speaker-0: episode of the ⁓ future finance. So it's our last episode before the summer break. We are going on a little holiday in August. So we'll be ⁓ returning in September with the new episode and ⁓ also we're gonna continue the cadence in Q4 so we're gonna do once a month until the year's end. Today we're gonna do a slightly ⁓ different format. So I'm sitting here with Alex Ravitz from Crossman. He's the head of AI and ⁓ we are gonna have a okay, microphone too low. ⁓ okay, now can you hear me better? nice, thank you very much. So I'm sitting here with Alex to cover the arguably the third hottest topic of the year. We covered tokenization, we covered stable coins. Now there it we we are covering ⁓ agenting payments. So very, very ⁓ important topic for institutional adoption of ⁓ a of course ⁓ digital acetrails in general. ⁓ Alex, ⁓ welcome to the show, man. Pleasure to have you.

speaker-1: It's a great it's awesome to be here. Yeah, I'm excited to chat.

speaker-0: Perfect, perfect. So we're gonna do a fireside chat format. So it's going to be up to thirty minutes, twenty five, thirty minutes. We're gonna aim for that sweet spot. So ⁓ I wanna ask our beloved audience to please submit ⁓ questions and ⁓ yeah, let's without further ado, let's start with let's start with ⁓ the the show. So Alex, ⁓ I know of you I followed you ever since you founded Dymo, so five years ago, ⁓ more or less, but now you are ahead of AI at CrossMint. So when you are discussing like agenting payments with your clients and when your institutional clients, what's the most asked questions ⁓ that that they have for you?

speaker-1: The most asked question is where are the agents that are going to be paying for stuff? ⁓ there is quite a bit of ⁓ interest in you know, everyone believes that this market is gonna be big one day, but ⁓ the demand is still very, very, very early, very, very, very nascent. ⁓ I would say that the vast majority of agentic ⁓ transactions that have occurred are pretty much just experimental. ⁓ there's you know, if you think about what people spend money on, what businesses spend money on, there's, you know, so many different categories of things. And the the question is like what category moves first? ⁓ obviously there's been a little bit of traction, teeny bit of traction in this agent-to-agent or agent-to-API based space. ⁓ but there's still no other clear areas where agent-to-commerce is really starting to take off. ⁓ that's the big thing that everyone's trying to figure out right now. Is it gonna be consumer assistants that buy you groceries? Is it gonna be business assistants that handle procurement? ⁓ you know, is it gonna be fashion assistants that help you buy clothes? Like what is the actual thing gonna look like ⁓ where we start to drive the flywheel of consumer ⁓ transactions?

speaker-0: out of all of these things that and use cases that you ⁓ right now mentioned to us, so where what's mostly hype and what's actually real right now in terms of sizable ⁓ volume amount.

speaker-1: The ⁓ I mean I would say a lot of it I mean pretty much all of it's overhyped. ⁓ I don't think any of it can be can be like proven as real at this moment in time. Right. ⁓ you know, there is ⁓ hang on guys, sorry. ⁓ I saw there was a note about my fan and I'm trying to see my if I can change the microphone. ⁓ yeah. I can't change the microphone settings, unfortunately.

speaker-0: Yeah.

speaker-1: Yeah. Sorry guys. Yeah. Hopefully it's not too distracting. ⁓ the nothing's been proved out. you know, there's there's just too little volume. I mean, if you look at X402, it's last I checked, thirty thousand dollars a day, roughly. Like that certainly is not what you would consider any form of of validation. ⁓ so we're still very much prevalidation that agents ⁓ are are the new economic actors. I think, you know, there there's certainly consensus that it will get there. ⁓ but there's I think both the technology issue and like there's a lot of ways that the technology here needs to improve, but there's also the behavioral issue. I mean, most people are not ready to delegate ⁓ payments to to an agent.

