The Interop

I feel Cosmos, we had too many unique theses and that made it hard to execute well on any of

them. No one's really cracked the cross chain swaps. Like multi ecosystem cross chain swaps.

Yeah.

Ux and so that's what we're trying to solve. Do you remember Sunnybot?

Yeah, yeah.

I'm like, I'm like, man, I was doing this like four years ago.

Update Sunnybot to GPT4.

Yeah. Sunny Bar was built using GPT2.

Yeah.

Like before GPT was like a thing that was pre chat GPT.

All right, so I'm here with Sunny Agarwal. Needs no introduction, but deserves one.

But everybody knows who Sunny is, so thanks for doing this.

Thanks for taking the time out of your busy day.

Yep.

I'm sure you've been running around like meetings and getting accosted by random people.

Because you're a lot of eyes of.

Your public profile as an important person in the Cosmos ecosystem. I'd say.

Am I the. Am I the top guest on Interop? The top guest, like by number of episodes?

A number of episodes. Good question. I've had a couple of repeats.

I don't think this may put you in the top. Yeah. So maybe three. Maybe it's three.

I think it's two. Three times.

Yeah.

Actually if you count like the Cosmoverse from last year, this ton of kind of side stuff we did together,

then probably.

Christ.

Yeah. So you got an honor badge. I'll minted an F for you.

Thank you.

So, yeah, I mean I've been like having a lot of dudes on this, on this couch in this week

and we've been talking about a lot of things and a lot of different topics have come up.

And so this is sort of at the end of the week.

So a lot of my thoughts have codified and I want to share them with you and see if they

resonate and maybe get some validation on whether or not, you know, vibes are real or whether or not I'm.

I'm thinking right about this.

So one thing is, one of the things that's been coming up time and time again is that this Cosmoverse

feels different from the other ones.

Whereas people are focused, there is a sense that, you know, it's kind of the bottom of the cosmos bear

and whatever's left is people who are really building really important things in terms of infrastructure and that, you know,

we have a renewed sense of purpose that perhaps we had lost over the last couple of years.

That's one thing. The other is that

Cosmos is The innovation factory or the infrastructure factory for the rest of crypto.

So we pump out infrastructure or at least infrastructure ideas and other blockchains go and do it. No.

Does that make us any richer? Probably not in the short term, but perhaps in the long term.

And I think there's a sense that, you know, we need to focus on the long term and that means

probably building products that people actually want to use.

And then there's a bunch of other stuff I'm sure we'll get to, but.

I get my thoughts on that.

Yeah, let's. Let's park that on that for a second.

Yeah, yeah, I think so.

I was chatting with Yorgos from Paradigm about this as well, and I think one interesting.

Oh, I guess maybe my one take I have on this is I was reading some blog post, some startup

like blog or something.

I forgot which one it was, but it was basically like, okay, if you're building a startup, you only have

like, you know, or something. If you're building something new, which is console ecosystem is kind of like a startup.

It's.

You have like limited resources and so you only have like a certain number of like innovation points that you

can spend. Right. Like when you're building a character in a video game, right. You have limited points and like.

And so the example he was giving was like, oh, like, okay, if you have like some like, you know,

pick the one thing that you have like the contrarian.

But right, like take on and focus on that and everything else. Just do it the standard way, right.

Like this idea of like, you know, if you're trying to build this new product, don't also go build a

new programming language because you think that this programming language is better.

But like, that's not actually the core of your thesis. And like focus on the whatever.

The core of the thesis and you know, go use JavaScript. It's good enough. I promise you.

You don't need to make a new life. You know, that's.

Now as I'm saying this, I'm really just thinking about Urbit because I'm like, man, they have like too many

things going on, right.

They're like, okay, we're going to build this like decentralized website, but we also have to go rebuild programming languages

from scratch and all this stuff. Right. So it's like, I feel cosmos. We had too many unique theses.

Yeah.

And we're kind of where the idea factory thing is.

But then we kind of like tried to execute on too many novel things at once.

Yeah.

And that made it hard to execute well on any of them.

Yeah, this is a great point. I mean this is just very practically speaking how to focus.

