Welcome to Zero Knowledge. I'm your host, Anna Rose. In this podcast, we will be exploring the latest in zero-knowledge research and the decentralized web, as well as new paradigms that promise to change the way we interact and transact online.

This week I chat with Zaki Manian about the Cosmos ecosystem, and I ask the question, "Is Cosmos dead?" We run through the key events in Cosmos history and map the teams and players who worked on the project during different eras. We tease out the strengths of Cosmos, but also the structural flaws that have led it to be in the state it is today. There is a key distinction, and one that I hope comes through during this episode, between the Cosmos Hub and the Cosmos dream. And one may be faltering, but I think you can tell by the end of this, the other is unlikely to ever die. We wrap up with a chat about some of the initiatives in place to try and revive the ecosystem and bring different factions in Cosmos together towards the central goal.

Now, before we kick off, I want to point you towards the ZK Jobs Board. There you can find job opportunities to work with top teams in ZK. I also want to encourage teams looking for top talent to post your jobs there as well. We've been hearing from more and more teams that they have found excellent talent through this Jobs Board. So be sure to check out the ZK Jobs Board. We've added the link in the show notes.

Also, a quick reminder, our bi-annual ZK Summit event is hosting its 12th edition in Lisbon on October 8th. Our one-day ZK-focused event is where you can learn about cutting-edge research, new ZK paradigms and products, and the math and cryptographic techniques that are giving us epic efficiency gains in the realm of ZK. The early-bird tickets are sold out, but we still have general tickets for sale over at zksummit.com. I've added the link in the show notes. Hope to see you there.

Now Tanya will share a little bit about this week's sponsors.

Today's episode is sponsored by Anoma. Intents are a new way to build dApps based on user-defined outcomes instead of virtual machine transactions. Anoma is making an intent-centric future possible by introducing a distributed intent-centric operating system. This will enable builders to create user-friendly dApps compatible with any blockchain. Use Anoma to add intents to existing dApps, or build on Anoma and settle wherever your users are. Anoma aims to vastly improve blockchain user experience, enable novel applications, and promote a more human-centric future for decentralized technology. Get ready to build with intention. Anoma is the universal intent machine, introducing a new era of applications where you define the outcomes you want. Follow Anoma on X to learn more at x.com/anoma. So thanks again, Anoma.

Aleo is a new Layer 1 blockchain that achieves the programmability of Ethereum, the privacy of Zcash, and the scalability of a rollup. Driven by a mission for a truly secure Internet, Aleo has interwoven zero-knowledge proofs into every facet of their stack, resulting in a vertically-integrated Layer 1 blockchain that's unparalleled in its approach. Aleo is ZK by design. Dive into their programming language, Leo, and see what permissionless development looks like, offering boundless opportunities for developers and innovators to build ZK apps. This is an invitation to be part of a transformational ZK journey. Dive deeper and discover more about Aleo at aleo.org.

And now here's our episode.

Today I'm here with Zaki, co-founder of Sommelier and a longstanding member of the Cosmos ecosystem. Welcome back to the show, Zaki.

It's great to be here.

Whoa.

Cosmos is a very, very old project.

Okay.

Wow.

And it does feel like we're at another inflection point. So it's a good time to catch up.

Yes.

What actually launched then? Was that like a full validator set?

Yeah. So it was the Atom Hub.

Yeah.

Okay, so there was this event called Game of Stakes, which was the first incentivized testnet, which the team executed. And sort of I led that process and sort of unleashed the incentivized testnet monster upon the world.

Which we've seen ever since.

Wow.

And that was kind of a pretty new concept in the centralized exchange world.

Was there already a PoS live? Was Tezos live?

Tezos. Tezos was live.

Okay, so it wasn't the first first, but it was like in the first five or something.

Yeah, I mean, Tezos had all of its early struggles -- all of its struggles. Right? All of the drama that happened with their foundation and everything. And so Atom was able to execute a very strong go-to-market. The token was pretty widely listed. We had a policy of not really engaging with centralized exchanges, but nevertheless, the demand for the token was very strong. So lots of exchanges listed us. Then this whole process of staking integrations, Coinbase made it, I'm pretty sure, their first staking integration. So that was a pretty big factor. Yeah. And so this very decentralized community of validators were operating the network.

But at the time it was just the Hub. So Cosmos literally consisted of like one chain, no connections. It was only that first part of the vision.

Yeah. It was --

There were no -- were there any other Cosmos chains at the time?

Yeah, so I guess at the time -- so first and foremost there was this other chain called IRIS, which is mostly forgotten at this point, but that chain launched. At the same time, Matic, which became Polygon, was also live, running --

No way. They started as a Tendermint chain.

They're still a Tendermint chain today.

Oh, I didn't know that, actually. That's funny.

And then BNB chain also launched very soon.

The Binance Chain.

Yeah, you find the Tendermint code kind of everywhere because it is this sort of very well understood implementation of a BFT consensus algorithm, and it is very stable.

I mean, and so what was going on behind the scenes very much, though, before launch, everyone was focused on one thing. Then we launched, then we have this asset. This asset is worth billions of dollars, even just in the early days. And there was always a lot of conflict in the team. There was always a lot of different visions. There was always -- you had this incredible pool of talent that was working together. Jae, Ethan, Sunny, Chris Goes, me, Jack Zampolin. Like, you had this pool of talent that have all gone on to build lots of things. And so here is incredible pool of talent, but they were all pulling in different directions, and it was like keeping them all aligned. But really a huge tax on Cosmos from the beginning was, there was not a lot of alignment.

The company at the time was All in Bits.

Yes.

But was the ICF already formed?

ICF existed.

Okay.

Wow.

It was always like, I can build -- like I want this thing to exist so I can build my own thing.

I'd resigned from All in Bits, publicly called out Jae.

There was a moment where it really looked like this ecosystem is imploding.

Yeah, it really did. So, basically, most of the team then left All in Bits, started their own -- many of them started their own entities. A big chunk of people then started working directly for the Interchain Foundation.

There's also Interchain GmbH, different org. Right? Or what is it?

It's a corporation.

Okay.

Which has an extremely weird structure, but it is a German company that is owned by the ICF.

Oh, so it is in the same umbrella.

Yeah.

I always thought of it as independent.

