BTC, ETH and BNB Trade At A Premium On Binance.US After Court Orders Asset Freeze

Binance Outflows Slow After SEC Crackdown 

 is the former CTO of SushiSwap, one of the biggest decentralized exchanges, and the founder of Astaria, a recently created NFT lending protocol. Joseph has been heavily involved in building and managing DAOs. With the recent controversial proposals surfacing on MakerDAO and Lido, we are once again reminded of the importance of DAO governance. We begin by talking about Joseph’s analogy on what would happen if a lemonade stand was run by a DAO and reflect on some of the DAO experiments that occurred over the last couple of years.

We go on to speak about the different factors involved in forming hierarchies within DAOs and how individuals assume leadership roles. Joseph also shares insights into his experience at SushiSwap and the difficulties he encountered as the CTO. Joseph goes on to share his thoughts on reward models across different DeFi projects. He believes yield farming in the long term damage projects as costs of liquidity incentives become greater compared than revenue. A major discussion in DAO governance has revolved around token holder concentration. Joseph also discusses the current 1 token = 1 vote model and shares alternative governance systems.

Podcast audio and video was edited by Daniel Flynn and Alp Gasimov. Transcript was edited by Claire Gu.

🎙Listen to the interview in this week’s podcast episode here:

 is a DEX aggregator that works to find the best rates across multiple networks. Why use a single DEX when you can use them all? Find your best deal at 

, a new podcast about the next internet, from a16z crypto

 makes sure you’re getting the best possible price by routing trades over various DEXs, without taking commissions. Head over to 

👀 Only paid subscribers have access to the full interview transcript below.

Joseph Delong is the founder of Astoria, a recently created NFT lending protocol, and the former CTO of SushiSwap one of the biggest decentralized exchanges. Joseph is heavily involved in building and managing DAOs. He recently wrote a Twitter thread satirizing how it would be to run a lemonade stand throughout DAO. Can you go over what you meant with that thread?

Yeah, this lemonade stand [is] run by little Jimmy and a thousand token holders of this lemonade stand. He has a profit margin of 80% on the lemonade. And it’s just all the random things that happen in a DAO that are trying to influence the product. Much of those are around these absurd tokenomics, and then secondarily it’s around product features and roadmap, which is difficult. So for me, this is more of an explanation about DAOs that are building products. So MakerDAO would be an example. I really don’t see any major problems with ones that are doing like investments as a collective.

Those seem to actually go really well, just a good way to pull money. Maybe the user is generally gonna be more sophisticated, but for the ones that are building products, it’s really problematic because you have so many cooks in the kitchen and the reason that the corporations don’t take the advice of every stockholder’s opinion is that’s impossible. What they do is hire the management team and then they build the product together and they do a bad job– they get fired, right? In DAOs. this weird volunteerism mixed with token valuations. It’s insane. I think most of it’s insane, especially VE token models. 

I think long term, what we’ll see is a drift away from tokens. I think what we’ll see is people making normal equity companies.

Can you explain a bit more what are VE tokens and how have they been implemented in DAOs and why do you think the space will move away from tokens in general?

What are VE tokens and how have they been implemented in DAOs

Yeah. VE tokens are this: I provide an asset for the DAO to use as an investment. And then I get this VE token. That’s like my claim on the underlying assets that are non-transferable. And this is like that same madness that started. I’ll go point out like SafeMoon. This is in the same realm of tweaking the tokenomics in such a way that gives it magical properties so that people can’t transfer when we saw that with Ethereum too. Right. The Ethereum too has been nontransferable, but, you know, Lido and Rocket Pool exist. And if that’s the case, then how come it’s non-transferable? It’s like people just naturally take riskier positions because they don’t hold the thing. They use an intermediary. So the VE tokens in particular were just like an asset that I can’t sell on the open market. And therefore it’s gonna be valuable. This is just the sort of stuff just kind of bugs me. But generally you just say DAOs are like a collection of people that are trying to accomplish a goal and if that goal is building a product it’s very difficult to do inside a DAO.

