Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
SKALE School: Stablecoins in the agentic era!
Recorded: March 24, 2026 Duration: 1:01:49Space Recording
Full Transcription
Scale school time! Ladies and gents, we hope you are sat down, you got your coffee, you got your avocado toast, that's what, you know, university students and stuff are eating, you know, stealing all their parents' money.
It's scale school time! Sawyer, how you doing, man? Having a good day?
I'm doing better now, that was funny. That was funny.
Yeah.
Yeah, I don't think I could afford an avocado toast when I was at school.
Yeah, me neither. Me neither know that dude i hate avocado um
honestly i'm not a good fan of avocado toast anyways i like it more like mexican food
oh respect respect uh fede how you doing man welcome to scale school hey hey hey everyone i literally
had avocado toast for breakfast this morning so yeah oh that's not even planned that's crazy let's
go yeah it's my go-to breakfast so yeah uh with x as well oh dude nothing beats like a nice i've
started eating breakfast i because i i don't really eat breakfast
i hadn't when i you know for the past like five ten years and my wife now is like i sit down at
my desk and she just brings me like just toast like just a little bit of buttery toast and i'm
like i finally did and um but it's been going for a couple days now and now i want to up it and see
if i can get it to make me eggs on toast because that would be a great start to the day. So what'd you have for breakfast, man?
Nothing. Coffee. Yeah. Lots and lots and lots of coffee. Lots of coffee. Lots of coffee. Well,
you're going to be ready for our discussion today
then as we talk about stable coin payments in the age of agentic commerce and of course that battle
between you know regular payments stable coins you know who's going to come out on top for the
world of ai um personally i don't think i would ever trust an AI agent with my credit card.
So yeah, what about you, man?
I can't answer that first.
I think first we need to put that into FedExCore because it's probably kind of a rude statement
to a man here representing the team behind Lobster Cash,
which is literally putting credit cards into...
Well, explain to me...
Agents don't have hands. Where do they put... Is it into the system memory, into the context?
Where do you put the credit card?
Oh, wait, are we just jumping straight into it? If that's the question, I would totally give my
credit card to an agent, 100%. But I come from crypto a lot, right?
Like I've been in crypto since 2017.
That's why I get my name.
So I've given money to worse things than an agent, like an untrusted human or like weird
tokens and weird ICOs back in the day.
So honestly, giving a card to an agent doesn't feel that risky.
But yeah, we can talk about that a little bit more.
Yeah.
Well, obviously, the big question is going to be, you know,
as agentic payments do become more real paying for API access, paying for, you know, random
storage and, and, you know, other things that, you know, the expanding agentic economy is,
is going into, um, what is, you know, open question to you both, you know, like for the
people that are not a hundred percent sure of you know what's what's
going on you know what is this difference between utilizing stable coins or you know agent link cards
you know how do they differ what are the you know the risks and and you know let's let's let's start
there man that's a loaded question before we we answer that though, can we just quickly lay the groundwork with one thing and then I'll let Fede answer that.
So I think this is really relevant towards something I've been thinking a lot about.
Almost every single thing you just mentioned at this point in time does have stablecoins in the lifecycle of it, which is kind of unique, I feel like.
kind of unique i feel like so and what i mean by that i've been thinking about this a lot um there's
So what I mean by that, I mean, I've been thinking about this a lot.
obviously been this this surge of agentic credit cards right that have kind of popped up the last
last month or so and so many of those are actually being funded with stable coins people are actually
making like an x42 payment or they're making a stable coin deposit to get which i actually feel
like it's more of a debit card than a credit card correct me if i'm
wrong here um because it does have an explicit limit based on what it's topped up to so again
tomato tomato but um there's that what is also interesting is if you're making a payment with
a true credit card or like if for whatever reason you're using maybe visas uh you know some of these
protocols actually do let you give a more traditional credit card, from my understanding.
They're, you know, they're actively exploring and consuming stable coins behind the scenes for enhanced settlement tax.
And so there's another example of stable coins being potentially involved in this process.
And then lastly, you have your direct stable coins, right?
An X for a two API or endpoint, a new MPP machine payments protocol
from the Shrek Tempo team
also accepts stablecoins.
There's L402.
There's a bunch of different ones as well,
but it all comes back to the same thing,
which I think is interesting,
which is stablecoins seem to be
a consistent component of all of these flows.
So I'm gonna pause there,
and I would love to have you answer that original question,
which is, what's the difference between all those?
Why do you give one or the other to an agent?
Yeah, there's a lot here, it's something happening.
Yeah, there's quite a lot.
So right now, so first of all, what I always like to clarify to the audience
or whoever we're talking that is not deep into these problems is that the three of us
on this call, we have our minds so ingrained and so deep into the problem that we talk
about these things like it's a solved problem or it's like something that is, oh yeah, everything,
like all of this that we're talking is done, it works.
Tomorrow you can totally build all of this. And most of the things that we talk about are kind of short term future, right?
And a lot of the things are new protocols. And that means things have not been adopted by most companies,
by most things working out there or most agents, right? So when you see OpenClaw, when you see Claude itself giving you
access to your Claude instance through Telegram, when you see all of these new things coming out,
these are all different approaches to the agent economy that we're trying to kind of define. So
all of these things we can talk about ours, but we won't know until
they are kind of adopted by the rest of the people, right? And the industry. So with that
said, the things that we are talking today and trying to push forward are the X4.2 payments
that we'll mention that involves stablecoins and then mostly agentic cards, whatever that
is, and we'll try to define that, right. So why do stablecoins show up in so many conversations and what these are?
I hope almost everyone has read about them somewhere, but it's basically tokens on blockchains.