But it always -- so there was a Berlin office for All in Bits, and that Berlin office all resigned. And they resigned and they went under the Interchain Foundation, but with their own -- with a considerable amount of independence. What was clearly lacking was a leadership team at that entity.

Okay, but the Interchain GmbH, that was the team that eventually shipped IBC. Right?

It did ship IBC.

What happened to All in Bits around this time?

So All in Bits went through a couple of iterations. So there was like the immediate post-crisis intervention. Then this early contributor named Peng took over as CEO. And Jae stayed on the board, but left. Still had the majority shareholder control of the entity, but Peng. So then there was the Peng era. This was pre-IBC. So there's the -- and in this era, there was an attempt for everyone to work together. So I had left, I was just working basically under my validator entity, Iqlusion, and doing my own thing. The advisory business started to expand. So invest advice business started to expand, but mostly working on shipping IBC. You know, we did a whole episode about this. But one of the things that was really painful was all of these centralized exchange had integrated Atom. And then as part of shipping IBC, we were making substantial changes to the code base.

The other thing that happened is there was this team called Regen, which has a chain that is like, again, not super famous, but exists within the Cosmos. They actually took over as maintainers of the Cosmos SDK. And so now you -- so you had the patterns that I think have been problematic for Cosmos get sort of cemented in this era, which is all of the core contributors have their own projects, all of their own core contributors have their own teams. There are a variety of different teams. No one has any value alignment or alignment around Atom. Like very few people have, very few people are contributing own Atoms have Atoms. Some of us do. Like, I have a bunch of Atoms from the early fundraiser days and stuff like that. But again, no alignment around an asset. And like most entities are kind of ad hoc groups of people who kind of were like, okay, we have to figure this out before we miss a payroll cycle. Let's just put an entity together.

Whoa. This is the lead up, though, to the Stargate upgrade.

Yeah.

And that, actually, I did an episode. All of these I'm going to add to the show notes, but this one was, I think I interviewed three different people to talk about this Stargate upgrade.

Yeah.

That was the launch of IBC.

Yes.

Yes.

So you made it through, like, despite this fragmentation, it got there.

Despite all of this, we shipped IBC.

Was Terra already live before --

Terra was live

IBC. Okay. What was it doing itself? Was it just a standalone thing?

Terra had been pitched as a payment rail in South Korea. Then they had pivoted to doing DeFi.

But as a standalone chain.

As a standalone chain.

It was kind of unusual at the time.

Yeah. And they had partnered with Wormhole, which is a bridge network. Also, early Cosmos people to connect themselves to a bunch of other chains in an early pre-IBC system.

There was no relationship. And Terra did not care about Atom. And couldn’t have cared less?

Yeah.

Yes.

So Sunny and Dev launch Osmosis, and then Terra decides to enable IBC. So like Osmosis launches, and there's not really anything to do on Osmo except--

Trade Osmo for Atom.

Trade Osmo for Atom.

And move stuff over IBC, which is cool.

And move -- but then Terra launches, and we get this massive injection of liquidity into the ecosystem as Terra has this massive fan base, massive user base. And a huge advantage for Terra was that Atom was widely available on many centralized exchanges.

So was that the link into Terra? Is that how people got Terra?

Yeah. So that was a big part of how people got Luna.

That was the on-ramp.

Like you could buy Atom, buy Luna, buy UST on Osmosis.

Cosmos had airdrops before anyone else. Airdrops are very easy to execute within the code. We kind of baked that in. We also baked in this idea of, you just would have a token. You basically launch an idea, okay, and you have a token.

And there is a massive graveyard of all of those chains too. Like most of them didn't work. Some of them worked for a little while, like Juno.

Yeah. But it's been a disaster. Just like -- I mean, basically, all of these tokens are down.

Yeah. Or dead.

Are down --

Or fully gone.

Or fully gone.

Yeah. There is only -- I mean, of that time, what launched? Stargaze, still around.

Stargaze is still around.

Juno. I don't even know. Is that still around?

Juno was around that time. It still exists?

Oh, it still exists. What else?

Sommelier.

Sommelier.

Yeah.

Anything else?

Agoric.

Agoric was around that time. And then there was a lot of random. I actually don't remember their names even.

So many, so much.

Wow. Gravity Bridge.

Oh, that was so fun. This was in Lisbon.

In Lisbon.

Yeah. That was like --

It was kind of like, you know, I have a house here. I, like, spend a bunch of time living in Lisbon now. That was my first introduction to Lisbon. Like I came --

That was such a special month. That was also pretty early after Covid era, so -- and I remember it was -- first, it was the Cosmoverse. I think it was first Cosmoverse.

There was NEAR, Cosmoverse and Break --

So I don't know which came first.

Breakpoint.

No, but there was an Ethereum thing too.

There was ETH Lisbon too.

And there was LisCon.

Yeah.

There were so much. It was actually, like a full month, maybe even six weeks, where every single week, a different ecosystem came to town. It was very cool.

Okay. It was fueling a lot of the --

Terra was fueling everything.

Okay. Because people were buying Atoms on exchanges to get into the ecosystem.

Yes.

So once they'd be having the Atom, maybe they'd hold a bit of the Atom, or they'd put it into other projects. Most of it, though, was going towards Terra because they -- and the reason were, like they would offer, what was it? Like, 20% --

APY on --

APY. And it was pretty consistent, unlike a lot of the other things, like Yearn, where you'd have really high APY and then it would drop, Terra stayed --

Yeah. It was just all artificially sustained by their -- by this sort of leverage that doesn't look like leverage that was built into the instability of the UST mechanism in Luna. And like all of these bailout. So there's all this drama, right? A lot of us didn't trust them. We looked at the UST mechanism, we were like, this is unsound, like this thing has a huge death spiral risk, but it was the thing that was keeping breathing life into the ecosystem. We all were starved for resources to build any of the things that were needed. To build wallets, to build block explorers, to build everything. And so we were also trying to get resources from that.

It's interesting, like IBC launching at that time. So far, there's never been a big bug in IBC, has there?

No, there's been a big bug in IBC.

Oh, there has been.

Yeah. So why don't we get to that part of the story?

Wait. What year -- is that pre Luna.

Post Luna.

So let's do Luna quick.

Yeah.

Yep. $60 billion gone.

So in the immediate aftermath, liquidity in Cosmos collapsed just catastrophically.

It was just gone. It wasn't even like people exited. Like, it was all in Luna and then it was zero.

