slides that he sent me, and I'll get his permission and send it to you guys, because what he was looking at was the numbers of services that are duplicated across the public clouds, right? And the cost of each and trying to make the argument that a private cloud was not only a lot cheaper, but a lot more manageable, etc, he's got a slide on that he was looking at the cost of the cloud. We
see it. The thing that makes me sad in that statement is that there's a lost set of skills as a lot of software companies became SaaS companies and chose to run a service rather than operate as software. And so I think that's true, but I also think that there's a skills gap in the market where services that people used to be able to basically buy software and run themselves are have have become SaaS outsourced infrastructure, and it's going to take some market correction. I think there's personally, from rackns perspective, right? This is where opportunity is. There's opportunity for companies to be like, Oh, we can run our own infrastructure. It's cost effective. We can manage the services that we need and only the services that we need. We don't have to, you know, maintain, you know, pay for Amazon to maintain services that we don't think are necessary. Or, you know, part of that footprint, right? There's actually a lot of overhead in those business models.
There is a lot of overhead. And the other thing is, as well, is you're removing some of the vectors in the cyber landscape that could be considered threat vectors, you're able to control them more closely, and so that's removing cost from having to have additional tools to manage the clouds that are public
to the extent that we start having some more standardized and repeatable processes for this. The challenge that we see from an from way off topic enterprise is that there's, there's no there's no commonality in practice, and so they aren't better, they currently, without some work or not better off than cloud. At least cloud requires some more consistent governance. Yeah, then, and we It depends it's a pendulum, it'll swing back the other way, looking at wrapping up on DLT, and I just got ping that I do need to wrap up on time. I really appreciate the check in on DLT, because I think that these are really important technologies for helping companies share and distribute work. And I do constantly forget the idea that even if you have your own data sources, the audit, external audit of those sources, is real. The need to have, you know, some trusts and capabilities. So I think there are some really powerful use cases in this, in these models that unfortunately got conflated with Sam Beckman freed, crypto theft and, yeah,
exactly, and everything else it there are two things. One is there's value internal to an organization, as you just mentioned. But when you're talking about communities of interest, closed communities or open
anytime you have a community of interest, you know a bunch of trading partners, what have? You in a, in a in an ecosystem around a supply chain, for example. Again, same, same argument.
I don't think you even have to go that far. I mean, we talk about software bill of materials all the time, and I think that if you had built a software bill of materials and then, you know, assigned it and put that in a trusted third party, right? If you're talking about an attack vector, one of the things that you need to do, just to for your own confidence, is to be able to say, you know, if I got attacked and don't trust my my own systems, I need to be able to, you know, validate that. So the I think these are very real, yeah, very real use cases. It'll be nice as we continue to detangle them. And I have incredibly mixed feelings on the idea that the Sam bankman freed. You know that they're actually I'm glad they're able to make the people whole that he ripped off because Bitcoin research, you know, Reese, you know, the resurgence of some of these cryptos, but at the same time, it just fuels additional feeding frenzies. And I think that's problematic.
Yeah, a different a different issue,
totally. I think they call that a cultural issue is
it is ultimately, and I think that's one of the things about all the crypto pieces, is that that was all the hallmarks of a scam, or people taking advantage of market and lack of knowledge and market, but just huckster. I hucksterman Freeze perspective, literally just, you know, stealing other people's money because they were flowing through his system. And it
was, yeah, it was available. It was a it was available. And
nobody was going to notice, because it was always going up, up, up until,
until it wasn't,
till it wasn't right the mountain box and all the other big scams that have come off of it. Nfts
just have to keep relearning the same lessons
history sounds too good to be true. Probably isn't
challenging the ability of the population
there you have it
is that if you catch it in the right window and you enter and you exit at the right time, you can make out like a Winklevoss,
oh, let's talk about moving shots. And, yeah,
no, I'm gonna go just doing things the hard way, in honesty. Thank you all for the conversation. Let me next week. We are we are on next week. And the topic that I have is, oh whitespace left by large tech acquisitions.
That'll be a fun one.
Oh the white space in Eco, in, in, yeah, yeah, good. Good topic.
Speaking of left tech acquisitions, just before, forget. The news happened today. Compushare has announced that they're dropping VMware for Nutanix. These are 1000s, 10s of 1000s of systems of licenses that they're switching over. Oh, and they're saying that they're going, like, even without the VMware license increase, they will come ahead monetarily. Now they're they have, like, a 10 to 15x saving this is
going to be the gift that keeps giving to everybody else. It's going to be amazing, except for the customers,
yeah. All right,
that is for next week. I appreciate the heads up. That's cool. All right, everybody have a good one. Yes, bye, bye, bye.
What a fun conversation. This is really about, you know, core technology that can be used in really positive ways, and yet in hype cycles, get so easily pulled into places where false promises are made. Hype wins out over objective thought, and people race ahead before controls all things we discussed in this in this conversation, and really fascinating to think about, I know we'll keep coming back to this, the court. Tech of DLTs, and I hope you'll be part of those conversations. You're always welcome to come in, join us, be part of the round table conversation. You can find out more at the 2030 cloud, or just join us on a Thursday morning, we would love to hear your opinions. Thank you for listening to the cloud 2030 podcast. It is sponsored by RackN, where we are really working to build a community of people who are using and thinking about infrastructure differently, because that's what RackN does. We write software that helps put operators back in control of distributed infrastructure, really thinking about how things should be run, and building software that makes that possible. If this is interesting to you, please try out the software. We would love to get your opinion and hear how you think this could transform infrastructure more broadly, or just keep enjoying the podcast and coming to the discussions and laying out your thoughts and how you see the future unfolding. It's all part of building a better infrastructure, operations community. Thank you. Applause.