It could be on different blockchains, whichever you know, maybe Ethereum, Solana, Scale, whatever, different chains,
different ecosystems, different approaches to get this done, right? These are tokens that
represent one-to-one the value of the US dollar or any other fiat currency, right? In this case,
stablecoins have gotten a lot of adoption for agenting payments because they are 24-7. They,
like stablecoins and the blockchain infrastructure
does not sleep, it doesn't need humans.
You have nodes running 24-7,
so the infrastructure is on 24-7.
So that is great for agents because you can have a script
running on a Sunday morning and it will work
across the globe as well.
They are also great for microtransactions,
meaning if I need to pay two cents for something,
doing it via credit card or ACH or wire doesn't really make sense because that is really expensive
and whoever has tried to send money across the globe has had this issue. The problem is a lot
of people have not even seen this problem because they usually just transact inside their country
or via payment rails that are already there. But when you try to go out of the system and try to seen this problem because they usually just transact inside their country or like, you know,
via payment rails that are already there. But when you try to go out of the system and try to
bring in new use cases, and if you're part of the crypto community, you probably have friends
all over the world and you've been in a conference, invited dinner, like, had dinner with some friends
and whenever it's time to split the bill or something like that, it's chaos. It's like, oh, do you have whatever software you're using in a specific country? And it's like,
no, of course I don't have that. I've always found the easiest way to settle that debt with
stablecoins. So yeah, that's kind of what I'm seeing now is like, so that everyone understands
whatever we're talking today is
valid today. It may happen that in three, four weeks, a new company comes out with a new standard
that everyone adopts super fast and then that is going to be the new standard that all agents are
going to be using. Yeah. Amazing. Amazing. I love it. I want to pull out a few things there.
Most of them very technical, one a little bit more, maybe just funny, ironic, especially if you're in a,
if you are from the crypto crowd and you've been around for a while.
On the technical side, I fully agree, right?
From a, I think the,
both the fun and confusing part for people is crypto historically moved
still fast.
And if you're not in the crypto space actually, or you weren't,
or you're not in this AI crypto space yet,
you probably hear people say that and they're like,
no, you work the same job.
They just don't get it.
They don't understand that a couple of weeks in crypto
is a few months in the real world.
And a few months in crypto is,
it feels like a decade in the real world
right like half the people you're talking to three months ago have pivoted 10 times by now
and are doing something else and the other 10 people are now somehow raised their series c
and they were a pre-seed company three months ago it's just like what doesn't make sense
uh but that's that's the world we live in right is like that that has literally happened and so
the world we live in right is like that that has literally happened and so i think your your call
out of just that these protocols are here they work but they're not necessarily here to stay
is such a good it's such a good way to look at the entire industry as a whole because um
i feel like the only thing again that has functionally proliferated itself
across the last what, eight years, nine years, 10 years,
is the concept of a stable.
Everything else for the most part
has come and gone multiple times.
But the only thing that has truly retained itself
is a stable coin.
And that's because all these other things require
such significant levels of adoption.
But the one thing that a stablecoin has that these others don't is it inherently has a value.
And when something has value, it's a lot easier generally to get adoption once you get part of the way there.
So obviously, I think probably they tether for this.
They just dominated for so many years from a USDT perspective.
Obviously they are still the number one stable coin
in the world from a total issuance on chain perspective.
I think they're sitting at what?
Like almost 300 billion or more these days.
It's definitely a spot.
I don't know the exact number, you may have it,
but I think the part that's really cool
is in the last couple of years, we've seen a really great growth from other statements.
And this has come from, I think, a lot of it compliance, which is one of the things I wanted to kind of pull out what you're talking about is why do these things change?
Like why do we have a protocol that exists today but doesn't exist next month?
Why do we have a protocol that maybe existed six months ago but isn't used today?
And I do think a lot of it comes down to people want to own the stock and at the same time,
they also need to be compliant enough to work with what they're doing.
It's no secret.
Stripe's a really big company.
Coinbase is a publicly traded company.
They can't just do things the way that, you know,
a random dev off the street can experiment.
They have a lot more they need to be compliant with
to ensure that they're meeting the shareholder requirements,
the public SEC requirements.
And in Stripe's case,
I have no insider information, I promise.
My assumption is that one day they will probably want to go public because that's what most big tech companies feels like need to do.
Maybe they won't.
But if they do, they also need to make sure all the payments they take are tracked and done properly because that all gets taken into account during
the filing for going public, right? So crypto is really complicated because crypto is really
complicated, right? We know that. So just want to kind of pull those two things out. It's really
interesting because most people don't talk about that, but it's a really unique perspective of hey like yeah uh we just had an mpp drop like is x402
going to disappear probably not but like could it yeah it could um at the same time someone else
could drop new protocol tomorrow and both could disappear right and we're all three could exist so
we really don't know um if you want to call it one funny thing before I stop my ramble here,
you said about the, hey, we're out at dinner, we're at a crypto event.
It's time to pay.
What do we do?
And I think the age-old joke about stablecoins is no one ever wants to pay with crypto.
Everyone wants to pay with crypto, but most people wind up not because they don't want
to dox their wallets.
And most people wind up not having a burner wallet or a payment wallet that is separate from the holdings, just because after all these years, apparently we're still terrible at doing this.
Maybe that's just the UX problem with crypto, right?
Maybe that's just the UX problem with crypto, right?
But I want to call out that this is actually something that we're working on at scale,
confidential tokens in on-chain privacy,
because it is a really big thing when it comes to not just humans, but also agents.
At the end of the day, whether you have $5 or you have have 500 million dollars it doesn't really matter
what matters is you don't want everyone else to know how much you have when you're buying
dinner for a friend or you're in a group splitting a payment um we call that the barista test right
there's a very simple benchmark that we've established which is the person you're paying
can know how much you're paying but but they shouldn't be able to see
how much you have in your bank account.