Yeah. And it was just like --

Did anyone get anything out? Like was there an attempt? Like did anyone settle into Atom? Did they get into Atom and just stay there? Or did everyone like move out whatever they could?

Most people, it happened too fast.

Whoa. Just wipe out.

So, yeah, so the ecosystem -- I mean, and the ecosystem has never economically recovered. But also we now no longer had a stablecoin, DeFi in the ecosystem became essentially non existent. We were really -- like the whole ecosystem -- like there was just -- because basically, what was the ecosystem at the time was there were a bunch of chains, but it was like, there was Atom, which was this token that didn't really do much, but it was listed in exchanges everywhere and you could stake it. There was Osmosis where you could do Uniswap-style trading, and there was Terra, which had like the stablecoin and anchor, but then like a bunch of other stuff, like apps launching and stuff like that. There was a lot of activity. And so once Terra was gone, that whole structure just was destabilized.

Wow. So this destabilization and it's not -- I mean, the Cosmos hit was early and hardcore and fast, but the rest of the market gets hit over and over and over again by other things.

Yeah. Then FTX happens.

But there's also bridge hacks and the culminating in FTX, the final pin that just let it all crash down. Cosmos during that time, did it experience actually the same level of crashing as the rest of the market after that initial Luna thing?

Yeah, yeah.

It continued to feel rough.

Just rough.

Okay, so --

So Luna is gone.

Luna is gone.

Or like Luna is gone and the refugees have not really coordinated into anything. That will come later. The builders in the Luna ecosystem, some of them came over and wanted to build stuff in Cosmos, but very little had launched. So it was really like, there was Osmosis, there was Informal Systems, there was Iqlusion, Zaki, there was Jack with Strangelove, there was the Interchain GmbH team, there was the Interchain Foundation.

There was Binary Holdings.

Binary Holdings didn't really become a player until later.

Okay. They had been a validator.

Yeah, so.

But, okay, I thought that they were already.

So we had Regen.

Okay, what happened? Is All in Bits not even a thing?

This is at the end of the year.

This was like mid -- like post-Luna collapse.

Post-Luna, but pre-Atom 2.0.

re-Atom 2.0, Jae comes back during the summer.

Okay.

Jae tends to be more active during the summer.

Okay.

He has this weird seasonality to his behavior. And so he tends to be more active during the summer. He comes back during the summer, another round of people getting fired from All in Bits happens. Another round of -- like Peng leaves. There's like a whole nother round of drama. But it's very self contained, because the rest of us are off dealing with Terra collapsing.

Okay.

But there still is an energy. There's like, there's an energy.

Despite all of the collapsing, I guess there is this hope that because there was some really interesting building happening in Terra, it had attracted not only money, but also talent. And those folks still understood Tendermint at least. So like --

Yeah, they understood CosmWasm, which was our smart contract. We had a lot of people who had been building with CosmWasm, and CosmWasm has been weirdly sticky. Like developers who got exposed to it don't want to leave, so they keep looking for places to build. That's been like weirdly sticky. CosmWasm is a smart contract language. It was this niche smart contract language that got -- it was never really organized, that Cosmos would have its official smart contract language be CosmWasm. There was a hackathon project, and then some of the devs from that hackathon project kept building it. And it was kind of not going anywhere. Then Terra adopted it without really asking them, they just adopted it. And suddenly this created an entire builder community around it, and that has like weirdly persisted.

Oh, that's the bug. And that's IBC bug. See, I always thought it was a Cosmos Hub.

No, it was a full drain everything.

Holy moly.

Double spend, but we found it. So what had happened --

This is Dragonberry.

This is called Dragonberry.

Okay. Who named that? The hacker?

No, no, no.

You found a bug.

So, Dave -- Dev from on the Osmosis team. So what happened was Binance is a bridge between -- that they used for BNB, the asset, was based on Cosmos tech. They literally used a bunch of code in Cosmos tech that Jae had written for a hackathon. And like --

Oh. Not CosmWasm. This is IBC code or something. Bridging code.

This was code in Tendermint.

Okay, okay.

So they'd written their own bridge, not using any IBC tech, but using this prototype of IBC that Jae wrote for a hackathon once. And the rest of us had -- like I mean, I had looked at it and been like, I wouldn't use that code, but, okay, like, whatever, move on. So someone had hacked that bridge for $50 million. And so, I remember I was in New York. I was walking down the street going to meet Shane from Stargaze for Sri Lankan food. And I'm scrolling Twitter, as I always do, and I see samczsun from Paradigm tweeting about this Binance hack, and he's halfway through it, and I'm like -- and I realize that it was a vulnerability in Cosmos-related code. So I called him, and I'm like, you have to stop tweeting. I need to go figure out --

If there's any effect. But why -- it wouldn't be, though, right? Because if they're using -- they're not using the IBC --

I hadn't put all these pieces together. So I'm just like, I leave dinner with Shane, and I'm like, sitting on a stoop in the East Village in New York on GitHub on my phone like tracing out the code, and I'm like, okay, IBC doesn't use this code. It's like Jae's weird hackathon code.

You're safe.

We're safe. But then we're like, as always happens, like, half of the Cosmos devs that are sort of core Cosmos devs are like, on a plane while this is all going -- I couldn't get in touch with anyone. So I'm just doing this by myself. So that happens, but then the next morning, everybody wakes up and we're like, well, all right, the bug was fairly exotic, fairly unique to our code base. We should go look and see if there's anything like this in IBC.

Okay.

And then that starts a 72-hours bug hunt that a bunch of people contribute to, and we have a bunch of near misses. And then Dave finds a full exploit.

Exploit. Whoa.

I mean, it was the first of, and then -- I don't want to spend the entire time doing the timeline, but because we got to talk about the philosophy of this thing. But anyways, so we find the exploit, and then we start this frantic effort to get all of the networks patched. It was like one 72-hours zoom call, where we ran a war room, and we would just call up the validators on different chains, and we would be like, we would just patch people's code without telling them, call their validators, say, it would be like Zaki and Ethan have told you, stop running Injective, start running this variant of Injective on pure trust.

Whoa.

And so we ran around, we patched all of the Cosmos chains. No funds were lost.

Why Dragonberry?

So, I mean, so at the time, we were using fruit names for --

Okay. For bugs?

For bugs. For security vulnerabilities we were using, and it seemed like a big --

Big one.

The big one.