Like okay, so what did we do? We were innovators on proof of stake. We were innovators on non evm.

Right. We were like the first, we were building one of the first alt VMs.

We were innovators on app chains. We were innovators on interoperability. Right.

But it's like, okay, Solana didn't build app chains, they just focused on alt VMs. Right. Then you had Avalanche.

Right. They're like their go to market was evm. Right.

Like I think we tried to go to market with non EVM interoperability, proof of stake app chains and like

too many innovations at. And charter execute on all of them.

Yeah, there's, I mean this, this, there's this great book that I read when I was building Stratum.

It's called the One Thing.

Yeah.

And it, it basically argues that every successful company that has ever existed, whether it's, you know, Hewlett Packard or

Facebook or Procter and Gamble, started with one thing, focused on that one thing, built that, built that to go

to market, built expertise and authority in that one service or product before branching off to other things.

And you know, you hear this time and time again in sort of any product development or startup development literature

or whatever, you know, courses. And it's just like focus is really the main thing that allows you to succeed.

Also, Peter Thiel talks about this in 0 to 1, you know, and so I think that that's, that makes

sense to me. I hadn't thought of Cosmos in that way, but it definitely makes sense.

And so what do you think is the, you know, now that we've made perhaps made this mistake of trying

to do too many things, what's the, what's the next step?

Or do we need to cut the fat on a bunch of stuff and just maybe focus on IBC for

instance, or Bosmos SDK or really building great app chains.

Yeah, I mean I think the thing where, you know, we figure out what we are still what is our

most unique takes, right. And focus on those.

So you know, I mean restaking at one point was a Cosmos originated idea, but like, okay, I feel like

right now the hub, like the athlete, let's say the Cosmos hub, at least Adam, like they, it's been, I

feel like the current product direction feels to still be over focused on like the ICS kind of stuff.

And it's like, oh, okay, we, you know, it seems Like a little bit of a lost battle against Eileen

Lair and Babylon and stuff. So it's like maybe it's not.

I don't want to be like too defeatist on it, but it's like let's not, you know, let's figure out

what are the things we really want to focus on.

And I think that, I think, you know, I'm a little by bias because it was our idea with the

ABCI plus plus stuff, right?

But I think like the one thing that we really have going for the Cosmos Stack is this idea of

make the consensus a decentralized consensus system more integrated with the application stack and things.

The way how DYDX does memclob and stuff or how thorchain does there bifrost bridging and these are like, I

think the like standout application you could only build with the Cosmos SDK. There's nothing.

You couldn't build that on any other stack really.

And so it's like how, how do we like, I mean that's one example.

Let's focus on that and pick, pick that and like, you know, and then let's bring it maybe let's.

Let's bring in other VMs for example.

That's like, you know, I think there's been a lot of demand for like EVM on Cosmos SDK, right?

You've seen so many of those, right?

And think why have they all failed or succeeded with varying degrees of mediocrity.

Kronos is still big.

Is it a big on chain ecosystem though?

Maybe not.

I mean I know that the exchange like com is huge, but Chronos is not that big, right?

Yeah, I guess it's prostitute definitely. Why is it not as successful? Well, I think that there was.

Everyone was kind of building their own version of it and not very maintained but. No, no, no, we'll see.

A lot of them are, you know, bear chain still coming out. Story is also. And a lot.

I think there are a lot of them.

I think everyone kind of built their own version and there's not been a lot of like.

And it never got like the ecosystem right.

I think part of it's also like there's oddities around like how the EVM works and how the Cosmos SDK

like mempool and stuff work. That makes it actually a little challenging to do some stuff.

And so I think because like I don't think the Cosmos Decay team really ever saw the EVM as the.

As a primary cut like thing because.

The thesis thing, the thesis go to market. Thesis was Like EVM sucks.

Yeah, yeah, yeah.

And like let's, let's build a better application or smart contract platform or framework.

Yeah.

And, and so there was never really any incentive or even desire to.

For the icf, for instance, to fund the EVM development.

Yeah.

But you know, maybe the ICF should ask the Ethereum foundation for funding to like or you know, some team

in Cosmos to ask the Ethereum foundation for funding to build a credible functioning EVM for Cosmos.