speaker-0: Okay, perfect. So ⁓ I have one more qu one more sub question to tie into that. So do you buy the projections that ⁓ age agent economy will be like one trillion in in in two, three years? What's what's your honest assessment how big is the market right now? You don't have to give me a exact ⁓ number, but let's say a range.

speaker-1: Yeah. These projections are always I don't know. Like in my experience, there's there's always a way to like fit reality to the projections if you want. You know, ⁓ somebody will come along and count some big category of transactions as agentic because an AI was part of the way in there. ⁓ I certainly don't think in three years there's gonna be a trillion dollars of agent mediated activity. And there's also a lot of like, you know, like what are we gonna count in that? bucket is, you know, right now everyone is basically trying to get to the point where the agent asks for my you know, I tell my agent buy one thing. It asks for permit it like confirms permission and goes and does it. ⁓ we're a long way away from the like, hey agent, you now have a budget and don't ask me to approve every purchase. Like I'm just gonna let you I'm gonna let you manage this budget. That's a long way off. And so when we say, okay, how much money will agents move in three years, five years, ten years, w what do you mean by that? ⁓ am I does that mean a human is in the loop approving every single purchase? ⁓ does that mean those agents n are autonomously r ⁓ running their own budgets? Does that mean ⁓ an agent is then assigning a budget to sub agents? Like we we we don't really know. ⁓ I certainly think that it's probably like three years away from getting into the billions. and you know, we're in the low tens of thousands right now. So that would be a pretty insane growth rate if we got there. but it's gonna take some time for both the behavior to fall into place and for all the technology pieces to come together. So

speaker-0: There are many questions to be answered before ⁓ like moving. So let's say from proof of concepts phase to actual deployment phase where agents move sizable I I call it like human check sizes. but ⁓ to that point, so you and I talked like a week ago and you told me ⁓ one very, very interesting take, which is that agents don't actually need stable coins. So you at the cross means you you're building you're optimizing building for both Rails, let's say the the the the the legacy card ⁓ network systems and also the yeah ⁓ X four ⁓ two, right? ⁓ a agent ⁓ payment standards. So but can you elaborate to us why do you think that agents actually don't need stable coins?

speaker-1: My so ⁓ the the the thing I would say is that like most merchants don't accept stable coins. and ⁓ you know, certainly for microtransactions, stable coins make some sense. And we'll get to the question that ⁓ Selena last asked earlier about ⁓ risk, because I think that's a really important factor in in this discussion. ⁓ but the reality is like how do people pay for things today? and and most people they're using cards or they're using they're using, you know, other payment methods. Like ⁓ I think that our thesis, working thesis here at CrossMint is that agents are gonna need access to a bunch of different payment methods, and that's what we're gonna provide. So today we can give if you wanna give your agent the ability to move money, we can give that agent a stable coin wallet and stable coins, or we can give it access to Visa and MasterCards. Eventually we'll add Amex once their product is ready. And I believe that when we started talking to other payments companies, other payments companies are building their agent native protocols. ⁓ and and that's gonna look, you know, in some way like, hey, I want to give my agent access to my Fenmo account, Cash App account, bank account, like you know, local rail, Pix account, B zoom, like yeah, there there's There's hundreds of different ways. You know, if you look across the globe that agents might need to move money. ⁓ and I think eventually every single one of those providers or most of them are gonna build ways that you can securely delegate authority to an agent to spend some amount. ⁓ we're in the very early stages of that. I just think that it's gonna go the way that merchant adoption goes, which is what do merchants accept today? I mean, merchants are famous for not trying like they don't want to change. They haven't like over 20 years almost of the crypto industry, they've not wanted to accept most of them crypto. and so, you know, I just don't think that it's necessary. And the other thing I'll say on microtransactions, people point to microtransactions as like the, you know, well, y you can't do that with a card, right? ⁓ I would say like I would not expect that the payments companies, broadly speaking, just roll over and say, All right, they got us there. In fact, I don't you know, in June, MasterCard made an announcement. They're rolling out a product agent pay for machines that's blockchain based, actually. but you know, the the the death of the payments companies because they can't do microtransactions has been vastly overestimated.

