Full Transcript

Dhruv Sharma: Hey there listeners, we're streaming live. I just realized it's episode 72 already, which means number 75 is coming up and if you want us to bring someone special on, please let us know. Today we're going to dive straight into our guest segment. We have Siddharth Shetty, who's the founder of Finternet Labs and who's someone who's actually worked in the India stack and a whole bunch of other digital public infrastructure projects. Siddharth, welcome to the show. How are you doing?

Siddharth Shetty (Finternet Lab): Thank you so much. It's great to be here with Savin Yadav, appreciate you bringing me on.

Utsav Somani: Welcome to the show and let's get started. So do you want to spend one minute talking about what Finternet Labs does now?

Siddharth Shetty (Finternet Lab): Sure, sounds good. So the real origin of Finternet came up from the work I've done over the last 10 years in designing and building digital public infrastructure, which has rolled out to hundreds of millions of people. And we felt that fundamentally the way financial infrastructure is architected today, it's sort of like every time you build a new car, you build a new road. And that creates all the fragmentation, all the complexities, both when you transact within a country or when you transact across countries. And so when we were looking at the newer developments, cryptographic developments around digital signatures, wallets, public blockchains, agentic flows, it allowed us to reimagine a world where individuals, businesses and agents can truly have control over their one digital identity and credentials. So credentials are broader proofs of not just who you are, but where did you study, where did you work, so on and so forth. The second is your assets and assets across the entire spectrum could be regulated assets like money and securities, registered assets like physical property and vehicles, attested assets like a solar rooftop or gold, silver and other commodities or user controlled digital assets. And then the third category that you'd have control over is your own catalog. And so businesses can allow for themselves to be searched, discovered, contracted against and negotiated by other agents. And so what we're really building within the Finternet, which is part of a larger foundation structure called Networks for Humanity, is infrastructure that brings your ID credentials, assets and catalogs together and gives you back that control.

Utsav Somani: Awesome. So, I mean, just for some backstory and for our listeners to know about your motivations a little bit more, you helped build the digital plumbing for this India digital revolution, right? Aadhaar, Ayushman Bharat, Digi Yatra, Account Aggregator, all of these were projects that you contributed to. Why choose infrastructure over individual products? Like, what are the non-obvious reasons? There are some obvious ones, but there must be some non-obvious reasons and motivations as well, right?

Siddharth Shetty (Finternet Lab): Yeah, absolutely. So I think one is when you really work backwards from the problem and let's say you're solving for financial inclusion, which means you want to democratize access to the financial system to a billion and a half people, then immediately you start to realize that, OK, could I build an app for that? You could, but would a billion and a half people use that app? The answer is pretty unlikely because you've only got 700 million people, 700, 800 million people that have smartphones. So you already got about 40 percent that don't even have smartphones. So what do they use? Even in the ones that do have a smartphone, you ask the question saying, will everyone use the same app? Unlikely because it's just so diverse if you think about it in the construct of India and people have different preferences. And if I'm an individual, I'm a business. Again, I have different preferences. And so as you start to unpack that, you realize that you can't just build one app that satiates the needs of the entire country. You need to start to build an infrastructure that allows for entrepreneurs to build much more personalized or targeted solutions. And so we end up, we are plumbers. We get excited by a lot of the underlying plumbing. And so our main focus is building the underlying roads, the underlying highways so that people can build their cars and trucks and other personalized colors and so on on top of that.

Dhruv Sharma: So that's with what you're doing right now is the ambition, is it Indian or domestic? Is it global?

Siddharth Shetty (Finternet Lab): Yeah. So Fintanet's a global initiative. It was actually born by a paper that was co-authored by Nandan Nilakirni, who had co-founded Infosys, Augustin Carstens, who till recently was the general manager of the Bank for International Settlements. Prior to that was the finance minister and governor in the Bank of Mexico. And so we released this in D.C. about two years back. And since then, we've spent a lot of time with both public authorities and private sector players like asset managers, commercial banks and others around the world. So it's a global initiative. We've set up a network of labs across Singapore, India. I'm based in Singapore, Switzerland and the U.S. And we're working with different countries plus different private sector actors to bring it to life.

Dhruv Sharma: And whenever things start with a paper, so I was just going to say whenever things start with a paper, it's generally a good sign. Bitcoin started with a paper, you know, Transformers, the whole architecture started with a paper. So and later, maybe we're going to ask Siddharth a little bit about that paper itself.