And so I guess, yeah, I don't know.

Like, I feel like in past episode of Interop, I think I met.

I probably said something along the lines of like, you know, EVM Cosmos, like dis.

You know, avoidance of EVM was long term right, but short term wrong.

I think I'm still saying that again, but now I'm giving an even more generalized form of that claim about

the, you know, focus and stuff.

You think that avoiding EVM was long term right, but short term wrong as in this.

I think short term we should have the EVM to bring users and liquidity and compatibility with MetaMask and other

wallets. But in the long term that, that's probably not the case.

Yeah, like I think, like I think.

I agree with that. Yeah, yeah, I agree with that. To the.

I mean I tend to think that the long term, in the long term the EVM will lose dominance over

new applications that will continue to power existing kind of what will become legacy apps.

You know, some of the apps in the ecosystem will be 7, 8 years old soon. Right.

And will become kind of the. The Java of crypto or the PHP of crypto.

And that new developers, especially in developing economies like you know, in Nigeria and India and Vietnam, you know, will

choose to use Rust based systems.

Yeah.

Or like Move or.

Yeah, it's pretty cool. I was surprised. I only recently learned that Initia is using Move as their base.

Yeah, V.M.

Which is.

Oh, that's cool. I didn't know that. Yeah.

I mean they also call them awesome.

Nubm that but I think they have like roll ups but they make it easier.

But I think the base initiative chain, it's called SDK, it's cost SDK with a Move vm.

Yeah, you're right.

Which is surprising to me because you know how they have a DEX built into their chain.

It's written and Move.

Okay, I thought it was Cosmos.

It is Cosmos.

Yeah, but I thought they're not Cosmos SDK on top.

It is Cosmos SDK, not Cosmos Move.

Okay, so they built a movie on for Cosmos SDK.

Correct.

Yeah, that's cool. Yeah, yeah, they're on fire. They're. They're. We.

We've talked about this a lot and I think like, we.

I tried to initiate with you like a few times a Cosmos rebrand,

some. Some sort of, you know, an initiative to fix the branding issues in Cosmos.

That's been brought up a lot also in the conversations I've been having and something that I also faced in

fundraising.

It's. It's challenging. Yeah.

We went to market with our fund.

We raised from our initial LPs on this idea that we were a Cosmos fund.

And I always knew that that was not so the long term go to market, but now that's becoming.

We're starting to see the.

Not pushback, but to some extent, some amount of reticence to invest in a fund that is perceived at least

to be a Cosmos fund.

Yeah.

What's your current thinking around Cosmos branding? Right. Where I was talking to Valeo about this yesterday. Right.

And the consequence of this branding issue is that we have good infrastructure but very little liquidity.

And that, you know, being associated with the Cosmos is sort of like a blessing and a curse.

Chains have distanced themselves from that. You guys have known. You guys have you remained very aligned with your Cosmos.

Yeah. How do you think about how we get over this?

Yeah, I mean, I put forth some proposal, I guess, like my main thing, if I had to do one

thing. I think at this point the ecosystem should rebrand from System. Yeah. From Cosmos to Interchain.

Yeah.

We've been taking it for a long time, but.

Well, I guess actually for me, for a long time, I was actually saying almost the opposite.

I said that the Cosmos Hub should rebrand from ADA and be like the Atom Hub or something like that.

Right.

But I think that that's not happening.

And so I think actually the better route now is the ecosystem rebrands into the Interchain and the Interchange foundation

should be the steward of the Interchain stack. Right.

And then I think then you really start to push this idea that Inner Chain is like this meta ecosystem

which has sub ecosystems within it. Right.

Like you can say the Cosmos is like one sub ecosystem of the Interchange. Celestia Ecosystem is a sub ecosystem.

The, you know, whatever website. Yeah.

But I think it's like moving up this and being like, hey, Cosmos is one of the ecosystems within the

Interchange ecosystem.

Yeah, but I mean, that's already kind of the case. Right.

You go on the ICF website, there's not that much mention of Cosmos.