speaker-0: Yeah, ⁓ agreed. And majority of population, let's say let's globalize, ⁓ they are using like card networks and it worked like what three, four, five decades back in the past, and they're not willing to change it. And merchants, to me, like every time we talk about ⁓ stable coins and the use case of the payments, ⁓ then the merchant friction is the number one kind of ⁓ the reason why the stable coin adoption didn't ⁓ you know catch up when it comes to payments. And they say like we don't want to deal with wallets, with accounting rules holding stable coins on our on our balance sheet, you know. many, many, many different aspects why why merchants ⁓ are refusing stable coins, accepting fiat, of course, today and they're not changing their legacy their legacy way of doing business. But ⁓ i to that point in your mind, do we need like ⁓ then ⁓ these protocols that are emerging agent to agent protocols, machine to machine protocols, or are card car networks with the human in the loop will continue to do the job just fine and will continue to prevail as a you know most dominant ⁓ payment method.

speaker-1: Well, so the card networks have launched, two of them at least so far, have launched agent native ⁓ protocols. And the big question is you know those protocols, just like X402 or MPP, those protocols are incomplete. They don't have, they haven't solved all of the questions, they haven't solved all of the things, notably around risk and and just to address the question that's now coming up a few times. ⁓ nobody yet knows what The KYA standards should ⁓ if I go and purchase something from a merchant, does the merchant need to be able to id you know, if I ask my agent rather to purchase something from a merchant? Does the merchant need to be able to identify who I am? maybe. Certainly the merchants want to, partly for a risk reason, but also partly because they want to be able to have a customer relationship. They want to be able to send me an email, they wanna be able to, you know. Thank me, send me a a promo later on. Like there's a lot of reasons that they want to know who you are.

speaker-0: you can return and buy more.

speaker-1: Exactly. And they want to create a good customer experience around that. and you know, we've had conversations with people, like everyone's trying to figure out, okay, how do you identify the agent and the human? Do you need to identify the agent itself? Do you need to identify the model the agent is used in? Do you want to identify the harness that the agent has? Like, ⁓ you know, how much detail do you need on this agent? ⁓ There's a lot of talk about using obviously like verifiable credentials around this, which is great. And MasterCard has also put out some spec about ⁓ verifiable attent and and a lot of folks are are thinking about how all of this fits together. ⁓ I I do think that there will be th you know what I think there's gonna be many protocols at the end of the day. Like it it's kind of been like It it's been protocol heaven for like every one every company's wanted to launch their own thing. And look at Shopify, they launched Universal Commerce Protocol and and there's ACP as well. And like there's all these different protocols. And frankly, like that's good for a business like CrossMint because somebody's gotta sit there and make all these things, you know, accessible to the end developers. ⁓ because developers don't wanna have to think about how many different protocols you need to support. And who knows? Maybe some of these went out and like lead to complete dominance, but like that's rarely the case that, you know, that a one single protocol becomes the the only dominant player. ⁓ on on the risk piece, since it keeps coming up, this is another area. Nobody knows. Nobody knows where the liability should land. And ⁓ this is an open secret. Like we talk about it w in the room with all the partners who we meet with. and if you think about it, like there's all there's so many ways an agent can go wrong. It could misuse, you know, it could go buy go rogue and like buy random things. It could buy the right thing that I ask it for, but it could get the quantity wrong. ⁓ it could fail to execute a purchase that is really important to my business. And now all of a sudden I don't have inventory or I don't have supply or something like that. And and you know, there's probably dozens of other ways that it could fail. ⁓ variations and you know now we're talking about a single item. It could get the timing wrong. It could get the shipping wrong like could send it to the wrong place. Like There's all kinds of ways that this agent can go wrong. ⁓ we have been like, I have met tons of builders ⁓ who are working on guardrails. And like there are some guardrails that are being built into Visa and MasterCards products. There's tons of indiv you know, independent builders who are offering their guardrails. ⁓ there's a lot of people in the space. I don't know whose guardrails are going to win. Like, frankly, ⁓ guardrails are a rather simple product. and You know, I think that many companies might choose to build them in-house because they might want, you know, for those that are doing like internal, like for companies, they might want more complex, you know, rules for their enterprise or whatever. ⁓ one of the really interesting categories that we started to pay attention to is those that are trying to price the risk of the of the agent's malfunction. I've met a number of early stage companies that are working on that problem as well. I think that like the guardrail alone is not all that interesting to me. The guardrail plus risk is super interesting to me. ⁓ because I think being able to price the risk is like really fascinating. Yeah. ⁓ you know, there's no data to do that today. So the people who are gonna go after that problem are really taking an interesting stab at things. ⁓ but yeah, we'll have to wait and see. I see there's a kind of another question here about