Utsav Somani: But what enabled this now? Like why was the time for this now? Like we've had a guest who've spoken about tokenizing real world assets before on the show as well. So maybe the technology evolution that has enabled this now and why is this coming to light now and how is this even possible now?

Siddharth Shetty (Finternet Lab): Yeah, that is a good question. So if you look at it today across the spectrum of assets that I mentioned to you, there are really two properties that are most important for financial flows. One is the property of verifiability, that is anyone else. So let's say I own a piece of land, anyone else should be able to verify I'm indeed the owner or I'm renting it, so on and so forth, depending on the use case. But let's start with let's stick with ownership for a minute. So anyone should have the ability to verify ownership around the world. So that's one. And so the work that's now happening traditionally, how would you solve for verifiability? You would try to create some set of APIs, expect people to integrate with you. That doesn't become universal, meaning it will only be a subset of actors around the world that can verify ownership because people permission their APIs. Imagine a local registrar in India trying to open it up to anyone in the world, someone in, let's say, California who's a lender trying to verify it's not going to work. And so that's where the work that's happened on digital signatures and wallets. So in the context of India, if you look at DigiLocker is fundamentally an example of that. It's digitally signed documents issued by public authorities at a federal or state or even local level. And now you've got over 9 billion documents, 500 million users. So one part, sorry, one part is the developments on verifiability and the online technologies that have unlocked that. And so you're going to see much more documents essentially being digitally signed and issued directly into a user's wallet, individual business or agents. And then anyone can verify if I'm the owner, if I'm the holder or broader provenance as well. So that's one. The second part pertaining to these assets is transactability, which is once you verify I'm the owner, let's say I want to transfer it to you. Let's take that simple workflow. I want to transfer ownership. That's where the ledger component comes in, because how do you ensure I've truly transferred ownership to you and it is yours? How do you have that normality? And traditionally, assets have been represented either in private ledgers. So if you think of the banking system, each bank runs its own ledger. Or if you look at the capital markets, it's represented in shared ledgers. These are depository institutions. So each asset manager doesn't run their own ledger. You have two depositories in the case of India that run a common ledger, shared ledger. Now, because of the work that's happened on public blockchains, you can imagine a world where assets are issued into a user's ledger. So it's evolving from private ledgers to shared ledgers to eventually the user's ledger. And that's the key shift, because think of it if you go back in time when assets were represented in paper. You and I meet, let's say we meet in some corner of the world. I have a physical token, like the US dollar bill. I can give it to you and you can look at that token and it is self-verifying. You can look at the watermarks and all of that. It's self-describing. It tells you it's $10 and I give it to you. Maybe I give it to you at 2 in the morning and it's yours. There's no idea that the computer is asleep. So you have a lot of portability when assets existed within that physical medium. Of course, you are constrained by the laws of physics because I can't tear up $10 into a dollar, but by and large, it had much more portability. When we digitized these assets, you lost a lot of that portability because now for that same $10 transaction, imagine the complexity that we created because I had to integrate with your correspondent bank and so on and so forth. And there was a whole complex chain that got set up. And then when you look at US dollar stable coins, anyone with an internet wallet can get access to it because it's part of a common grid. And so we're moving back, really a mental model I like to look at it from is you're moving back to a world just like you have those physical tokens, but you now have these natively digital representations. And so those are the underlying technologies that are enabling us to imagine this future. And then, of course, you're supported by a lot of legal developments that are taking place around the world to now formally create the trust infrastructure for tokenization to happen across different asset classes. It's still going to be a 10-year journey, but a lot of progress, particularly the last two years has already taken place.

Dhruv Sharma: Dharth, I think this is great. You spoke about a lot of concepts. How about we wrap them in a real-world use case? I believe you guys do something around agri-loans. Do you think that'll be a helpful example?

Siddharth Shetty (Finternet Lab): Yeah, that's a good one. Let's build on that.

Dhruv Sharma: So maybe we'll lay down a few questions that we have for you. One is, you spoke about ownership and verifiability, et cetera. And so once you have a real-world asset that turns into a token, how do you map all of the ownership attributes of the real-world assets into the token, bring all of that information on the ledger, and see if that asset is going to be transacted across land borders between two different jurisdictions? It's unlikely they're even sharing a common ledger. Does your architecture allow people who have different ledgers to communicate using a common language?