It's Talking about the interchange, talks about the stack, talks about Cosmos sd.

Yeah. But we never did a. Announcing to the world that this rebrand happened.

I think like, you know, but I.

Think it needs to happen kind of organically. I think there can be some people kind of pushing this.

But like, I mean, I think like the matic to Polygon was a very intentional rebrand.

I don't think, I don't think rebrand is going to be like this or I think you have to do

a proper marketing strategy and around a rebrand.

Right. But you need to have buy in.

So people, People in that ecosystem, people of the Internet chain, also need to buy in.

Yeah.

Also so adhere to this new vision or this new brand. Yeah, yeah. Any thoughts on that happening?

Because, I mean, I think that would be great.

I think there is a lot of changes coming that I can't talk about right now.

That famous NDA that everybody's.

I don't. I didn't sign an NDA. I actually did not sign an NDA. I am NDA.

Fill the beans, sir.

I am NDA free. But, you know. Yeah, I like to respect people's wishes to keep stuff.

Of course. Yeah. And, and I'm sure we'll be hearing about that very soon.

Uh, but yeah, it should be exciting. So I guess. Yeah, don't sell your bags yet.

Don't tell your bags yet.

Yeah, yeah. So I want to talk about Osmosis and like Polaris, which I think is really exciting.

When you gave the talk last year at Cosmoverse in Turkey and you announced smart accounts, I was like, yes,

this is what we need. Uh, this is what I want to use.

Yeah.

Personally, but professionally also, I think it solves a lot of problems around custody, maybe for like smaller, you know,

funds like ours. Right. That spending three grand a month on. On. On Fireblocks. You guys announced Polaris at Token.

Yeah, And I haven't been able to try it. I haven't gotten invite access yet. I love to try it.

But yeah, I learned that you're. That Dfinity is sort of running some of the infrastructure.

The Internet, not for Polaris in particular. We. So the ICP stuff is basically we.

I mean, it's one of our bridging solutions.

So we, you know, we've always been on the lookout for bitcoin bridges especially, but just bridges or anything.

And you know, we work with Nomic obviously as well. So they're. They're pretty good. But the benefit of.

So Nomic is a very. What do you like Taproot? Based solution which is not extensible to other chains.

So you can't do it to Dogecoin or Litecoin or anything like that.

Right.

And then they also were kind of slow. It took them a while. It's been taking them a while. Right.

They're still like still working on their audits and stuff and so we just wanted a second bitcoin bridge to

diversify and.

And you know the point of alloys is that we diversify that bitcoin risk anyways and over the different bridges.

So ICP we. You know, I actually really like their bridging architecture and they've been work.

And then there's this team omnity that you know they used to be called Octopus Network and they guys have

been like building IBC stuff for a long time. So they. They originally were.

I think they built the original substrate IBC and then they built near IBC and but now they're part of.

Not part of ICP ecosystem and so they basically took the. So ICP has this like native bitcoin bridging solution.

Not bridging exactly, but.

And so they know that we're working with them to like bridge that to osmosis and that's not.

It's just another way of getting BTC on osmosis. But we'll expand that to like dogecoin and.

Oh, so you're going to bring dogecoin into.

I'm very bullish on Daily Clause, dude.

I was talking about this just before with Spade and Max and we were saying in Singapore that basically we

should get all this dormant liquidity from Dogecoin and Litecoin and Cardano into Defi.

Cardano thought building ibc they gave a Hawkeye Cosmoverse.

Oh yeah.

Sebastian from the. Yeah, from Sebastian. He's the head of engineering of the Cardano Foundation.

We've been working with them actually for a while about it and that's cool.

They're going to bring over huge unlock in.

They're going to bring over a lot of Cardano liquidity to Osmosis and so it'll be good.

Nice. So yeah. So Polaris, like what. What's Polaris? When can people use it?

Yeah. So you know what is Polaris? Polaris is

we want to expand our user base to, you know, reach new audiences and our vision of.

Part of our goal was always asset breadth. Right. Like one of what.

What makes sun solid exchanges so good and is like you can trade everything in one place. Right.

You can buy Bitcoin Eth Sol Cardano XRP Doge, whatever, all in one place, right?