speaker-0: In fact.

speaker-1: No legal personality, what weight does an agent-driven transaction carry at all? Yeah. Yeah. OB, I'm not sure if you can clarify that a little bit. I'm not quite sure what you mean by what weight the the agent-driven transaction has. But th I think, you know, I'll I'll just speak about another kind of like interesting point around that question. ⁓ the you know, a lot like one of the really interesting things that people come to Crossman and they ask about all the time is like, I want to issue a card to my agent. I'm like, well, you can't do that legally. Because you can't issue you have to issue a card to a human. And maybe that'll change in three years' time. There'll be some legal framework to give agents some form of legal standing on their own. And typically what we see is like some small jurisdiction, and we saw this a lot with crypto regulation, of course, like some small jurisdiction like a Bermuda or a Malta like is like, Hey, we'll be the first ones to try this out and figure out what it looks like. ⁓ and then like, you know, other countries will watch and see how it goes and And try to adapt from there. ⁓ but there is, you know, you can't issue a card to an agent. So you have to issue it to the human, and then maybe you, you know, you give the agent access to that card via some protocol. So it's you know, i there's gonna be a lot of different ways that that that changes over time. ⁓ and then yeah, where does the legal liability land? ⁓ is is this gonna be something that gets navigated and probably litigated in courts? Many times over, as we enter this new era of commerce.

speaker-0: Yeah. How how long in your mind will it take for this ⁓ legal for this liability case to be f ⁓ f like ⁓ f to to enter the stage of formality? We know that okay, if this transaction fails, I know that someone ⁓ can be ⁓ held accountable for that, like in terms of years.

speaker-1: I mean it never really gets settled, right? Like there's there are c always disputes all the time, right? Like I don't know if you've ever disputed something with, you know, your your card company. I I had to like dispute this subscription and like the vendor kept telling me we don't have a record of your subscription and the card company was like, Well, we can't see anything in their systems. ⁓ like we like I've been through hell on disputes before. So It it always is ⁓ there's always some amount of of like, you know, gray area. Like that's that's why the law that's why their lawyers exist, you know. If the law was binary, then like you wouldn't need lawyers to go and fight about what what the words mean. So I think it'll get better. I think we're it'll probably improve to the point in like two to th like two to three years is generally what what I'm thinking about for a lot of these things. Like

speaker-0: ⁓ fair point.

speaker-1: when the guardrails will mature and the risk will mature and the liability stuff will mature. I just think we'll have established more established patterns for that in like roughly the next two to three years.

speaker-0: Two to three years. Okay. So that's like an approximate time frame. ⁓ Ob ⁓ went back to us with a better formulated question. So he said that how can a transaction be legally characterized, for example, declared null or or void if one of the parties is has no legal personality and therefore cannot be subject to legal ⁓ resource such as an AI agent. Yeah.