It's a weird one. It's exotic.

Yeah. And so we were like -- it was like -- but I think the entire experience created this, at least within me, the sense of crisis in the software development of Cosmos.

Ooh, you're saying, like why was that allowed to continue? Why didn't they find it?

Exactly. Why in the name of God did this actually get through our software development process?

Had it been incorporated later, or was it a very old bug?

So the bug was in the original IBC implementation. It had been there from the beginning, but it showed a lack of capacity in our auditing process. Like, all of the things that I had been a part of and the rigor in which we had built the Cosmos Hub, launched everything, was just fragmenting because of the fact that you had -- like it was no longer one team. It was like, well --

Although it hadn't been for like two years at this point.

Yeah. And it's like one team is writing IBC, another team is auditing IBC, different parts of the code are owned by different people. So it was like one team owned this part of the code, which is not even part of the core IBC team. There were just all of these -- you know, you had this fragmentation of development, and it was like, clearly, we have a problem.

Yeah.

I mean, I remember it as something to vote on in the Hub. It was a signaling proposal, as far as I could understand, included in it was sort of like, what ICS, this Interchain Security, would mean, but yeah.

ICS, yeah.

Yeah. And so actually, the first -- so the way Jelena, who's the founder of -- CEO of Noble and co-founder of Noble, and I got close and started working together, was, I was helping her with go-to-market --

For ICS?

For ICS.

Maybe just briefly, ICS is a bit of -- it's funny, I just did an episode recently on restaking, and it's a bit like --

Yeah. It's the first restaking primitive.

Yeah. It also had a bit of an echo of Polkadot, with the Hub being the center and then chains branching off it. Yeah.

Was it like going to redistribute like --

It was going to redistribute a massive amount of Atoms?

Whoa.

It was a huge restructuring proposal of a dramatic nature.

Wow.

And because if you wanted to get most of the core builders back on the same page, back aligned around Atom, back aligned around like on one team, essentially, this is what it was going to take. And when Atom 2.0 failed, basically the fragmentation of the ecosystem accelerated.

And just like a note, because the last episode that you came on to talk about this, at the time, Atom 2.0 had been voted down --

Had failed.

But it was just a signaling proposal. But you sounded still pretty positive about it. Like was there a sense, even after the signaling didn't work, that maybe it would still happen?

There's something coming.

There's a bunch of stuff coming that's very -- that's exciting and is likely to have a positive impact.

So, Jelena had left Informal and started Noble. Celestia had raised a lot of money in the bull market and had become more and more of a player in the ecosystem. The modular narrative had really kind of hitting its stride.

2022 was the first Modular Summit. '23 was the big

The big one.

The big one pre --

Yeah, so the L2 narrative was kind of strong. The shared sequencer stuff. There was, like, lots of -- there were just all these modular things, lots of tokens. There was everything. A lot of this sort of infrastructure madness was kind of very strong in the VC investor community, and there was a lot of energy around that. And concretely, we were like, we're going to launch dYdX. It's going to really show what this technology can do. USDC is going to come to the ecosystem.

Native.

Native USDC, that's going to inject a lot -- like restore a lot of liquidity to the ecosystem.

This is replacing what Luna had sort of destroyed.

Yeah.

Because even though there were always stablecoins, but they were like, bridged.

They were bridged from Axelar, primarily.

Yeah.

People were starting to really engage with interoperability. You know, I was frustrated with the IBC team because they were not really focused on modular. They were not focused on outside of Cosmos, all of that stuff. But that was -- and so, I helped co-found Hyperlane as a kind of IBC alternative in that era. And so, yeah, there had been like a lot of -- there was just like a lot. So there's a lot of stuff coming to the ecosystem that was like, to be excited about. Even if there was less hope in my mind that would buy time for Atom to kind of figure itself out.

And so at least you had that experiment in action.

Yeah. And the flaws of the ICS product became apparent pretty quickly.

Okay.

It became -- Neutron had a lot of halts and stuff like that. And that was clearly -- and this idea of not having a very aligned validator set where you -- because the first version of ICS had to be opted in. And the other reality of the situation was, is that it became very clear that people were not willing to pay -- projects, were not willing to pay a lot of their tokens to use ICS. So Neutron ended up doing this airdrop to the community pool, which the Hub can't figure out what to do with, and no one knows what to do with. Stride provides like 10% of their fees or something like that. Some small percentage of their fees to the top stakers, and people are like -- it's like, all right, what is this? Right? The excitement that would then -- this was the first restaking thing, but then EigenLayer starts their restaking process, or like their contract that you could walk funds in --

On Ethereum.

On Ethereum. And that becomes 3, then 5, then $15 billion. And you're like, no one remembers this ICS thing.

Yeah. There's also, I don't know when, but liquid staking on Cosmos. Was it like --

Yeah. So I really, really pushed liquid staking on Cosmos, put years of work into --

Yeah.

Still pretty --

So Celestia bull market is still running strong. Like that airdrop narrative with Atom had continued

So Celestia's airdrop to Atom holders did really well.

Yeah, yeah.

And people got really excited about it.

Totally because it was unexpected and nobody really could game it in the way that -- I mean, I think there was this assumption like, you should stake your Atoms or you should stake Osmosis maybe, but there was no big game.

Yeah. The Celestia airdrop was well executed. The Celestia launch was well executed. It was executed well in the same way that Atom's launched, you know, most exchanges listed it, it had a strong narrative, strong team. All of this stuff. Like really good.

But come the new year, there's these other airdrops in another ecosystem that sort of sour, at least the spirit of airdrops. People get a little bit more annoyed. That sort of, I mean, in a way, it's the false activity gets called out and shown for what it is. But it also --

Airdrops are never a reason for an asset to exist.

Yeah.

Like, let's just be, from my point of view. I think, what did people understand Cosmos assets as? So like, first, there was just this like, we don't really understand where the yields coming from, we don't understand inflation, we don't understand unit bias, we don't understand any of this stuff. But we're like, the markets, like, well, I just get more Atoms in my wallet and I can do it from centralized exchanges and I can do all of this stuff. So it's like yield coin. All right. Then people got retrained as this is airdrop coin. Okay?

So you hold your asset because you might get airdrops.

Yeah. You hold -- you stake your Atom, you hold it, you might get airdrops. None of these things are sustainable reasons for anything to be valuable or exist but those have been dominant narratives. And in general, as one would expect, those narratives have been running out of steam.