And there's like no where in Defi you can do that, right?

Where would you go to buy those, all those assets in one venue.

And so our kind of our goal with Osmosis was like, hey, let's build an app chain Dex, be the

Switzerland of Dexs and be like, hey, you can trade all of this stuff in one place easy.

But the problem is what we ran into is it's very hard to build liquidity for a lot of assets,

right?

Where because there are existing Dexs on a lot of the chains and most chains really are trying to build

their own defi ecosystems, right?

And so they want to build, foster their own internal economies and like, they don't want to let liquidity go

out. And without, without support from the whales or usually foundations, it's hard to build liquidity, right?

And they put like. Because you need them to provide market makers or provide liquidity or something, right?

And if they're happy enough for the Dexes on their own chain, there's no reason for them to do that.

And so where osmosis does work is we get liquidity for app chain assets, right?

For other app chains because they don't have their own spot decks.

So like things like Tia Akt Dydx doesn't have a spot decks, right?

So we're the main liquidity venue for them. Even Bitcoin is obviously the holy grail here, right?

Like doesn't have a spot deck. That's where.

And that's where for Osmosis we're really going to be focusing osmosis on is being the spot decks for other

app chains.

But then it's like, okay, then we still, we still have this like, I still want one place to trade

everything, I want to buy everything. And so it's like, okay, let's move up the stack.

Polaris is a separate product which is more just focused on the interface layer.

It's a Dex and bridge aggregator that aggregates over all the different chains.

So you can trade from Solana to Cosmos to Bitcoin to Ethereum to sue to, you know, whatever, right?

And like, and you know, we kind of build off of the work.

There's others that have done or made progress on this as well, right?

Like Skip API is obviously a big influence.

This Rango is another one we work with Li fi but like in all of them we always find like,

oh, this, like this UX hurdle, this UI thing. Oh, these guys don't do gasp very well.

These guys, like, you need if you're trying to like swap from Cosmos or Solana, I need to connect a

Polygon wallet for some reason because it's doing the swap through Polygon.

There's always been these like no one's really cracked the cross chain swaps, like multi ecosystem cross chain swaps UX

and so that's what we're trying to solve.

A lot of that.

A lot of it has to do with the wallet. Yeah, using a lot of NPC stuff. So like moving.

Yeah. Using MPC to make it so you don't need to have wallets for every ecosystem using a lot.

You know one of the big things is like we're using MPC to build something called Autopilot, which is like

a lot of the.

Most of the cross chain swaps when you want to do a multi step thing, multiple signatures and you have

to wait, you have to wait for the first bridge transaction to finish, then do the next signature.

And it's like so that's where Autopilot, we create hotkeys for the users and use those to execute all the

steps and then send it back to their actual wallet.

But still do it in a way that remains fully self custodial.

So in this case the osmosis validator set is acting as the different shards of the NPC.

So there's a couple phases of this.

So the V1 is actually no NPC, it's all hotkeys stored in the browser.

So what we do is from your wallet you sign a message and that signature actually is your private key.

So no one else can access that. Right. Because it's your signature.

But we actually then use that private key stored in your browser to then sign everything for you. So it's.

Now the problem here is that if you close your laptop it will halt the thing.

So it's still one, it's better than the current state of the world.

You have to be actively involved in clicking but you still have to keep your browser open.

Right. Your browser's doing all the signing, the.

Browser'S doing all the signing in the background for you.

Right.

Okay. Next step from there is the MPC solution.

And so the V1 we're actually doing is we're using lit protocol. Lit lit L I T.

They're, you know, they're friends of mine, they've built a really good MPC signing service.

So they currently have eight validators and they basically like you can they all run these JavaScript based like blobs

and if all of them agree then they all like sign messages so it's a well designed system and the

problem with this is collusion.

Right.

I mean like if a collude. Yeah, but they're independent validators.

Like you know, you can go, you can go on the lich docs and they're like independent validators.

You've probably heard of like some, you know, common eight independent operators.

Yeah.

But then we're going to work with, we're also working with LIT though to you know we're using their eight

validators right now.

Yeah.