speaker-1: I would say that like at this point nobody really thinks of the agent as an independent actor. And like everything is take is is is acted upon by the human. You know, you can't I like like I said, nobody is is really at the point right now where ⁓ they would recognize the agent's individual action as like the agent decided to do this, therefore, ⁓ Like we don't know how to go after this agent. It's always, at least for now, going to come back to the human and it's going to be like, you sent some prompt. You gave the agent some intent that kicked it off. And that that's where the liability is going to go back to. and ⁓ you know, it again, like if you look at what MasterCard has, you know, published around their verifiable intent, spec. ⁓ whole goal is to try to coalesce the industry around some standard ⁓ to to to make this ⁓ possible. Yeah. And so Arb, like, you know, like well, so you're saying like there the you don't have to necessarily like have a upfront KYC, but like ⁓ if you end up in a dispute where like your Yeah, exactly Laura, where your agent is like, you know, well, your agent wants to dispute something, then like somebody will You know, you have to identify yourself 'cause otherwise you're not going to be able to like resolve the dispute that that you're looking to accomplish there.

speaker-0: Nice. I hope that ⁓ answers your question, ⁓ Alban, ⁓ good good point, Laura. So ⁓ Nathan asked ⁓ also very important question, like they are bombarding the chat. It's ⁓ very good guys, keep keep it up. So he said ⁓ i isn't the trust the main problem of this ⁓ sector. Don't people want something that works hundred percent? Because one percent error is not possible with money. What do you think?

speaker-1: I think a lot about ⁓ autonomous vehicles here, which is you know, autonomous vehicles don't have to be a hundred percent safe. They just have to be better than human drivers. and ⁓ and so the question is like do you need to have a hundred percent reliability with an agent that's handling purchases? No, I d I don't think so, you know. ⁓ if it let's talk about like a business that wants to automate its procurement function with AI, do I need to have it ⁓ be a hundred percent right? No. Do I need it to be better than the humans I'm paying to do the procurement function right now? Yeah, probably. with AI, there's this other element. And you know, you could argue with other things too, or even with autonomous vehicles too. It's like, what is there's probably some value to the convenience? Like, are you willing to To take some hit on reliability if the convenience is off the charts more convenient for you, maybe, or the cost is off the charts. Like, you know, would I take an AI if I had one pro like, you know, if I had a procurement d division of 10 people and I could replace it with one AI, would I take a few mistakes? Well, it depends how costly those mistakes are. I think there's gonna be a lot of calculus, but but Nathan, I think your point overall, the the rile reliability of this stuff does have to vastly, vastly improve. And everyone gets very excited and I talk to founders and I talk to people all the time who are like, Have you used Fable? Like it's so amazing. You can do anything. And I'm like, actually I have used it a ton. And like the more you use any AI product, the more you realize the mistakes that it makes and the shit it hallucinates and how it ain't perfect. ⁓ and so I think like we're actually quite a ways away from like really having this stuff dialed in. and, you know, it's it's just it's gonna take time.

speaker-0: Yeah, definitely. And ⁓ good ⁓ like your ⁓ analogy with autonomous self driving with the FSD for example from Tesla, it resonates very well because they it it with each new upgrade of course it's getting better and better, but ⁓ these one percent errors will essentially happen. It's happening currently, although God, ⁓ like good thing is that they're not fatal. ⁓ and ⁓ the the the FSD will get better of course with with the bigger updates and more usage. But ⁓ I hope that answers the the question for you, Nathan. Before I ask you one last question from my side, ⁓ there is one good ⁓ question from Gabriel that actually asked about what you're doing at CrossMint. So ba he's question is Are you working on agenting on chain payments for capital markets use cases currently?