And they sort of managed to keep them going for a while by slightly altering the plan. But, yeah, it was getting kind of into -- I think the Ethereum airdrop story became more -- like there was just such a complex amount of activities that people are supposed to do. They kept sort of adding, like, oh, it's not just hold it in this asset. Now you have to do something. You have to move it around, like five times, you have to spend this much, you have to do that. And everyone -- like a lot of people trying to guess what they're supposed to do.

Yeah. But on the builder side, for Cosmos, the story started to become, if you airdrop to Atom holders, all you do is get tens or hundreds of millions of dollars of sell pressure for your asset that, if you are lucky, reproducing the success Celestia had with their airdrop on the builder side. And so, more and more teams started asking the question, okay, if I want to airdrop within Cosmos, does it make sense to airdrop to Atom stakers? Do I want to look for users? Do I want to try and find other signs on-chain that people are actually users? Because a big frustration that builders had within Cosmos is they would look at these numbers and every builder in Cosmos would look and say, okay, I see 3-400,000 accounts staking Atom. Great. I see 50-60,000 accounts swapping on Osmosis in a month. My app has 100 users. How do I grow? How do I do anything?

And this was a problem that every builder was having in Cosmos. And so people started really -- there became this huge, became a very strong misalignment between Atom stakers and the staker community in Cosmos and the builder community in Cosmos. Which I think, again, has propagated as a problem, because it's like, I want apps that people, all of this stuff depends on apps. If apps have no users, why am I building in this ecosystem? And if all of these people who are staking which, who appear real, like it's not just -- you know, it's not one person with 400,000 accounts. There's like, at least a few hundred thousand people in this community. But if those people can't be engaged to engage with any apps, like, what do we do?

Yeah, we're at an inflection point.

So now we're recording this in September, but I feel like a month ago, two months ago, you just started to see kind of core people, like people in those teams that you've mentioned, say things like, I think it's dead, it's over. I think even maybe you've tweeted something that suggested it's over, which I think my team at ZKV flagged being like, whoa, if Zaki thinks this --

Okay.

This is post Atom 2.0.

This is post Atom 2.0.

Wasn't Interchain GmbH supposed to be doing that, though? Weren't they sort of meant to be like the IBC center point?

Yes.

Okay.

But --

No one could be the center point.

You had different users who wanted different things. One of the big problems that Cosmos has had is the teams were not very commercially oriented. They were all these very researchy-oriented teams funded out of the Interchain Foundation, who were not -- so you have this division in Cosmos right now between teams that were very commercially-oriented, Neutron, Skip, in some senses, Strangelove, me, who are very commercial-oriented, and then this not very commercially-oriented team. So users of the tech were like, we are missing this feature. And frequently the attitude, like what you would get back from -- I mean, again, this is just my perspective, but on the whole, what you would frequently get back from the Interchain GmbH team is, yes, we will work on that in two years. And everyone's like, yeah, that's not going to help. So I'm going to go find someone, Strangelove, another vendor to make it to like --

But that's kind of dangerous then, too, right? Because that's when you get other teams adding and, like maybe it won't be vetted as much and checked.

Exactly. And in my mind, we created an entire design that was supposed to mitigate this bug at a design level, and then that design was subverted because of incoherent design. So we get this bug. Again, it gets white-hatted, these wonderful people at Asymmetric Research who seem to be finding lots of Cosmos bugs, who seem great, like, report the bug through our bug bounty process. We get this bug bounty. The chains get patched. Now, in June, somehow, well, we kind of know how -- sorry, this is like a lot of stress.

Oh, no, this is still fresh.

Yeah. So in June, Terra --

Terra? Rears its head again.

Rears its head.

Oh, no.

And Terra has reverted the fix. So they reintroduced the bug --

That you guys had just fixed. But just for the Terra --

Just for the Terra chain.

Okay.

But then Terra gets exploited for like $6 million.

Who's still on Terra, though?

But it involves a very sensitive topic.

Yes.

Terra. Not just like Terra for you, but also Terra, I think it triggered lots of people.

Yeah. And so --

So Terra broke again.

Terra broke again.

Something, something Cosmos.

Something, something Cosmos.

Something, something IBC.

Something, something we don't really understand, but we were just like, Jesus Christ, what is going on here? And so this triggers this just avalanche of -- like there had been in Cosmo circles, negative sentiment kind of brewing for a long time, but this somehow escapes the bubble and triggers this outpouring of negative sentiment that has just been sort of with us --

Since then.

Since then.

So Osmosis, the team has just been grinding on the core product.

Yeah.

Like, Osmosis --

They got it back on track.

Osmosis is just killing it from a product point of view. The product is just unbelievably good. It's as good as any DEX. I mean, I use everything. I use Solana DEXs, I use Cosmos, I use Ethereum. Really, the product is as good as anything out there.

Nice.

And like, just -- and the team has done such amazing work and it contributed a lot of amazing work. They've fixed Comet BFT Tendermint bugs that I've been chasing for years. Just amazing, amazing, fantastic work that has not been rewarded in any way, shape or form by, in token price or users or anything, but it's there -- it's still fucking amazing.

Okay. What about Informal?

And Cosmos SDK?

No.

Oh, okay.

So they own Cosmos Hub. So the Cosmos Hub team is Informal. So Jelena onboarded two teams, Stride and Neutron. Then no --

To ICS.

To ICS. They announced, Jelena left, Informal executed, got them onboarded, launched them. They've been running for more than a year. There have been no new teams. The teams that Informal has talked to about onboarding have all come back -- came back with a litany of reasons why not to use it. Informal has dutifully worked on all of those problems. And so, there is a version of ICS that is sort of about to launch that fixes a bunch of obvious problems, but there is also absolutely very little reason to feel imminently excited about those product improvements. It seems very -- it would be -- it's like a huge lift to get from where we are today to something that would really get people excited about Atom and the Cosmos Hub through those product improvements.

What about Strangelove today?

So Strangelove is still a contributor to IBC. I don't know, like their main thing right now is -- so one of the things that's very exciting, actually, about Cosmos is so one of the biggest Cosmos chains. Oh, actually, that was -- it wasn't live, but they started building it before Cosmos Hub launched. It's called, is THORChain.

Oh, yeah. Is it also Tendermint, somehow?