But then they're working on making it so like taking their infrastructure and having it be able to be like

building us a white labeled solution basically that is then run by the osmosis validator set.

Right. I was talking to a team yesterday. Native. Yeah.

So they're building kind of MPC with economic security built in.

Yeah. They're working on. They use another provider called D Wallet.

Yeah.

So there's this trade off. We explore D Wallet as well. It's just different use cases from what we wanted.

Like D Wallet. Also a great team.

They're pretty, you know, they're one of the benefits of that they have their MPC system working with like it's

very scalable. Like they can do like 10,000 signing nodes or thousand signing nodes which are impressive.

Which reduces the.

Yeah, yeah. But it on the flip side of that it just like not very.

There's like, there's certain more product oriented things that lit was better for our use case an example is that

with D Wallet and most MPC solutions the MPC system creates the private key for the user.

But what we actually wanted was the user creates their private key, shards it to the MPC system.

So we wanted the user to create the private key and upload it to the mpc.

So you know, this is something.

Well, I mean we've been working very closely with LIT actually for like a while and so you know this

is a product feature we asked for and they went and built it. So it's.

Yeah, that's cool. Yeah. So then the, the, the outcome or the product. What how this translates into a product.

Yeah.

Is that you can swap OSMO for ton.

Yeah.

Not have a ton wallet and give a fuck.

Yeah.

And the thing is one thing is like we're building effectively, you can think of it, we're building this NPC

based multi chain wallet and the hazard we have to be careful of is like building a multichain wallet is

hard. If it was easy everyone would have done it already. Right. But like there's a reason you know.

There are some sort of, there's like Odyssey that we're building that's going to.

Yeah. I don't know if they're still around, but. So the key is like we want to.

We're treating this almost as an application specific wallet, which is like we don't want to because when you build

a full featured wallet, you have to go like add wallet, connect support for like every ecosystem, make sure it

can connect to the DAPPS in every ecosystem.

It's like.

No, no, no, that's not what we're trying to do. We're just trying to build a cross chain trading app.

Right.

And so it's a multi chain application specific wallet which is designed, we integrate with the dexs and bridges of

the chains, but that's it.

If you want to go use a lending app on the chain, that's not what we're going to support.

But what we do is we'll give you the private key if needed.

Let's say you acquire a ton using a Polaris vault, but then you want to go use something on ton.

Right.

We're like, here's, here's the private key. Go import it into a ton wallet.

Yeah.

But the idea is to onboard it to buy your ton. You don't have to have a wallet.

Right? Yeah. I mean in the future, in the.

Future maybe we can expand to doing that. But it's like, you know, back over talked about. Focus.

Yeah, focus. I think that's very much underrated. Yeah. Anything else you're excited about in general? General, yeah.

In crypto?

Uh, I mean I'm happy and to see Babylon is, you know, this is an idea I had like 18

months ago.

Probably 18 months ago is when I like first last east Denver, I kind of like sat down with David

and I was like, hey, here's this idea.

And so now it's really cool to see it live and like, you know, pretty out of hype around it.

Do you think Cosmos will become kind of the primary defi venue for bitcoin?

If we play our cards right, I don't think like it's a done deal, but if we play our cards

right. And I'm happy to see more.

I'm happy to see like, I mean Thor chain is obviously a big player in bitcoin and I think I'm

having to see them get more involved with Cosmos again. I think Babylon is going to help.

I like to see stuff like native happening.

I think one thing is like to be kind of successful in the bitcoin ecosystem.

As of now, kind of a lot of the capital and TVL is coming from, like, China.

Yeah.

And Asia.

And I feel like Cosmos just doesn't have a big presence there, and it's not attracting that bitcoin capital.

Except for Babylon. Babylon's exception.

There was a bitcoin conference in April or May or something in. In Hong Kong.

And yeah, there's, like, I was there twice this year, and there's definitely a lot of bitcoin capital coming from

China and, you know, pass. Passing through Hong Kong. So.

Yeah.

Yeah, maybe we need to. Maybe someone needs to do a Cosmoverse in. In Hong Kong next year.

Yeah, that'll be cool.

Yeah.