They shouldn't be able to see how much you got paid last week,
et cetera, et cetera.
The only people that should really be able to see that is you.
And in some cases, right, like maybe the government
or your accountant or an auditor or a regulatory entity,
these are all depending on who you are, what you are.
These are things that most people
have become comfortable with in today's society if you don't have a bank account because you've
been debanked or you just really don't like banks then maybe this isn't for you but for the average
person who does have a bank account this is the level of confidentiality you've come to expect
right which is only a very select few groups of people can see it that's actually
something that we have um i was actually posting some screenshots of it last week
uh which is making confidential payments with x402 and mpp using usdc um on a scale chain
because well again if i'm paying for things i don't need everyone in the world to know how
much i have in my account it just seems ridiculous um And the cool part about that is it can be used at any part of
the site. It's kind of putting two and two together with one of the first things I said is stable coins
might be used, just not where you see them. You might be using a credit card or an agent card,
but there could still be stablecoins involved with the deposit,
the withdrawal, the settlement, etc. And all of that can now become confidential as well,
which is really cool. I'm not sure, Fede, if that's something you guys have come across or
been thinking about. Well, you mentioned a lot of things, but privacy is really important,
especially in crypto, I think, because of the thing you just said.
What I think people still... So I am a believer that crypto is going to have much more use cases
than just payments. I've chosen CrossMint actually. So like I have been in CrossMint for like nine to
10 months now. And I purposely made the choice to choose CrossMate because I love
the intersection where we're at, where we're trying to solve like wallets and crypto infrastructure
for traditional businesses that want to move money around. Because that's kind of, as you said,
one of the main use cases that we've seen crypto solve, right, and blockchain in general.
But I also have my personal views on that. I think that DeFi is actually
revolutionizing the world. And I know a lot of people are saying it's dying. I understand
the risks of it and a bunch of other things, but I do think crypto has a lot more applications.
The thing is, it's a really, really young industry. Like people don't understand how
long it takes for this kind of things to evolve. It's like, you cannot expect to just a random debt, build a smart contract and then create a whole economy on it.
Like that does not work like that.
If you go see the banking industry, the funny thing is that, so my father is an economist, right?
And he's been working at banks for his whole life.
So I do have a little bit of education on that form, but I don't have formal education on economies and stuff. I've just dug in very
deep into crypto and stuff. And every now and then, every year or so, I would come back to my
family and be like, hey, dude, look at this cool thing of DeFi that you can generally yield doing
this, this and that. And he's like, bro, that already exists. That is called this in the banking
industry and that is called this. And I'm like, oh, that's crazy, right? So we're replicating everything the banking industry has
been doing in finance for like years, just bringing it on chain. And that makes kind of a lot of sense,
right? The problem is you find a lot of issues with that and it's not really straightforward.
And then making it permissionless brings a lot of issues with regulations and a lot of other things.
So what I think we're seeing with crypto, with blockchain, and the crazy thing that people are not acknowledging is
most people are talking about it now.
Like it's an actual industry that's respected.
Companies like Stripe, JP Morgan, and every big company in the world is exploring blockchain, hiring
blockchain experts and building their own approach to it.
So that means it is an area of interest.
Right now it may seem that AI has taken over the whole crypto industry, but it's not.
It's just that we've gone more institutional.
And this is what we were shouting and claiming for in 2018, when we were like, oh, when are institutions coming in? Well, they are now, right? They're here. Now we got to fight
with regulations. Now we got to fight with how to make sure we have a good user experience and how
we can pay back someone that made an incorrect payment, for example, or like, you know, charge
packs and stuff like that. So yeah, everything that you're saying, you're like really resonates
with me. There
is the professional side of things where you can work at right now, and then there's people
working more on the visionary stuff with zero knowledge proofs and more forward thinking
things that are solving problems that are going to come in the next year or so. And
AI can accelerate 95% of all of this, but there's still a part where humans are involved, which
is regulations, which is governments, which is the banking infrastructure, the old one.
And I'm going to go back to the subject of agentic payments with this.
So this kind of history on crypto, and I love talking about that, but why stablecoins or
credit cards, right?
And I think we gave the perfect introduction for this.
And I think the answer for that is stablecoins, I think stablecoins are the best form of payment
that we can have today for agents.
It is the best.
And we have a lot of infrastructure backing that, that could make it work even better than it is today with X42 and stuff. But if you want agents
plugging into the economy today and buying things today and getting access to 99% of
the internet and the things that you can buy online, then you need to give them credit cards because that is the
number one form of payment online. It's credit cards. And the whole world has worked really
hard to gain the trust of people enough to put their credit card into a website. If you remember,
I don't know, like 20 years ago, my parents would not, like they would never put their credit card details on a website. They were like,
no, no, that's dangerous that I'm going to get my card cloned. I don't want to lose all my money.
So there's no way I'm putting my money into this catchy website. Now people will put their cards
anywhere. And that is what I think needs to happen. The same thing with agents. We need to slowly add guardrails, slowly have security issues where people lose a lot of money and that's going to happen at some point. And we're going to learn from that. And once we earn people's trust, then people are going to start giving cards to agents a lot more and giving them access to their money as well. This has happened with everything, right?
Whenever you're delegating your own money or access to your money.
So yeah, that's kind of what I think is going to start happening.
So to me, stablecoins are the best way for agents to exchange value 24-7, always online, global access, international, really small fees, and instant settlements
as well.
But then credit cards have built crazy amounts of infrastructure.
They are recognized by humans, they are trusted by humans, and that is the hardest thing to
accomplish.
Anyone can build the best protocol ever, the most optimal, whatever,
but then getting the trust of billions or millions of people,
it's not really easy.
That is the hard part of option.
Yeah.