Siddharth Shetty (Finternet Lab): Yeah, a hundred percent. So we do believe in a world where there will be multiple ledgers. And so you really need to put in a lot of work when it comes to interoperability. And interoperability has both legal characteristics associated with it. So how do you contractually make these business transactions interoperable? And then there's also actual technical protocols to make them interoperable. But if I go back to the use case you mentioned and use that to just illustrate how this works, let's take agri-loans, let's take care of farmers. You're a small landholding farmer. And in a country like India, over 50%, almost 50% of small landholding farmers don't even have ownership of the land. So they're called tenant farmers. Now, as a lender, imagine I'm a bank or a regulated NBFC, a non-banking financial corporation, what happens? I need some amount of, in the traditional method, I need some amount of documentation to prove ownership, to prove that this is a legitimate borrower, to prove that they'll use it for the intended purpose and so on, which is very unlikely that these farmers will be able to share any form of documentation because of the asymmetry between them and the landowner. So now what we've launched, vouch.finance, if you take a look at the website, it's the first product on the Finternet, essentially allows for credit-based vouchers to be created. What does that mean? You can now create a programmable instrument as a lender that says, I'm attaching it to this farmer. So you're connecting, how do you know that they're a farmer? It's a combination of your, in the case of India, your Aadhaar ID, which establishes who they are. And then many states have farmer's IDs that are issued into their DigiLocker. So the combination of these two establishes the base identity linkage. And then I issue you a programmable credit voucher that says that I'm giving you a credit for, say, a lakh, and you can purchase a set of agricultural inputs against that. So the farmer cannot take that lakh and spend it anywhere else. They can only spend it for, say, pesticides or fertilizers or seeds or other related inputs. And so there's a live program that will soon actually be using this infrastructure. And that's just one example. Because at the end of the day, these tokens are natively programmable. So you can create different kinds of use cases or purposes. But this is just a simple one where any financial institution will create a credit-based voucher, give it to farmers, and that changes the nature of risk because they don't have all the documentation. Or even if they do, these are much more smaller tickets. So lenders don't have the unit economics to go out there, check in person. But now they can give you a voucher. And they are much safer because you can only use it for authorized input purchases.

Dhruv Sharma: And you're saying even if it's my first time as a farmer walking in with, I don't know, a sack of rice, and I have no prior relationship with you, I can walk out with a credit voucher. No questions asked.

Siddharth Shetty (Finternet Lab): Exactly. Exactly. That's what will go live later this year.

Dhruv Sharma: Yeah. That's pretty exciting. What would the application layer look like for the farmer?

Utsav Somani: And also, yeah, exactly. Just adding on to that. What else is Finternet enabling now, which was not possible before?

Siddharth Shetty (Finternet Lab): Yeah. So I'll touch on the farmer situation and then we can extrapolate further. Is from a farmer standpoint, the vouchers could be given both in an offline mode or in an online mode. In an online mode, that could be on any, because the network is interoperable, it could be on any application that they already use. So it could be existing UPI apps. It could be existing farmer apps. So in many sort of rural areas, there are much more targeted apps that farmers use. It could be on any of those that hold the underlying voucher. So the experience is extremely simple for a farmer. They don't have to go out, download a whole new application. And then it's not just the farmer side. It's also what you shared, which is I'm walking into some shop which is selling me those inputs like say pesticides or fertilizers. They could be live on their existing UPI merchant apps. They don't have to be on boarded all over again. And that's where the interoperability really comes in. And then you can use the same rails for, let's say tomorrow you have flooding in a particular area and a set of farmers are affected. You can now hand out disaster based vouchers where they only get activated if the disaster strikes. So you can upfront, there's no infrastructure in the world that allows for that today. I can upfront issue vouchers. So people are mentally at peace when a disaster is about to strike. And when it does strike, you can activate it. And the issuance is not just from the government because it's an open loop system. You can issue it from other private actors. So it could be CSR arms of companies. It could be individuals themselves. So there's an effort now where individuals can create vaccine vouchers for, let's say, people that you work with at home and so on and so forth. So that's some of the other use cases. But here's broadly what the experience would look like from a farmer. And we'll serve back to your question. I think fundamentally the unlock where assets are being represented on a user's ledger rather than just on the institution's ledger is what's allowing us to imagine a lot of these use cases. And a simple mental model to think of it is just like the internet where everything is a packet. But then you create layers on top so that when we're having this Zoom call, in that situation, the packets are represented in this multimedia form versus when you do a UPI transaction, the same packets represent money. So while everything will be existing in a token form, of course, how that token is represented will also come with additional rules, who can issue it, who can transact with it, what are the limits. And so there needs to be a broader classification. We are seeing that set in place within the Indian ecosystem, but there needs to be a lot more education around how these different tokens can also be managed and classified, because based on that, their governance will vary.