So there’s the question of whether this VE token model where you’re kind of placed staking tokens in a DAO so that you’re linked to the DAO long term because the idea is that you can’t withdraw these tokens after a certain period of time has expired. So that’s one aspect of it, but I thought it was interesting that you mentioned that you think DAOs will move away from tokens in general. Is that true? And how can that [be]?

No, not DAOs. I think people building products will move away from the DAO model. Here’s an example. I was working in a DAO before, and I was basically making a salary to do that, but talent like myself, or any of the other people who work at a lot of these DAOs is valuable and DAOs aren’t really necessarily treating talent [a]s valuable. And so what they do, and this happened in a lot of the DAOs I’ve been in, [is that] they just break off and start working on like side projects where they get poached by other startups or they just leave and start their own startup, which is what I did.

So do you think DAOs will be mostly used to pull money together and invest? Will they serve as mainly investment clubs, I guess, or decentralized or semi-decentralized VCs, but it seems like you don’t think they can be a viable alternative to companies, to corporations.

I don’t think so. I think if we do, there has to be an understood hierarchy and governance process for installing and removing that hierarchy. I think a lot of the DAOs start not knowing that they’ll be successful, because people are just running experiments. I don’t wanna bash any of this experimentation either because I think that that’s like really valuable for the ecosystem, but they have these imperfect births most of the time. And then because of that, they’re not allowed to have explicit hierarchy. And when you have explicit hierarchy, you have a transparent system for how the hierarchy works. But the counterpoint that I think a lot of people want to believe is that if we don’t install a hierarchy, [or] that there is no hierarchy. But there’s a short story written by a feminist called the tyranny? Hmm. I have to remember the title. But the point of that story was that having no hierarchy doesn’t create no hierarchy. Having no hierarchy creates an opaque hierarchy. And so I don’t know who I’m responsible to, who is calling some of the product direction, et cetera. Is this problematic?

Yeah.I saw this. The Tyranny of Structurelessness.

The Tyranny of Structurelessness. Thank you. Yeah. That was in the talk I did at ETHDenver.

The Tyranny of Structurelessness explains how leadership and power can be ceased when left in a void. We go on to speak about the different factors involved in forming hierarchies within DAOs and how individuals assume leadership positions. Joseph also shares insights into his experience at SushiSwap and the difficulties he had encountered as a CTO.

So on December 27th, I was, I was working part-time at Dapper Labs on the Flow protocol. I had come from Eth2 research. I had done that for about two years– Eth research and implementation. And then I was working at Dapper Labs and Maki whom I had known from DeFi Summer called me and asked me if I wanted to be CTO. And it’s like, oh, okay, that’s interesting. Because I thought sushi was a really fun experiment at that time. I had some like tokens, not very many, but I had some and I said, yeah, okay, I’m interested. And I talked to my wife about it and she said I should go for it, which actually shocked me because she generally is pretty conservative and trying to work at a thing that isn’t a thing is difficult. So then I said, okay. I called them back and said, okay, it has to be CTO because it’s gonna be, you know, beneficial for my career. None of this is made-up role kind of stuff, the ed chef of blah, blah, blah, something.. And he agreed and they wrote up like a proposal and that proposal passed on January 5th. And then I was helping build out products and roll out products and manage a team and work on technical architecture.

And then what happened that you left? I know that this is very much tied to how DAOs are structured. From the talks you’ve given, it seems like that was a big part of why in the end, you decided to leave.

Yeah. The team was fighting a lot. There was this continuous infighting. The ability to build anything, which was the whole reason that I was there, to build something interesting, became diminished and almost eliminated. And so I just said, you know, this isn’t for me anymore. Like, we would have these kinds of long group calls that would just lead to nothing. It would just be everybody yelling over everyone. And it’s like, insanity. And I’ve experienced this not only at Sushi. I’ve experienced this at other places that are decentralized. There’s no hierarchy. I came from the Fortune 500 world. I was working at a bank, [and] it’s like if you want to go for the polar opposite. I always anticipated the goodness in people was the reason that things like banks and DAOs themselves would be somewhat functional because there’s a limit to the level of human cruelty.