Definitely.
Oh man, there's so much to unpack here.
That was awesome.
I wish I could just clip that and replay it over.
I want to start at the highest level and then we'll kind of work
our way down because there's just so much good stuff there. So as far as kind of your vision of,
you know, hey, payments aren't the only thing for blockchain. I agree. The part that's always been,
I think, I'm a developer. I like building tools for developers. I always have. And so I'm not always the best person to comment on the consumer-facing side of things. But the a, I was talking with someone, they were like, hey, you know, can't we use an agent to go make some, you know, options trades on blah, blah, blah.
I'm like, yeah, we can.
Like, I have, you know, here, let me go, you know, create a small little testing algorithm, you know, a paper trade it.
And I've spent a couple hours on a Saturday, had a little fun.
And then I'm like, all right, well, here's the problem.
I can't get access to an API to actually facilitate the trades.
They're like, what do you mean?
I'm like, I've applied to all of them.
I didn't even get a response back.
They're just like, yeah, we'll let you know if we let you in.
Never got a response.
And they're like, well, what else can we do?
I'm like, we could throw it on hyperliquid.
They're like, what's what's that well it's basically
like well go trade them right it's just crypto but they have you know some top etfs and stocks
and things like that but like okay it sounds kind of cool well you kind of know how the story goes
right but it's such a it's such a common thing in wanting to build anything.
If you want to build on any application today, not everyone has an API.
Not everyone has open access to their product.
And what a lot of people don't realize is our contracts are essentially the equivalent of what we call an API.
like an API. There are functions that you can quite literally tie into for almost every protocol,
There are functions that you can quite literally tie into for almost every protocol.
right? That's why everyone's built on top of core primitives like an AMM or a lending protocol,
flashlones, things like that. They're functionally APIs, right? They're application programming
interfaces that tell you, hey, here's how to consume me. And so I fully agree. I think
as we continue to see it grow and we continue to see the traditional finance world look at blockchain more and more, I think we start to get to this exciting part where the word crypto leaves a bad taste in a lot of people's mouths and probably for good reason. I'm sure everyone up here has gotten hit more than once with bad things.
And most people in the crypto space have.
And that's why I've always kind of, I mean, for better or worse,
I've always said like, hey, crypto's kind of its own little use case.
I look at it more as like the blockchain industry.
There's more opportunities for this immutable public ledger than just a single token.
I know crypto encompasses more than just tokens, but I feel like that always seems to resonate well with people.
And yeah, I'll pause there for a second, so you've unmuted yourself.
No, no, sorry. I was just testing a thing.
But yeah, no, I completely agree.
And that's the problem is crypto has a horrible, horrible image that I think it's going to be hard to recover.
And it's only going to come back by bigger companies using blockchain without telling people that they're using it.
And we're going to be the unsung heroes of the history.
And it's going to be not what we want it to be, but it's going to be what we kind of, yeah, it is.
It's just, I'm seeing it with CrossFit.
Like we sell to traditional big companies, like remittances companies,
payroll companies that are moving money around like in the billions and billions.
And they're like, actually, we just found out that if we use stablecoins for all this,
we have installed settlement, we don't need to move any around and it's much cheaper for us. It's 24 seven. So like, how do we do it? And it's non-custodial,
meaning we can explore business in countries that we wouldn't have been able to, or we would have
had to make an insane amount of investment to create the legal structure in there that now we
don't have to, and we can explore those markets and go international. So it's hard.
It's hard to explain to me. I've had this conversation with friends constantly. They
are like, well, but what is the number one use case for crypto? And I'm like, there's a lot.
The thing is they are boring to you because you're not even close to the problem. And this also
happens with agentic payments. A lot of people have been asking
me as well, like, why do I want an agent to pay for things? I don't get it. What's the
problem of me just going and filling out the checkout myself? And I'm like, there's nothing
wrong with it. The problem is because you don't have the ability to... you cannot give
the ability to them, you don't see the use cases it's like
if the car was not a thing you wouldn't even think to go 500 kilometers away from your house
right like you would only think about the plans that you can reach walking or with a horse but
once you can you once you get a car once you get a plane you're like oh wait i could actually you
know like what if i just fly over the whole ocean and go to another continent? Like what's going to be there? Like whole new possibilities. So that is what I think
it's happening with the authentic payments. And what at least I've been struggling a little bit
is to share this vision with people. It's like, we need to create demos and show them a video of an
agent doing something because if not, they just don't see it.
And I don't blame them. It's not easy to picture. We're putting together five different,
very futuristic technologies and trying to come up with a use case that makes sense for people. But
at some point, the community needs to take over and people need to experiment on their own. So
if you're listening to this, this is your call to just go on Twitter, find all the toolings that are around and start
playing with them and try to understand why they're building that tooling. Why are even
people building emails for agents? Why? Well, there is a very big reason. If you go on the
internet today, a lot of things require an email, most like 90% of every account that you need to create.
So this doesn't mean that emails for agents are going to be a thing in 10 years, because if the agentic narrative succeeds,
then you would not have your agent create an account on Uber, you would have Uber prepare the APIs and do the open,
like the public access to their app that you just said, Sawyer. They would open APIs that are
pay-per-use, maybe via X402, maybe MPP, maybe X402, MPP, and also credit cards, all of them together,
credit cards, all of them together so that your agent can order an Uber.
Right. And yeah, this is, I don't want to go into that rifle, but maybe I do want.
Yeah, this is a great. Okay. So this is the shift I think we're going to see.