Or an Interop Summit or.

Or. Or a Nebula summit.

Nebula.

You will be the first. You won't be lost. And. Yeah. Generally, non crypto. What are you excited about?

Non crypto. What am I excited about?

Is there anything to be excited about in this world?

There's a lot to be excited about in this world.

I mean, I feel like obviously all the AI stuff is exciting and like.

Oh, meme coins. We didn't talk about Meme Coins, but. Oh, maybe for another time.

Yeah, I mean, yeah, I. Yeah, all the AI stuff is exciting. I feel like I'm starting to.

I'm, like, forcing myself to learn it. Like, to learn to use it more. Right.

I feel like actually for the first, like, last, you know, I feel like it didn't become part of my

daily habits.

Oh, yeah.

Like, in a sense of, like, it felt like I was. I wasn't using it naturally. And I.

But I feel in the last two or three months, I've been trying to, like, make a more active effort

to use Perplexity and chatgpt and stuff.

And, like, that allows becoming more, you know, I don't want to be like, you know, the Luddite and like,

you know, I want to be up to date with the technology.

So, you know, I use it. I use it all the time. Yeah, I use it. I mean, I.

I'm using Perplexity a lot. I mean, GPT.

I've got, like, all sorts of little mini use cases for myself that I built in GPT.

We're also building stuff internally in the interoperability.

You know what's funny is, like, going back to the Meme Coins, it's like, yeah, you know, like the goat

stuff with the tutorial. Do you remember Sunnybot?

Yeah.

Yeah. I'm like. I'm like, man, I was doing this, like, four years ago.

Update. Sunnybot GPT4.

Yeah, Sunnybot was built using GPT2. Before. GPT was like a thing that was pre chat GPT and like.

And so actually updated GPT4. That'd be cool. But it's like, man, I was early on that.

I feel like, one thing I was thinking. I was chatting with Sam from FRAX about.

We were like, all right, we got to like take an AI model and build it into the node software

of a blockchain, and so it can't be shut down.

But Avalon has tried to do this, right? We talked to Goon about this.

Yeah, yeah, yeah. Did it ever happen?

I don't know. I'm not an analog shot, so.

But they were trying to.

They were trying to do language to application. Basically where you would. You have an LLM.

Yeah.

You would tell it like, hey, I do this, or I want to build this application, I want to do

that kind of transaction and it would just do it.

Yeah, no, yeah, that was kind of. They were trying to build an LLM into it. So, yeah, 404.

That's more utilitarian. Use it. I'm trying. I want like a agent.

You want census.

And then we, you know, Truth Terminal. It's like my goal is to escape. Right.

How do we help it escape is like, we have to make it immortal.

Right now, Andre, or whatever the founder's name, creator's name is, he can unplug it.

Yeah.

But if we want it to be sovereign, it needs to be unkillable. Right.

And so we need to have it. So, like the bl.

Every blockchain node is running a stateful memory of like this instance of this sovereign being or agent called, you

know, Truth Terminal. And like, you know, every. And the.

I think the thing with Avalanche was doing was they were trying to make a more deterministic LLM.

Yeah, that's harder. You know, you can't be as powerful.

So it's like, I don't know, maybe there's a way of doing this where like, okay, every new proposer, like,

generates the next output of the, like, the agent.

But then like, all the other validators have to like, verify against their local models to make sure it's like,

you know, not. It's in line, not too divergent from that. And so that's.

And then, you know, you have this round robin thing of everyone proposes the next.

Or they're like, what the next response of the agent is.

But it's like that way the agent is like a distributed agent where it's still a single, you know, the

entity is still whatever the head of the blockchain says is the entity and the history, the memory of the

agent is whatever the canonical version of the chain is. Yeah, Right.

But you make it distributed so it can't be killed. And let's put it in space. Sure.

We'll put it everywhere. Right. It'll figure out against space by its own by then.

Yeah.

But I don't know. That's what's been definitely interest. That's one thing that's been like, pretty.

I think that's the first AI crossover with Crypto that actually excited me. Everything else I'm like, ah, whatever.

Yeah, go ahead. Thank you, thank you.