Utsav Somani: And I mean, zooming out a little bit, all the DPI initiatives that have scaled beautifully in this country have had institutional champions behind it, right? And they needed some sort of regulatory buy-in from all the players involved and the regulators involved. What are the regulators that are supporting you in this journey?

Siddharth Shetty (Finternet Lab): So it depends on the type of asset, again. And so in some cases, yes, if you're talking about tokenization of money, tokenization of securities, then you do need the banking regulator like the RBI, you do need the capital markets regulator like SEBI to come on board. And many of them have made public announcements about what types of tokenization products they're looking to offer. E-Rupee is one such example within the banking system. However, I could also extend tokenization to non-regulated assets. So for example, one of the pilots that is being worked on within the lab is tokenization of agricultural produce through warehouses. And so that's a completely different supply chain, does not have a regulatory requirement upfront. You do have regulated warehouses or accredited warehouses. But from a farmer standpoint, I can go put my produce in an accredited warehouse, get a token, and that token can then be shared with a lender or it could be shared with an insurer or other financial institutions. And you can do that flow in a permissible manner today. So it depends on what type of asset we are tokenizing. And based on that, you will need regulatory buy-in, but all assets don't require regulatory buy-in.

Dhruv Sharma: So let's say if that tokenized asset was to be sold in exchange for tokenized money from elsewhere in the world, is that possible today or will that be possible only a little bit in the future?

Siddharth Shetty (Finternet Lab): So it will be possible a little bit into the future. I think what you do need is compliant infrastructure, particularly in a country like India, where you have capital control rules that exist. It's important that the KYC, AML, capital control rules can be embedded, particularly on the outflow or when you have an inflow into the country from another country, you can tag it with the identity so that the funds can be repatriated. And so as long as infrastructure that allows for this gets built out, I don't fundamentally see an issue. We are a bit at the early days of it, but I'll give you a simple example. You've got about 30 million, 30 to 40 million NRIs outside. They have no way to really access India's capital markets. It's often paper-based transactions that they want to invest. However, if even say 10% of them, 3 million of them could allocate $10,000 to investing in India, which is a very conservative estimate, they would be able to do that. You're talking about $30 to $40 billion coming into the economy, right? And so there is an urgent economic imperative for us to actually build out infrastructure that implements these compliances of tagging it with identity, embedding capital control rules, et cetera, so that the LRS limits and all of those norms can be followed. But that's pretty easy to do. Those are nothing but simple configurations on the money flow. And now with a lot of this, it's much easier to track. You can have native observability. So I think over the next two years, you should see a lot of this fall in place.

Utsav Somani: Dhruv, any final closing question for Siddharth?

Dhruv Sharma: Yes. Siddharth, I think in your opening comments, you spoke about individuals, businesses, and agents. And so agents are all the rage. And so we have to ask that the agent question, when the world goes from the way we currently transact to machine to machine transactions, how would they work on the back of infrastructure that you guys are building at Fintanet Labs?