So just one question for you, Sawyer. How tired are you of applications like in your phone
apps like how many apps do you have in your phone how many apps do you have in your computer now
um i would guess more than 25 at least i'm gonna go with over 100. i'm notorious for downloading
things and then never using them because i wind up a i'm definitely
i'm a creature of habit but also i'm very much so like i like the simplification so i'm more and more
definitely finding myself like i've got on my phone i don't know i've got i've got like 15 pages of
apps and i can guarantee you i use less than three of them. On my computer, I've got a lot, but obviously like engineers, so build a lot of random stuff.
But I find myself more and more these days using agentic tools that aggregate behind the scenes,
right? Natural language to it. And I think that's the direction. So I'll pause there.
Perfect. So that is one point, right? I hate apps.
I'm tired of apps, right?
I'm tired of them.
I don't want to download any more apps.
If you're building a web app for the little thing that you want to solve,
I'm happier because I just need to go somewhere, do it, and forget about it.
Number two is, it has happened to me that I'm typing something really long,
like a message or something.
And at some point, I'm like, what am I even doing?
I'm like a monkey with these two little fingers just tapping constantly.
And I'm like, this is not optimal at all.
My brain already has a whole text and I need to be typing every letter.
This is just horrible.
And I've tried different ways of typing, the whole swipe thing through the keyboard
and it kind of works, but it's not great. So that is the second thing. So the third thing is why people say
the agent tech infrastructure or agents kill apps is because imagine you would have your phone with
an operating system that is natively integrated with the way agents
work and you would just have one terminal where you can send the voice note or just type and just
indicate your intention. So with me telling my phone I need a car to go from my house to my
mother's house, it should already know what it needs to do. It's not that hard to like, it's not even automation. It's just intention. It's like, you know what? Find me a car. I'll pay
the things that I like, the form of payments that I already have saved in my phone. And
just, I just need to approve it. I just need to see the price of it, how long it will take,
and then just approve it. I don't care if it's Uber, if it's Lyft,
if we have D.V. here, there's 100 million apps. So what I think is going to happen, and hopefully,
I don't know, whoever out there is building this, Apple or maybe Microsoft or whoever it is,
we need to start integrating agents into our operating systems. And you see that with Cloud. Cloud has Cloud
Co-Work now, so you can have an app that goes on top of your operating system
and you can manage apps now and do different things. That is not optimal.
That's just a patch to how we're using operating systems today. So I, and this is
all personal opinion and what I think is gonna happen, but I really see the
future where you would have your phone with just one terminal like a Siri or a Gemini or whatever it
is, where you would talk to it with intention and it would know how to get those apps between quotes
that would be behind the scenes working. And the nice thing about that is that agents are so optimal
that Uber would have to compete with every other car provider out there because your agent is going to look for the cheapest and the most relevant option.
So it would know with time, it could learn what your preferences are.
I'll stop there. This is like a lot, but yeah.
I love it. I don't know. Maybe Mr. Ben Davey remembers the episode we talked about this,
but I think I've mentioned this at least once on here. Something very similar, which is
Apple and Google have the best chance of making this happen for us, right? Because
unless you are a person who works on your computer, you spend more time on your phone now than anything else.
It's kind of sad. Humanity has gotten to this point where that whole, like, if you've seen that thing of, like, the caveman growth, you know, it's like you went from, like, monkey, chimpanzee, big gorilla, human, and now it's, like, human like this, right?
big gorilla human and now it's like human like this right um and it's very true though right like
sometimes that like gets the end of the day and i like sit down and i'm just like wow i spent the
entire day looking at a screen right it's like phone to computer to car screen to phone to
computer to television to whatever it's just it's just non-stop
um but it's it's such a good proof point that you're saying of well who kind of how do we truly
make this intuitive right and the part that chat gpt and open ai brought to the world that still i
think is ultimately underrated is just the
ability to talk in natural language.
It's such a big unlock that I still don't believe people really comprehend how much
of a historical change this is going to create for the world.
I mean, I'm a great example, right?
Like a year ago, I was not using Agenta coding.
A year ago, I was using Stack Overflow still.
A year ago, I was using all the things I used to use, right?
Google, whatever.
And these days, you know, I've now,
I historically never paid for anything.
And now I've got subscriptions left and right for all these tools, because the reality is, I am way better than I used to be.
I can process and do 10 times plus more than I used to, because it's just like, if I need something, I go delegate it out, go back, grab the information, plug it into the next tool.
And how we've worked and lived in tech has changed and the reality though
is those are for these very precise flows right those are for coding research writing content
creation etc we haven't gotten to the point where the average human can pick up their phone and be
like hey mom i'm coming over and your phone is just well, you don't own a car, so I'm going to order you a car. And boom, now there's a Waymo or an Uber or a Lyft or a Bolt sitting in front of your residence or wherever you are.
heads up, I just ordered you one, just double checking, like, are you going anywhere else
first? And if it does, it changes it for you. But it's such a deep conversation because
to a lot of people, it sounds like, hey, you're trying to rebuild how the world works.
But the reality is you're not. What you're actually trying to rebuild is how humans interact
with technology. And by doing this, we actually take a step, in my opinion, back toward being human beings, which is great.
The problem today is everything requires so much attention that you're forced to look at your phone all the time.
You're forced to be online all the time because if you're not and you live in the tech world, you're behind.
And if you don't live in the tech world, you have to use it to do everything.
You want to pay your bills.
You want to do whatever.
Like you're looking at a phone now.
There's really no choice. You want to get paper statements from your bank guess what five bucks a a month they're charging you to get your information where you don't have to stare
at the screen it's kind of wild right um and so i actually see this as a very net positive for the
world where by improving our use of natural language tooling and being able to just be like hey fed it like
let's go grab a beer after right and then it just like sends you a text it's like if my agent pings
your agent and says hey can you securely tell me approximately where you're at i don't know exactly
where you're at for security like hey like what area are you in and your thing says oh i'm in
curious like hey like what area are you in and your thing says oh i'm in the south side of new
york whatever and i'm like well i'm in san francisco right uh maybe we should meet in chicago and like
next thing you know our agents are like hey we're booking you plane tickets to go meet fedeg or a
beer in chicago okay cool like i just look at my phone just like we show up we have beer it's cool
right like that i know is obviously like a really weird example,
but the part that I really love here is everything you and I have both ranted
about here in the past like 15 to 18 minutes has one thing in common.