Also Comet Tendermint, everything. So Strangelove is actually involved in this massive project. So THORChain, while it was built with Comet BFT, was never integrated into the IBC network?

Yeah. Yeah. I think of it as very separate.

Okay.

So, you know --

That's what Strangelove is doing. Binary, because we didn't really -- they sort of entered.

So there was this team, Regen, who had been responsible for a lot of the Cosmos SDK work post AiB collapse. And I personally was very unsatisfied. And I think a number of other contributors were not super satisfied with Regen's work as leaders. And Marco, who started in Cosmos as a DevRels guy, took over as product lead and frankly, quietly has made the Cosmos SDK a state-of-the-art blockchain building toolkit and fixed numerous problems. But it's basically taken 18 months of really, really hard work from his team to kind of turn Cosmos SDK into a truly state-of-the-art system.

Nice.

So a lot of builders are very -- like this is what we see, current generation of builders is super excited about the Cosmos SDK. You know, you have new projects that are like, we raised $100 million and we're going to go use the Cosmos SDK every -- like on a regular basis.

Story. Who was it?

Story, yeah.

Yeah, Story, okay.

Which I'm an advisor to.

And yet this summer we hear Cosmos is dead. And when they say that, do they mean Cosmos Hub? Do they mean Cosmos ecosystem? Yeah, what does Cosmos is dead mean then? Why is Cosmos dead? Or is it dead?

So I think what people very specifically mean is its, I no longer associate this asset, Atom and Cosmos, with the success of the ecosystem in a meaningful way. The fragmentation of the ecosystem now has reached the point where it's like, I realize that people will continue to be -- there will still be a Comet BFT, there still will be IBC, that IBC is the third largest bridging protocol by volume on DefiLlama. But I can't -- I no longer have any reason to believe that Atom will gain any value from this.

So it's not Cosmos is dead, it's Atom is dead.

Yeah. Because anything else doesn't make any sense. Right? Like is Celestia dead? No, there's new stuff launching on top of Celestia all the time. Is IBC dead? No, it's the third largest bridging protocol by volume. Is the ecosystem dead? No, there's $270 million of USDC on Noble doing DeFi stuff. Is dYdX dead? No, it's the second largest perp DEX. Obviously, Cosmos isn't dead. But what people are really saying is, where does the attention go? What is the core of the ecosystem? What are the things to be excited about? And this thing where it's primary property was its decentralization, that was -- many people viewed as the -- which was to a lot of people, the core of the ecosystem.

And the great part.

Yeah. Is there's nothing to be excited about.

Or its own decentralization made it hard to have that central leadership.

Yeah. And so one of the problems that I think, and we could get to -- I mean, it was probably a good time to kind of analyze how did we get here? One of the reasons I think that we really got here was, so All in Bits, the organization that we all worked at when we created Cosmos, shambles, Interchain Foundation has gone through many iterations of turnover of core people, of leadership, is now off on a search for an executive director again. Is doing stuff -- there are talented people who work there, I'm not trying to be negative about everybody, but it has been a deeply -- there's also a governance proposal right now to audit the ICF or to convince the Swiss to audit the ICF, which I don't know exactly what the theory of why the Swiss government cares about governance proposals on the Cosmos Hub. But maybe they do. But also, who knows if it will pass? I'm not bullish, this auditing process, but I was --

There's a sentiment of anger towards the ICF.

There's a sentiment of anger towards the ICF that you see in the community.

That's the foundation. This is the thing that gives out grants, this is the thing that -- this is separate also from the community Hub.

It doesn't really give out grants. It funds these various fragmented development organizations.

It used to do grants --

It used to do grants.

Yep.

This year, Brian leaves.

Brian left.

I don't know who else leaves, but it seemed like there was another big shift.

But don't all the foundations have that exact thing, though? Like, they're all meant to be non-profit, they're not supposed to be super financially driven? I mean, EF --

Yeah, bit then --

NEAR Foundation, Solana Foundation, I guess Solana Foundation's the exception. Right?

Solana Foundation. What we're seeing right now is Solana running a very -- the Solana Foundation being far more attuned to the commercial needs of the ecosystem. And just doing a really good job. People come into Solana, they want to accomplish something. They will get help from the Solana Foundation. There will be people at the Solana foundation will help them. The Solana Foundation will have an opinion. They will say, these initiatives are important, these initiatives are not important. There isn't this effort to please everyone. And it has shown, I think, really positive effects.

Is it the only foundation that has such a strong showing so far? Because EF is currently, and has been over the years, lambasted for paying too much or not paying enough or hiring too many people or firing too many people.

The EF, in my mind, is not very commercially-oriented, but pulled off such an astounding engineering feat with the merge and proof-of-stake and all of this stuff, which is just unbelievable that they were able to pull something like that off. And, but the fact that there has been very little commercial reward for that accomplishment, like this is the challenge that you have with crypto. Right? You set some engineering goal. The engineering goal is really hard. You go do it, and sometimes you don't get a -- there's no commercial reward on anytime horizon. But also sometimes the commercial rewards for those engineering goals can be delayed by months or years.

Sometimes you have people, though, be kind of pleasantly surprised that they get a retroactive or whatever. It's a weird space that way, but it makes it hard, like at the beginning to commit, I guess. Although I think, why have people committed in the past? I think it's because there was an assumption that there would be something at the end of that tunnel. Right?

Yeah. That like something would work out.

Yeah.

You know, you have a bunch of things, something works out. We're coming into almost like, basically, year plus of flat ETH price, Solana going up, but then flattening out. Like, things have been just like -- you know, one feels super rewarded for all of the hard work.

And that goes beyond Cosmos, then.

But that goes beyond Cosmos. The problems of Cosmos are to the rest of the ecosystem as well. I'll just cover like a third thing and then loop back, which is there's AiB, there's ICF, and then there's the validators.

Okay.

And in many ways, the sort of weakness of AiB and the weakness of ICF and the way the builders of Cosmos consumed of it is like, we are turning this thing over to the validators. But from the Atom 2.0 era onward, any sort of clear, this is where we want this thing to go, this is what we should do, from the validators, they've never viewed it as their responsibility, but at the same time, there was no one else. And they had the power. They had the power to reject everything. And so everybody has put so much time and energy trying to make incremental changes, oh, let's fund AA DAO, let's incrementally improve, let's constantly test the waters of the validators, because if we try to go too fast, the validator -- what we were in Atom 2.0 as the validators, we'll just say no. But that was too slow enough that the validators would accept is at this moment too slow --

To same the ecosystem.