But yeah, I’ve found out the opposite. There’s kind of no limit to the level of human cruelty. And that makes DAOs really impossible in my view without some sort of structure. So we were fighting a bunch and I could feel it that Trident was never gonna ship at this rate. We would have to offboard a bunch of the team. We would have to onboard a bunch of new team members and I didn’t have explicit power to fire, but I only had some explicit power to hire. So I just figured I would go take a break and then found my own company, which is what I did. I went to took a month’s vacation, went to Hawaii, went to Dubai, went to like everywhere, [and] then started Astaria.

So what are the, the main lessons that you took from Sushi and are now implementing in Astaria?

Yeah. Explicit hierarchy is a big one, like how everything is laid out. What else is there? Well, I’ll tell you one thing. Building is a lot faster. I feel like I’m on super speed with a team, a close-knit team. Another thing that I learned that is probably not necessarily related to DAOs, but was premature scaling. I think Sushi was definitely premature scaling the team. We were I think at the top 24, 26 team members, and 18 of those were developers. And fast, small, effective teams are a lot better. Counterparty risk, you know. Where the control for the thing you’re holding is what the Bitcoiners always say is, not your keys, not your Bitcoin. Something like that. Not your ownership, not your GitHub. Not your keys, not your multisig, et cetera.

So do you mean team members should, should have ownership over the thing they’re creating? Like there should be accountability, is that what you mean?

Well, has to be somewhere. There has to be ownership of something somewhere. Like somebody has to own a domain name. Right? There has to be a legal entity of some kind. So be a person or a company, someone has to be the admin or the owner for a GitHub repository or GitHub organization. Somebody has to be. So there, there are all these touch points of control and that wind up being a disaster when things start to unwind it’s like, that is where the literal control rests. I think like one thing that I thought about at ConsenSys is, ConsenSys was another one of these leaderless organizations. And my time working at consensus was… by the way, I love ConsenSys, but it was a bit chaotic. And when it’s leaderless, this counterparty risk that I was talking about, where the control of something actually lies wound up being financing. So can I ask you to write me a check to let me do X or Y or hire a new person, or build a new software. And that’s where the explicit control comes from.

And that’s, I guess very much tied to hierarchy, just like having a specific and explicit structure for the team and who’s in charge of what and who responds to whom and so on. So I guess the main lessons are having that structure, that hierarchy within the team. Having explicit control, explicit ownership, scaling gradually, or scaling according to the project’s growth. I guess that’s a matter of efficiency as well. What about communication? I wonder the fact that there are so many contributors in doubts coming and going and this picture… It has to make communication pretty hard. Different time zones. What kind of lessons did you learn from that?

Totally. I think that communication is a very key one that I didn’t touch on at all. So I think for Astaria, one of the things that we did was early we picked like our channels for communication, which for us it’s slack for another team, it might be something different. Like with [the] DAOs I’ve been in, I’ll be stretched across Telegram, Discord, sometimes Signal if somebody wants to use that, and Twitter DMs. It makes you all over the place in terms of communication. Your communication graph is just huge. Yeah. Picking a venue to communicate. I also think the way people communicate with each other is really important, right? 

I think we always equate professionalism to suit and tie and well-mannered not cursory et cetera, but I think I have an alternate definition of professionalism, which is this idea that you commit to something and you hold yourself accountable to that thing that you’re supposed to do.

 As well as your communication should be with your team and your coworkers should be at least polite. I think that’s really easy to have a polite conversation with them about something, because none of this at the end of the day matters super deeply. I mean, we’re all just trying to go home to our families and trying to be the most successful organization. So, regardless of whether you are feeling one way or the other about the way an organization should go, the conversation should be polite. And I‘m probably guilty of not always being polite in those instances. So again, as a general matter of fact, I think that we should endeavor to be more polite with one another,

For sure. Where do you think token holders have a role in organizations? In DAOs the way that things have evolved is that there are these instances for governance to give power to investors and users of the thing that you’re managing the protocol. And I think that’s a very inspiring idea for the future of organizations, like democratizing decision-making to the users of web3. And from what you’re saying, this makes organizations less efficient, but I wonder if you think that there could still be a role for that.