We did not mention stable coins.
We did not mention how we were going to pay.
We did not mention crypto. We did not mention how we were going to pay. We did not mention crypto. We did
not mention agenda wallets. It proves that these are the rails that the world can run on. The same
way the world runs on credit cards and banking and all this stuff, which I know you mentioned
that briefly, it's just a better way to move money. It's just an easier way to move money for a computer.
And by making it easier for the computer to do it, it makes it more powerful.
It's not a computer that can actually do it, right?
Anybody trying to make a swap on an AMM and figure out how to manipulate your own slippage,
I promise you, you won't be able to figure it out.
It's really trick and complicated.
But if you ask ChatwPT or Quad or any of your favorite tools to figure it out. It's really trick and complicated. But if you ask ChatWT or
Quad or any of your favorite tools to do it for you, not guaranteed, like double check it, but
it'll probably be like 98% accurate with a decent model because it's just code. It knows how to
read and comprehend a specification of information. And that's all APIs are.
So going full circle here, if we open up APIs to things and we make agentic consumption the default for a human, we can actually improve our own lives, which is kind of cool.
We don't have to stare at the screen so long.
um which the cool the crazy part about that it then brings us into a whole different topic
which is well if uber opens up their api to everyone how do we know which agents are good
how do we know buyers are legit how do we know cars are legit which now gets us to this next
part of well this is where blockchain becomes even potentially more of an able as this confidential
technology we need these things to not just be available.
They can't be controlled by necessarily one entity. We need the ability to have many different forms of
verification, reputation, feedback, and provability for agents and agendic systems, stable coins,
pretty much every piece of information, but in a private way, because, well,
your pretty much every piece of information, but in a private way because, well,
that's how these things are going to share information. Today, if I go use a service
credit card, you know, I've got XYZ type of credit card, a Visa, a MasterCard, an AMX, whatever,
they've done their lightweight KYC on me. They verified my social security number, they know my credit's good,
and they can say, well, he's not a risk that we should decline this charge. And so it goes through
all, I know you work at a payments company, so you probably know there's a ridiculous amount of
checks that these things go through automatically every payment. If we're using a virtual card that
has no history, we no longer have that guarantee.
And so as we start to get more toward this future, it also opens up the door of
not just, Hey, we need better rails to move money.
We also need all the peripheral information that you want to talk about
people not thinking about.
They're not thinking about how banks share information about the payment.
They're not thinking about the switch behind the scenes.
That stuff isn't sexy.
That's not fun.
But isn't that critical toward making this a reality?
It is.
And I want to make a clarifying point on the cold cards thing
because there is a very tricky, but really nice detail to it.
So we're working right now with Visa's intelligence commerce.
So what Visa does, for people that don't know how Visa works, Visa does not move money around.
They just know how much you are spending and how much money you're owing certain merchants.
And that's how they just move information around. They manage tons of information around
the world on how much money each person owns each other. And then what you mentioned as
well, and if a transaction should be approved and stuff like that goes through Visa. A lot
of the times it's your own bank that tells the merchant, hey, this person has enough
credit, enough money to pay for this, we can approve it or not. Depending on the things
that you're buying, depending
on the amount, on the place of the world you're at, different regulations play with you.
So for example, with Lobster Cash, because we're using Visa's intelligence commerce,
what we're doing is using a lot of Visa's APIs to use your physical card right now.
So let's say you currently have a Visa card, you would input your details through a secure enclave, you would save those card details and tokenize it, not tokenize it in the crypto side, but like kind of a token that you can then use to retrieve a Visa virtual card. So what happens with this Visa's platform, because Visa is already integrated in a bunch
of places and they hold their security, what you can do is use your own card, store it securely
on a file so that the agent cannot access that. But what the agent can do is generate a virtual
card with your approval, with a max spending limit and with a specific thing in mind.
So for example, what you would do,
like the kind of flow that would happen in Lobster Cash,
for example, is you would put your one time,
upload your Visa credit card.
It would be, it's not going to be in the cloud.
It's not that your details are going to be anywhere,
but you're going to be tokenizing those details with Visa's
APIs. And then whenever your agent needs to spend on something or you want to give your agent some
money, you would just create a virtual card for that specific use case. So let's say you have an
agent that's running your marketing strategy and you want it to be paying ads on its own.
So then you can assign a $200 Visa virtual card for it.
Mastercard is doing a very similar approach as well.
So we're going to be working with them as well on that.
And again, all of these are very proprietary
to each of these companies.
And you would say centralized
because there is a ton of regulation in all of this.
And it's not an easy problem to solve, as you just said.
What happens with stablecoins?
That is credit cards, right?
So just to clarify, we've mentioned also credit said. What happens with stablecoins? That is credit cards, right? So
just to clarify, we've mentioned also credit cards that are funded with stablecoins. So there's two
different things. One is traditional credit cards that you may have from your bank. Then you can
tokenize it, like generate virtual cards from those. But all of those spendings that your agent is going
to be doing are tied to you and your KYC and your bank. That is why you don't need to do KYC if you go through this flow. That is why we're choosing that
for LobsterGash. No KYC, but you are linked to your card and your agent spending is your
kind of responsibility. We'll see what happens, right? If your agent goes and does something
illegal online, I have no idea what's's gonna happen on a court or what's
what's the implications of that but you have lots of conversations about that with a lot of
different people um uh like on the on this show on our other show the consensus for that is it is a
hundred percent your problem um yeah that is the not that is not legal advice or any sort of advice whatsoever, just so we're clear here.