For them to save the market of the system.

In a way. It's funny. So as mentioned, as a validator, especially during that era where those votes were coming in, we were one of the few independent ones in the top ten. It really looked like it was exchange, exchange, VC, strange kind of company game thingy. It was very --

Strange weird Japanese investment vehicle that no one really understands.

And then -- yeah, and I remember we took a lot of those governance decisions, especially the contentious ones, very seriously, and we would have big conversations trying to figure out what to do there.

The builders always viewed that ZK Validator vote as one of the most reliable allies. And with you guys gone, we're just -- we actually, we don't know. I don't know. I don't know what's going to happen.

Yeah, ZKV in the last month or so has -- you know, it's no longer in the top ten. We're still in the set, but we don't have the same backing that we used to and felt when it changed as like a bit of an end of an era, because we'd been up there for about five years almost. But yeah, I do wonder, who are the people? Because even if you look at the bigger VCs who had been there, even VCs that could have been friendly, they've been selling off. Like, you don't see the Paradigm or the Polychain.

The Paradigms, the Polychains is almost gone. There still are -- I mean, there's a lot of tokens, like Coinbase Custody is the biggest validator.

But they never vote on anything.

They don't ever vote on anything. And it would be great if they were -- you know, because it's just like retail and retail users on Coinbase, but they don't get any voting rights or anything, which is just really frustrating. We've tried.

Yeah. Okay. So now this brings us to today.

Yeah.

Atom is dead. Cosmos is not dead. But in a funny situation, the ICF is looking for new leadership, and yet we hear little murmurings of proposals, ideas, one of them is public, which is this Osmosis merger.

So one thing that has been sort of -- you know, is not in the form of a serious proposal, but there is this idea of merge Atom and Osmosis in some fashion that has been out there. And as I have said, I've been very, very positive about the product progress that Osmosis has made and been very conscious of the complete lack of price action on Osmo. Right? I think that another frustrating thing for crypto builders in this last year, not just in Cosmos, but just in general, is that product progress is very rarely rewarded with token activity. There's a few examples where people have been able to make both the token and the product work, but you are very hamstrung by the market realities these days.

Can control revenue.

Can control revenue, can grow revenue. Let's go focus on revenue.

Okay, but the proposals. So there's an Osmosis-Atom, like an Osmo-Atom --

There's like, a soft Osmo-Atom merge proposal that is out there.

01:09:0: Anna Rose:

But you think it's not super serious. I wonder if it's worth it to explore, or if you think --

What I would say is the dynamics are very clear, which is Osmo is now a great product, and Osmo is very much the entry point for people into the ecosystem. It is the most widely used thing, it has the most daily active users.

Although from centralized exchanges, it's still Atom, right? Or is it more like ETH into Osmo?

So one of the things, another form of fragmentation that has occurred is there are multiple front doors to Cosmos. So Noble is a front door because it has USDC from Coinbase, which is the only centralized exchange that it supports, hopefully more soon. But we integrated this thing called CCTP, which is a bridging protocol that is native to USDC, and that has provided -- you know, like you can go Solana Noble in under a second.

So it's a new on-ramp.

So that's a new on-ramp.

Okay.

And then you have Osmosis has gotten Binance and multiple other exchanges to list Osmo. So there are multiple entry points now into the ecosystem.

And TIA as well.

And TIA is another entry point into the ecosystem. So now you have all these multiple entry points to the ecosystem, but they all meet like a good 30, 40% of the USDC is on Osmosis. A good -- there's TIA volume, there's Atom volume, a lot of this stuff happens on Osmosis. So Osmosis is for users of the ecosystem very much now.

A Hub.

A Hub.

I feel like it has been for a while, but maybe more so now.

More so now.

0kay.

Because before Osmo was like --

Step 2.

Was DeFi, but it wasn't like where all of the different on-ramps met. But now you have TIA, Atom, Noble, all of them meeting in Osmosis. Osmosis is much more, and Osmosis is a much better product now. The UX is much better. Like, the functionality of the DEX is contemporary to other DEXs in the ecosystem. So all of that stuff is happening.

What would a merger look like?

I mean, in my mind, I tweeted this.

How does that even work?

So there would be the mechanics of it where there have been very few mergers in crypto. So the big recent one was this merger of Fetch and Ocean protocol, and I think a third protocol. Fetch is actually a Cosmos chain to do this AGI asset where they -- like you had three things that have all pumped because of sort of the AI bubble that have merged. So there was that merging process.

You and I were talking just the other day about, the NuCypher plus Keep --

Yeah, Keep.

No equals Threshold.

Yeah, it became Threshold --

A few years ago.

And that's also happened. So that merger has occurred. Very few mergers have happened. One thing, though, that is a problem that I think Cosmos and all the other ecosystems are now having is everyone's incentives are new token, new token, new token. You have this, get this, what I call fractalization of incentives, where everyone has their own token with their own ecosystem, which is their own thing. And you have so many of those in Cosmos. You have Celestia, then you have Initia, then you have Astria, then you have --

Stargaze.

Then you have -- you know, thing after thing after thing after thing, which is all producing their own ecosystems around them. And you're like, okay, what are the incentives anymore? Now, consolidation has to be the way that we put all of these incentives back together.

Yeah, I mean, I even remember when every project you talk to, I mean, still kind of the case, is about to launch a token, and you have to ask yourself, how many of these things are we going to hold on to?

Are stuck now.

Are now -- you know, EigenLayer tokens are going to become liquid, tradable. Like all of these tokens suddenly. So now you have this like another massive wave of tokens into a market that has just, like there's no buy side.

Whoa.

So it's a painful time not just for Cosmos.

That idea of merging tokens. I mean, I think this proposal is Osmo like, Atom, may or may not make sense, but I think that generally, that's a kind of good idea. Or has anyone absorbed tokens? Could one stronger token absorb another?

Yeah, so, I mean --

I think Polygon did that a little bit with the Hermes token.

I mean, somebody could write a thing that would, for instance, allow people to deposit Osmo and get vesting Atom of some kind in some module. Like, you could build this. Someone could build a Cosmos merger, smart contract module, whatever. They'll do this, and I suspect someone will eventually.

So you would have, though, Osmosis sort of give itself over to the Atom.