Siddharth Shetty (Finternet Lab): Yeah. So two things. One, we actually built out a protocol for agents almost now, I think it's four or five years back. If you go to the beckonprotocol.io, my co-founder, Sujit, was the primary architect of that, along with my other co-founder from Obama. And the two of them built out this simple protocol that allowed you to search, discover, negotiate, contract, and have dispute resolution between any two nodes. And the protocol basically made your entire catalog machine readable. And this is what got implemented. If some of you are familiar with Namma Yatri, which is in the mobility sector, where people can do peer-to-peer ride discovery and booking, or ONDC, which is the Open Network for Digital Commerce, and now more recently, the India Energy Stack, where people can... It basically makes the copper cables programmable so people can buy and sell energy packets peer-to-peer using the existing grid infrastructure. So the generalized protocol actually already existed a couple of years back. What we're now doing within the FinConnect is... And this is sort of the goal I've set with my team, is it's a safe assumption to make that agents will be the primary customer. And they're going to be pretty ruthless, because you cannot take an agent out to play golf and get it to buy your product and service. They're not going to have a lot of those biases. And so fundamentally, it pushes you to actually have infrastructure that's the fastest, cheapest, lowest cost, because at the end of the day, an agent can... Like, let's say... I'll give you a simple example. I moved from India to Singapore. I needed to transfer 1000 Sing dollars back to India. How would I do traditionally as a human? I'm going to open three, four apps, or typically two apps between my bank app and say, why is... And then try to compare rates and make a decision. When I have an agent, the agent's actually going to go out there, search, discover, find the best liquidity providers. And that will vary from transferring 1000 Sing versus, say, a million Sing, and can actually piece the most optimized route together and give me back the best offers. And then depending on what time of the transaction needs to be fulfilled, I can go ahead, agree to that. And so it's safe to assume that now I'll have the lowest rate or lowest rate at a particular point in time versus earlier on me being locked into one app, unless I invested the effort in going to 10 different apps and then trying to manually put that together. So I think that's just a simple example of how consumer experiences will change for the better for users, but it requires a fundamental rethink in how people build platforms and products and assume certain types of lock-in today.

Utsav Somani: And when you were building all of these, I mean, large ecosystem infrastructure plays, I mean, you built the consent architecture as well, and you got a billion people to buy in as well. What happens in the future? Will people... I mean, Dhruva and I were having this discussion before the show started about giving cloud access to certain folders and stuff. Do you think convenience will always win over control when it comes to scale and adoption, or will there be a little bit more restriction, at least until the earlier adopters wet all of this out?

Siddharth Shetty (Finternet Lab): Yeah. So I feel it's safe to say that eventually convenience, if it serves a user's self-interest, will win over. And that's the right thing. If people are able to use solutions that serve their immediate self-interest or some future selfinterest goals that they have, they should be able to go ahead and adopt it. Now, in that journey, however, you will have new technologies, much like we saw the evolution of the internet, much like you see the evolution in crypto today, where you start adding more safeguards, because there's fundamentally a difference when early adopters use it and adopters are much more aware of the risks, versus you then have 100 million people or a billion people or 5 billion people using it. And then that's where the risks get magnified, or that's where you can lose trust and then people can revert back to a worse state than it was before, particularly when it comes to matters like financial assets and other non-financial assets as well, where trust is even more critical. And so I feel like what will happen is you'll have the early adopters migrating to it for convenience plus just curiosity. But over a period of time, you will have new entrepreneurs that build out a lot of the control mechanisms, safeguards, security, privacy, and that starts to mature the technology as a broader population comes on board.

Utsav Somani: All right, Siddharth. Thank you so much for coming on the show. All the best in your journey ahead.

Siddharth Shetty (Finternet Lab): Awesome. Thanks a lot, Atif. Thank you, Dhruv.

Utsav Somani: Have a great rest of the day. Bye. All right, Dhruv, it's you and me. Let's cover the news quickly. We don't have too much of it, but there are some big numbers which changed hands. IPL, 3.4 billion week, tech, money, Blackstone, old. Are the teams selling? The teams are selling. RCB and the Bangalore team and the Rajasthan team, 1.78 billion for RCB and RR, which is Rajasthan Royals, 1.63 billion, both closed within 24 hours, 3.4 billion combined. RCB was bought by Satyen Gajwani, who's representing the Times of India group. There's Aditya Birla group represented by Aryaman Birla. Some next-gen coming into play as well. And then we've got Blackstone, which has experience in, of course, owning different kinds of assets. And there's David Blitzer, who owns 76 as well. And there's a name that might ring a bell when you talk about Rajasthan Royals, the Walmart family, and then, of course, the Ham family, which owns the Detroit Lions. And there's a tech entrepreneur who's led all of this, Kal Somani. He was the founder of InterEdge and TrueView.

Dhruv Sharma: I'm sorry, what's his name again?