That is one man's opinion. That is not the opinion of Scalapse or the Scalapse system.
Yes, 100%. Because none of us know and cannot know what could happen in court. We have no idea, this is all new, so again,
please just be very careful. Again, at least for me, I am okay with giving my agent a card
and making spends and whatever, and I'm okay with seeing what happens because I'm an early
adopter and I love that. So we have the traditional credit cards. Then we have virtual cards that
you may have seen like RAINN or YREX or other cards that you mentioned. This confusion between debit cards and credit cards. They work as debit cards because they extract your stable coins or whatever it is in the moment you make the purchase, but they are actually credit cards. So the tricky part there is that they work as credit cards online,
but they in the back end, they work as debit cards because they take your money instantly.
That is kind of taking the money to pay for the credit later. It's not like it's getting
instantly settled. So that is kind of another category. But in that front, you do need to do
KYC because they don't have your details, you're going to be spending money and it's an actual card. So you do need KYC for those
kind of cards. And if they're not getting your KYC, that company is probably closing
soon. And we've seen this a million times because with Crossman, we have on-ramps, we
do a bunch of things with payments. And a lot of clients sometimes come and say like,
hey, look at that one. They're creating cards with no KYC, this, this and that.
And we're like, well, let's see if that lasts.
Because you can play and do things for a certain time until compliance gets to you.
And then the third part is table coins, which is the most, I think, controversial here.
Because what if you create a non-custodial wallet for your agent?
Only an agent has access
to those keys, you send it some money, and then it goes spend money around.
Like through X402, for example.
That is a new thing.
It's an agent, like an economy player that is not linked to any person or any KYC, and
there's no way to do that.
At that point, I really don't know what's going to be the legality of it.
Yeah, for sure. I think this is where the AI crypto crossover is something. We've heard a lot of people say, oh, it's just smoke a mirror, it's not real. But to me, it feels very valuable because
it's not real but to me it feels very valuable because um i'm not condoning that stable coins
today don't have the same level of compliance as a credit card my personal opinion is that
businesses at the end of the day and at the outset are going to have to treat them all the same way
uh we're already at the point where most stable coins, if they follow the Genius requirements
or the MECAR requirements and whatever, they have to be considered legal tender with similar
holding requirements from banks and all that. Lots of new regulations, which is exciting.
It's actually what we've been asking for for years. And as part of that, I feel like it's as simple as, well, if I'm a business and, you know, this is maybe the whole
MPP strike tempo play here is I want to be compliant. This is my life. It's my business.
And so I'm taking payment in a way where someone's doing this for me, right? They're helping me with
know my agent, know your customer, know your business. They're helping me with fraud monitoring and transaction compliance and OFAC sanctions and blah, blah, blah.
The list goes on and on.
Again, compliance is massive.
I think Stripe PCI compliance is like 270 rules or something.
There's a reason people pay the 3% fee.
You don't want to do that yourself.
It's way more expensive.
And so I feel like we're at a nice inflection point here where
we're going to see probably two types of things occur in parallel. One, we'll probably see the
hardcore privacy freedom maxi people who are entitled to their own opinions say, I don't
want my stable coins KYC. I don't want any of this stuff to be tracked or done.
And they're going to go do what they've always done, which is experiment and try to avoid any sort of regulation.
And again, they're entitled to their own opinion. We're not condoning using that stuff.
What I think is the stuff that will be adopted is the end to end.
This actually works in the real world, right?
Which is I'm giving my agent
a payment method i'm giving my agent the ability to consume this resource my agent has an identity
that is somehow linked back to me i'm using models and providers and inference that is good enough
that has gargreens it is know, there's prompt injection protection.
There is, you know, by default, it doesn't have access.
You know, we love to say with like X402,
oh, we can give our agent access to 15,000 resources.
In reality, you probably don't want that out of the gate.
You actually probably, what you want is in the first go-round,
your agent says, hey, I'd like to use this.
Here's my recommendation. This is like agent says hey i'd like to use this here's my recommendation
this is like a nine out of ten safe to use and here's why can you please approve it for future
use right and then again like i think there'll be a whole group of companies that pop up and
probably do this i think seawth is probably the closest right now but again like people and maybe
again your heart or core crypto crowd are going to say, I don't want that.
That's against my beliefs.
And again, I get it.
I respect their opinions. But if you actually want to give this to a human being, you kind of need this end-to-end.
Everything is checked.
Everything is tracked.
Everything is recorded.
Everything is compliant.
Because that's how you get real adoption.
That's how you get a business to use it.
That's how you get the payment companies to adopt.
And I'm pretty excited because I feel like we are actually seeing that happen in real time.
I think Stripe's entry to the game was a very good timing.
Coinbase doesn't get enough credit for this.
If you use the Coinbase facilitator on Base and Polygon and Solana,
they actually do know your agent and transaction monitoring and a lot of that stuff for you.
They don't get enough credit for it.
But again, I think, you know,
obviously Stripe exists for basically that reason.
And so the whole that being involved is,
in my opinion, very exciting.
And I think teams like you guys are also doing a great job
because you're bringing these as more direct primitives
that aren't custodial, right?
So they can be used by anybody, but they're still following in, from what I can tell,
you know, these best practices, right?