speaker-1: We don't care as much about like what the use cases are. ⁓ you know, our goal is to be able to support like we're just the plumbing at the end of the day. It's like how does the agent get access to the money and then how does it handle executing on certain transactions? So ⁓ we are we're talking I wouldn't I don't know if I would call these like capital markets use cases. We do have some ⁓ agent builders who are building, you know, agents that trade for you and stuff like that. Like In those instances, maybe we provide the the wallets and the on ramps. but the plumbing in terms of like what APIs the agent calls, that's you know, ⁓ it's gonna call the Robin Hood API or some other markets API or something like that. That's that's not what we're doing in that instance.

speaker-0: So you're basically use case agnostic where you just supply the tech. If I some

speaker-1: Pretty much. Yeah. I mean there are s yeah, there's some use cases that are we're not as touching as much. ⁓ but ⁓ yeah.

speaker-0: Alright, ⁓ fair enough. So you know already that ⁓ majority of our audience and and clients are based in Europe, so PSPs, asset managers, banks. So I'm interested in knowing your your case on ⁓ and if you can give advice non-advice what what should what should they be doing ⁓ today? ⁓ so is this ⁓ agent ⁓ like payment stuff, is it ⁓ twenty twenty six topic will be as hot as it is in ⁓ right now in twenty twenty seven or can actually ⁓ should they all already everyone that's doing PSP have like a lie proof concept implementing agents in their in their day to day business? But strictly th thinking about Europe right now.

speaker-1: Just well, you know, I don't know if I would give it that much of a different spin in Europe or the US. I think some interesting European ⁓ some interesting developments. I mean, like there's the whole like ⁓ PSD two in Europe, which is like aiming to make verifiable credentials and pass keys like more uniformly adopted. that is I think something that like pass keys are a great user experience. ⁓ they are an upgrade over 3DS right now. And ⁓ they are also gonna enable like a lot of this stuff to fit together better because an agent can't really handle 3DS on its own right now, but with a passkey that the agent serves back to its human operator, that's a lot more feasible, a lot smoother. ⁓ I definitely think companies should be l you know, planning their their roadmap around that. you know, if you're like It anything that improves the user experience there is gonna drive volume and like that's just gonna be the most important, you know that that's ultimately what what drives about here. So ⁓ that's something that I'm thinking about and that that we're seeing ⁓ on on so

speaker-0: You would say that UX is like the top North Star ⁓ for the for these companies. Like regardless of their location.

speaker-1: I don't know if I would say that. ⁓ UX is always an important North Star. you know, there there there are other things to consider, but ⁓ you know, it's it's it's definitely a technology. Like the passkey itself actually is is a handles a handful, like it adds security, it improves the UX, it opens up a range of ⁓ it just like really streamlines a lot of agent activity in particular. So ⁓ it kinda kills a few birds with one stone. Perfect.

speaker-0: All right, ⁓ so that's the top solution that it should implement. ⁓ I hope ⁓ some people from these institutions are actually listening right now. and ⁓ we have like two more questions or three more questions in the chat. So I'll just ⁓ read them out loud. So Lana asks, touching back to the ⁓ cardrails, so what's your take on how the agent commerce and agenting payment ⁓ technology will affect car la cardrails in the near and long term ⁓ future?