Utsav Somani: Kal Somani. I have no relation to him. I wish I did. We would be having him on the show to talk about all of these things. There are some big multiples, though. I think the most valuable team in the world right now is Dallas Cowboys, valued at 13 billion. We've got LA Lakers next, New England Patriots, which has won the NFL Super Bowl a few times. And then 10th is Manchester United, just below Real Madrid. And then we've got RCB at 1.78 billion. Why this number is exciting or a little bit cautious or somewhat forward-looking is because RCB is valued 32x of price to sales, basically. And other teams, like the Cowboys, are 12x. Lakers is 14x. Man U is 9x. And many of these teams make losses. RCB is, at least in the green. Our league is unique, where BCCI controls most of the media rights. So that's why franchises get a revenue share. Our colleague told us that it's similar to what NFL does as well. And they're thinking that the current BCCI deal is extremely sweet, 6.2 billion, which expires in 2027. And buyers are looking at the next cycle, hopefully increasing that number and the cap. So, yeah, that's what's happening in the world of IT.

Dhruv Sharma: It explains why every month or so, I'll hear from someone who wants to organize an SPV of some sort with some domestic and some foreign capital to acquire a minority stake in one of the teams.

Utsav Somani: Are you talking about one of our guests who's come on the show?

Dhruv Sharma: Oh, no. I mean, just in general, like every other month, there's some sort of interest in doing this.

Utsav Somani: Very cool stuff. RCB is actually pretty hot in the unlisted segment, by the way. And Mohit Menon also keeps on tweeting that he wants to organize an SPV for RCB.

Dhruv Sharma: He does that too?

Utsav Somani: Yeah, he does that too.

Dhruv Sharma: RCB is the one with Virat Kohli, right?

Utsav Somani: RCB is the one with Virat Kohli.

Dhruv Sharma: Sorry, I'm a cricket illiterate, so.

Utsav Somani: And season is starting on Saturday, so let's tune in because, I mean, there are different waves that happen in terms of sponsorships, right? We had the FMCG companies going big when the league started in 2008. Then we had a bunch of different waves. The most recent before the current AI wave, like now we've got Google Gemini, we've got ChatGBD with sponsors. And now, before that, we had, I mean, the phones, I think the mobile phones, which was Nothing and Vivo and all of these Chinese companies which were sponsoring the league. And now, I think before this, we had RealMoney Gaming was a big sponsor, right? Dream11.

Dhruv Sharma: A lot of startups who ended up burning their fingers and some think of it as a curse, but hopefully the curse will break sometime soon. And the front of the jersey, right?

Utsav Somani: Yeah, I mean, it's one of the easiest ways to distribute, but I think you get diluted as well. I don't know if it stands true for F1, like F1 media rates still keep on going up and up and up. But I think there's so much saturation of cricket in the world. F1 still is restricted to a very limited number of races, and they have a fixed calendar across the year, unlike cricket, where they're just, I mean, different teams like Women's League, I mean, Junior League, U19, and everyone's doing cricket all the year round. So I think the value of, I think, per game, I think, goes down significantly.

Dhruv Sharma: I'm starting to think the F1 calendar for this year, I really hope it doesn't look different.

Utsav Somani: It is affected by the Bahrain and Saudi is not happening.

Dhruv Sharma: It's not. Oh, wow. Fewer races though. Okay.

Utsav Somani: Yeah, fewer races. I hope the rest of the races go on as per time. I think latest is they're doing a five-day ceasefire.

Dhruv Sharma: Yes, we're in the middle of it.

Dhruv Sharma: That's the situation changes every few minutes. So yeah, everybody's guess. Yeah.

Utsav Somani: Looking on the positive side of things, Jensen Huang on Lex Friedman podcast said that AGI is here. Did you listen to that podcast or that clip?

Dhruv Sharma: Not yet. But the thing with AGI is I don't know if we'll have a clean, we're here moment, right? It's not like we'll get to AGI, there'll be a mushroom cloud, you could see it from like 10 kilometers away.

Utsav Somani: Man, for me, like, I mean, when I use cloud for many other things, like even simple things for me, that's AGI. Yeah, man, I don't want it to create the next billion dollar company in front of my eyes. Although what Huang said is OpenCloud can build a viral app that earns a billion dollars. But I think even a 1 lakh agent or 100,000 agents cannot build an NVIDIA.

Dhruv Sharma: So I think these guys are going to end up with a my AGI is better than your AGI kind of situation.

Utsav Somani: My agent can talk to your agent and figure out AGI. Alright, but that's it from us and we'll see you on Friday. Have a good one. Thank you. See you on Friday.