So, hey, you can actually introduce compliant agent wallets, compliant on-rames,
things like that to get these things moving, but you're not just kind of floating in limbo
saying, hey, I'm not doing the right things. Yes, those are great points. Someone asked me the other day in a group chat, they were asking,
I don't get it. And this is a very developer mindset, right? You understand, like Sawyer,
you must really understand how developers think. And they were like, I don't get it. Why did they create MPP
when they could have used X402, you know, like Coinbase is one. I'm like, business bro. Like
they want to own, it's like Apple, they want to own the whole stack. They want to have, you know,
every end to end part of the process that your agent's going to pay, they want to own it and they want
to extract as much value from it as possible. And with that, if you're building the whole end-to-end
stack and the whole protocol, you can provide a better experience, better developer experience,
better user experience, lower fees as well, because you're owning the whole supply chain,
right? That's the only reason.
It's not because MPP is more optimal than X402.
It doesn't mean it's better.
It's not, it's just a different approach.
And again, as has always happened in tech,
you can go and look for patents, IEEE standards
and protocols, frameworks that are out there
that solve the same problems.
Some were more
adopted than others and that's just history and how well they solved the problems for people and
how people adopted them. So that's why we'll see what happens. Right? Yeah. Totally. No, I agree.
I mean, we call it going vertical, right? Like at the end of the day, like the best companies own
the vertical stack of, you know, if you look at the computing companies, right?
They own the hardware, the software, the coordination, the bandwidth, the data centers, the software on top of that.
It's just never ending.
And I think they've done a great job.
I mean, I will always, especially for developer tools, I'll give credit for credit to you.
The WEVM team is one of the best i've ever seen at building developer tools which is only
ironic because the only team that i may have given a compliment to before that is the stripe team the
strike team has some of the best docs apis developer tooling in existence anywhere and so
the two of them together is just incredibly dangerous.
X402 has been awesome.
I mean, we're still very, very excited
to continue to support and see the growth there.
But I mean, I will say like getting MPP to work,
it took me a fraction of time.
The developer experience is just top notch.
And so I think to your point-
It requires like the problem with that
is that it's not open and it's not like you can integrate let's pause there let's pause there
though because that's not that's not true that's not true mpp mpp is a spec okay and so just like
x402 mdp is a spec the part where everyone's getting hung up, and I think this is incorrect, is they came out with the vertical scheme for them
because it's in their best interest to do so.
However, and I'm happy to talk to you more about this,
I'm actively talking with the Tempo team
and a couple others,
and we're actually in process of writing an EVM general spec
because MPP is an open protocol.
It's just no different than how X402 is functionally backed by Coinbase.
This one's just backed by Stripe and Tempo.
But it is an open protocol.
You can build anything you want on top of it.
The part where I think everyone got a little bit like myself included, maybe got a little bit confused out of the gate was their charge for
tempo is 100 built for them and the charge for strike is 100 built for strike but you guys can
build the same for crossmit you can build we're building you know we're building a generic one
for the evm at the end of the day a spec just helps us standardize it makes adoption easier so that
right now what's happening is I'm talking with
a dozen teams and everyone's writing their own EVM SDK. The problem is that means we're going to
have 12 different EVM SDKs, right? Whereas if we create a standard on top of their NPP stack,
we all wind up with the same thing. And so what they came out with, yes, is more geared toward them, but the
core spec at the end of the day is just basically a, here's the headers, here's the expected body,
here's the requirement. And that can be extended in a very open way. Now, as long as they allow
these types of specs to come in and be put in as part of the repo from third-party contributors,
we have no problem.
So far, like I said, I've been talking with them.
They seem very open.
I, you know, fingers crossed we'll be all good.
Obviously if they don't, different story.
And then you maybe wind up with more of your like,
you know, open closed game.
But I think we hate, it's kind of like, you know,
innocent until proven guilty.
They haven't done anything yet to tell somebody like,
hey, they're not going to.
So I'm going to err on the side of it.
These are all great opportunities to help agents make payments and help humans make
pay-per-use payments.
And I think as the ecosystem pushing these forward, it's just the most important thing
is we try everything, right?
We shouldn't just say, no, we don't want to touch it because it looks closed out of the gate.
We should say, hey, it actually looks good.
Let's try to make it as close to how X402 works as possible from a scheme perspective.
And now everyone can use both, right?
Maybe an agent prefers one over the other because their owner does.
It's kind of like Mastercard visa right yeah you go you ever go to a restaurant and they say hey yeah sorry
we don't take amex happen to me happen to be a lot in europe um yeah they're like we don't take
amex like well this one's all right yeah i'm i we are very very over guys. And I'm cautious of, of Fedez time, but dude, please come back.
We would love to continue this conversation.
You know, that was, that was great.
Before we, before we say our goodbyes, anything, anything you want to just throw out there, Fedez?
I don't know.
Just thank you so much for the invite.
This was an interesting conversation.
I think we could be talking for hours.
So yeah, thankfully we have a call tomorrow also to discuss more stuff with you guys.
So I appreciate the invite.
And just for everyone listening, if you're interested in agentic payments and wallets
infrastructure, cross-border payments, unwraps, etc., just
feel free to reach out to Crosby in sales or myself. I'm happy to chat about these things
a lot more and to know what you guys are building because we're constantly looking for builders
pushing what's possible and testing these frameworks out. And it really helps me know
your opinion on MPP, on X402, on your experience in different toolings.
And if you had any issues with our staff
or if you have any recommendations,
I love being in the ground with all of you.
So yeah, just thank you for the invite
and inviting everyone to just try out stuff
and share it with us.
100%, 100%.
Soya?
Couldn't have said it any better. Come build with us there's a lot of stuff
that works on scale based from crossfit they're embedded wallets um if you need other stuff just
reach out again we're here to support developers so don't hesitate to ask and um yeah this is
awesome so much good stuff um excited to continue to explore and let's build the future. 100%. Guys, thank you so
much for joining me and thank you to everyone for tuning in to another episode of Scale School.
Join us next week where we might be talking about something special.
Thank you. Goodbye.