speaker-1: Yeah, I mentioned this a little bit at the beginning. ⁓ I'm not convinced that like it's gonna dramatically change the cardrails in the near term. ⁓ maybe in the very long term, like cardrails continue to shift and and and morph over time. ⁓ you know, there's certainly but but like I just in the near term, like that's still the way people primarily pay. I think a lot of people in the crypto industry. like to think out to that time. I don't know how long it'll take, but like once you get everyone on chain and once they have a little money on chain, maybe you can keep them on chain. ⁓ but like we've been waiting for that day. I've been in this industry for a long time now. Like we've been waiting for that day for a while. And ⁓ it, you know, just it's it's a ways out there. So I don't think in the near term there's going to be a huge impact on the cardrails. ⁓ and you know, I'm I'm looking at some of the other questions that are here. Zach, you know, you asked about how is it different from algorithmic trading. I mean alg algorithmic trading is just like a very specific use case. I guess you could call it agentic commerce. ⁓ it's just, you know, it's one, you the the question is like how do you have whether it's an algorithm or an L L ⁓ whatever, how do you have that thing, that bit of code, go handle a whole bunch of other types of payments and and purchases and and stuff like that. ⁓ so maybe it's more of an expansion. ⁓ and then Laura asked ⁓ Do you ever see see a time when there will not need to be a certain level of human oversight? I mean, I think that like that's what the trend of all these kinds of things are. You know, whether you're again just whether you're looking at autonomous vehicles. When they launched, when the autonomous vehicles first launched, there was a human driver. You know, Waymo had a human had hundreds and hundreds of d human drivers sitting in the car ⁓ at all times. And then They had a driver remotely and they still have people remotely. And like eventually maybe, you know, you'll have completely unsupervised. ⁓ but like that's gonna exist of all of the agent, whether we're talking about agent payments, agent commerce, anything else, like, you know, there's a human in the loop now. And the whole goal of where we're trying to get to is like how much can you take that human out of the loop?

speaker-0: Out of the loop. Perfect. And to to give like food for thought to the audience and ⁓ to wrap up with this question, like one last. ⁓ y we you are of course coming from the deep in space and you're familiar with crypto. So where is crypto oversolving today?

speaker-1: ⁓ I don't know that I don't know that I would say that crypto is oversolving. ⁓ I mean, I'll tell you the area where crypto's oversolving. We definitely don't need any new L ones and any new like there's no such thing as an agent native L one. And like I continue to see these things launch where it's like, ⁓ like agent payments need this new L one to work. It's like, no, they don't. Like ⁓ That's probably an air the area where crypto has like most provably oversolved. I saw something on Twitter the I think in the last couple of days, which was like seven point five billion dollars raised by these protocols in the l like the last thirty days of revenue on these protocols was like ⁓ you know, fifteen hundred dollars. And you know, it's a bunch of stuff like EOS and Barrachain and like all these things that at some point launched with tons of hype but like have effectively no usage. So The p the place where crypto is most oversolving is in the number of protocols that exist ⁓ today. Yeah.

speaker-0: Exactly, exactly. That's the point that ⁓ we were like we are trying to send this year is that the great consolidation is needed without a doubt and ⁓ majority of this ⁓ I don't wanna offend anyone but of not used chains and the infrastructure like layers need to need to well disappear in order for the whole industry to to thrive and focus on the handful of winners. Like it it it's the normal thing, like it it a as in any industry.

speaker-1: Yeah, we'll I mean we'll see what happens and and we'll see how many of them disappear or try and pivot again and you know, ⁓ it's it's it's not about the technology. Like sometimes good stuff just doesn't get adopted and you know, there's all kinds of reasons for that. But ⁓ no, I I think that's like it it's it's just it's one thing about what's happened in the past. It's another thing of like people showing up with their new L ones still today in twenty twenty six. ⁓ that that surprises me.

speaker-0: Yeah. And pushing it to your face like give us money to raise. But no, no, no. great, great point, Alex. So I I hope ⁓ guys you're satisfied with the answers. I really enjoyed the the chat. It f time flew by extremely quick. I I didn't realize it's like 35 minutes since we started live. ⁓ Alex once again, thank you very much for tuning in from NYC. ⁓ I wish you luck at Crossmint and ⁓ yeah, recordings will be available on our YouTube and ⁓ our podcast platforms, so shpotify and ⁓ an Apple podcast primarily, but ⁓ yeah, ⁓ re you can re always reach out to me if you want to watch a re replay where you can find it. ⁓ with that being said, I wish you all ⁓ happy holidays and will be returning in ⁓ a month and a half. You are welcome.

speaker-1: Thank you very much for having me on.

speaker-0: Bye bye guys.