But what I would -- in my version of this, the only version of this makes sense is, the entire Atom product roadmap of today, which is this ICS product roadmap is abandoned. The Hub product is the Osmosis product. The Hub asset is Atom.

But you'd have to -- how would you possibly do that? You'd need the buy in from the validators, I guess.

You would need the validators to --

And now no one.

No one knows.

No independent is on the -- is up there.

So yeah, we don't -- so, you know, you don't know.

Wow.

But like yeah. And not only, you would need the Osmosis validators and the Hub validators to agree.

Okay.

And these things are hard. It is manifested, but this is just one of many plans that exist in some way.

Can you share any others or hint at?

So, I can hint at things. So Skip has a plan. I think we'll see when this comes out, whether or not this thing has leaked onto Twitter. But I'm not going to be the one to leak it.

If so, I think listeners should check your Twitter when this comes out and see if you retweeted anything around a Skip proposal for the future of Atom.

The more commercially-oriented teams -- so I would view Osmosis as a commercially-oriented team, I would view Skip as a commercially-oriented team. They would say that there's such demand for what Cosmos makes, which is unability to build your own sovereign chain. Now, it's really unclear exactly what a Hub asset should do in that world because sovereignty is all about doing your own thing. Why do I need --

Why wouldn't it just follow the Uniswap model, though? If it is sort of like DeFi? I guess you don't think of it as that anymore. Now it's a center.

So what I would say is the demand for sovereignty is huge. Again, all of these problems with Cosmos and Story, Berachain, all of these people are building with Cosmos tech, Babylon. A lot of the things that are like late this year, early next year that people are actually excited about, in spite of this massive supply of tokens and everything, the stuff people are actually excited about, so many of them are built with Cosmos tech, and so many of them are -- I'm an investor, advisor or whatnot. I've been very much -- I shifted out of Cosmos Core and I put all my energy into the more commercially-oriented stuff, which has been a struggle making it work. It would have just been far easier to build infrastructure and live off the ICF. But I threw myself -- like both me and Kristi threw ourselves into this vortex of trying to make commercially-oriented things work in Cosmos because we were like, the ecosystem will die without more commercially-oriented stuff. So we worked on Celestia, we worked on Babylon chain, we worked on Berachain. Like very, very commercial teams.

Last question, because I know we're almost at time. So you've talked about some of the proposals for how to kind of save Atom or Cosmos. I mean, Cosmos isn't dead from what I'm hearing, but it's not booming, and Atom is somewhat dying. There's people selling, there's people undelegating. I mean, we saw that. But so far, you shared a few. There's some proposals to fix it or turn it around. So there's still a desire to do that. But my actual last question is, when would Cosmos actually be dead? What would it be? What does that picture look like?

So the picture of Cosmos being dead -- okay, so another sort of meme out there is the Cosmosification of the rest of crypto. So Solana, the new buzzword is network extensions --

Which is Cosmos.

Which are Cosmos -- which are sovereign chains that extend the Solana network.

The like rollups.

They're a bit like roll ups, they're a bit like Cosmos chains, whatever you want -- I don't know. I have not read a formal definition of what a network extension is. It's just branding that is different from L2s. Then people are like the Cosmosification of Ethereum. ETH is the new Atom like meme is out there. Ethereum does feel like Cosmos. Now, it's like you have this fractalization of incentives where everyone has a different L2, everyone has a different token, the core devs get sucked up, scattered across multiple L2s. Ethereum Foundation is not, you know --

The AVSs are going to do more of that even.

You have core Ethereum Foundation, people doing, like, EigenLayer advisory and all of this other stuff. And the revenue -- you know, the funniest thing is there are so many people who were pushing this, like Atom as gas as the solution to Atom's problems. And you see that Ethereum adds the tiniest little bit of scalability and protocol fees drop to like nothing. Right? So what would the death of Cosmos really look like? It's probably that the Cosmos idea doesn't go away, because I do think Cosmos as a mechanism of giving developers sovereignty, this ability of, or the idea that developers want sovereignty, the developers are not going to just accept that the choices have already been made for them. You must build with the EVM, you must build with the SVM. You must launch on top of the Solana Network with the Solana consensus. You must pay your MEV to the Ethereum validators or the Solana validators. You must do it in this prescribed way, and there are no other options, will never be the way that the most ambitious builders want to build.

So the dream you don't think will die?

The dream will not die, but the death of Cosmos is people no longer use Comet BFT, IBC, Cosmos SDK for doing this. There are no more Storys or Berachains or Celestias or anything. There's no future pipeline of all of that.

And I guess the teams that you mention close up shop or they shift, they start working on other things. But currently it doesn't look like that's the case.

We're not there yet.

Okay. Do you have another timeframe, another timeline? Do you have a 12 month, 6 month, 3 month?

I mean, I do think that the ecosystems -- what the biggest question is, is Atom relevant to the future at all, I think, at this moment? And I think there may or may not even get to be a decision point for this, I think is the question. You know, it may be that just again, the more commercially-oriented teams in the ecosystem just decide that they want to build something outside of the framework of the ICF and are just like, let us move on. And then it may be more of a individual decision among Atom holders. It's like, do I stick with the legacy stuff or do I move on? But I think the question is more, is like the ecosystem needs to be at a really fundamental level, restructured and reconsolidated and new incentive patterns need to be created in a much like Atom 2.0 way. But I do think that the appetite among builders to ask the validators now is pretty low. They may get one more chance, they may get one more ask, but it will be a very, like take it or leave it. And if leave it as the answer, here's our very established plan of being done with it.

Whoa.

I think people will be surprised at the willingness to do something big.

Cool. On that note, Zaki, thanks for this epic journey back in time to the beginning of Cosmos, all the way to today. Some sort of hints at what might be coming. And I mean, I think there's lessons to be learned for kind of anyone in the space from also stories of the Cosmos.

Yeah, I mean --

So it's really cool.

I mean, people like @chainyoda tweet always about learn from Cosmos because you are all walking down this path.

Also learn from Polkadot but --

Yeah, but I'm not good at writing all this stuff out, so it was a really good opportunity to embed all of my experience and learnings and trauma into a single podcast that will hopefully the AI can learn from.

Nice. Thanks, Zaki. I want to say thank you to the podcast team, Rachel, Henrik and Tanya, and to our listeners. Thanks for listening.