Yeah, totally. I think governance tokens are critical, but I would say that it is a governance token and not a management token. Right. So we should use it as a governance to install management, install hierarchy, and remove hierarchy. And then the first few proposals should be explicit about that. I also think governance tokens are nice, but there are problems with some of the mechanisms. One of them in particular is yield farming, or what do they call it? Token distribution, staking, staking rewards, et cetera. This style of token emission is so problematic. The whole design is essentially to get people in the market to be buying the token, to get people using the token. And then now the ability to emit these tokens has some sort of value, right?

And so in instances of some of these protocols that emit tokens for liquidity, that makes sense to do it in a bootstrapping sense, like to get people interested in a project and amid some tokens. And then many of these projects are making revenue, but they’re not making profit because they’re emitting tokens that are, that are far exceeding their revenue in the protocol. And the expectation is that sometime in the future we’re going to succeed. And now all these tokens are gonna be worth a lot more, but that’s not generally how it works, right? Because we’re emitting tokens that aren’t exceeding revenue, and then we continue to emit and emit and emit emit until they’re completely diluted. A lot of these teams, aren’t getting tokens, like a significant portion of their tokens for their work. Just like I was talking about earlier, talent flees. Talent will go where their talent is respected.

One of the biggest challenges, DFA projects face is onboarding users. Typically users are incentivized through liquidity mining and yield farms. However, in the long term, these mechanisms damage projects as cost of liquidity incentives or far greater compared to their revenue. Joseph goes on to share his thoughts on reward models, across different DeFi projects.

Reward models across different DeFi projects

I always said that it was habit forming. Like the token emissions are habit forming. There is a very specific narrow band of when you should use it. And I think that is to get initial flywheel effects for something. And there should be a very short window that it’s using, because if you continue to emit, and your revenue is not even crossing the threshold of your token emissions, then you’re just paying people to use the platform, and it becomes a shell game. So you’ll see this with Spell, you’ll see that with Sushi, you’ll see some of these protocols that continue to emit because this becomes a shell game of where is the actual value being accrued.

So I completely disagree with this model. And I think people are coming onto that. Now the alternative is, when it first started and I’ve come like 180 degrees on this. They used to be the DAO maxi, token maxi kind of person. But if you look at Yearn. Yearn was really successful because they had a really great product. And then Sushi was successful, I think, partially because they borrowed a really great product. And then after those two really good use cases, the staking rewards [or] the notion that people had about staking rewards was that, oh, this is a good and useful mechanic to get uptake on your system. And they became so broadly adopted and so abused like the mechanized farmers who would farm specifically. It became so abused that it’s a completely irrelevant sta. And another thing to think about is the staking rewards should be somebody providing you a service. In the case of like Sushi or whatever. It was people who were providing liquidity. Or in the case of Uni, it was people who were early users. They got airdrop, but it should be somebody who’s providing you something because you’re kind of giving away something of value. So I think people are just totally drifted from this. You look at Gem, you look at Genie. Those were two equity acquisition. I know Genie had a token, but they’re wound up being an equity deal essentially. And you look at Opensea, who’s never issued a token. I think it’s clear that if you have a winning product that’s generating revenue is a real business. Why would I give that away?

It was interesting to see that that Uniswap– when they bought Genie, they announced these rewards in USDC. I thought that was very curious. Do you see this as a new model emerging? Why not give away protocol tokens or native tokens?

I think Uniswap Labs is one thing that I aspire to model Astaria after. Their model was very good. I would say with the exception of the UNI token. They did this airdrop for genie token holders that was in USDC. I think that was kind of brilliant because once a token has become emitted, there’s nothing you can do. Like creating a merger requires you to make a contract that you can redeem for. You have to have people like swap for their original token, and then you have to have them actually do it. And you can’t force anybody to do anything. And so I think that airdrop was actually brilliant.

I think that was the way to go and get the tokens off the market and go straight in as equity businesses. Uniswap Labs didn’t buy this for UNI token holders. Right. This is a Uniswap Labs’ acquisition. And the Uniswap token is actually ownership of the protocol, ownership of these V2 and V3 smart contracts like switch their governance and their BSL license. So I think tokens are great for some instances, but everybody is seeing them as out there. Their perception is that it’s like proto equity and that’s wrong. That’s completely wrong. We’re seeing tokens that should be ownership of things that are literally on chain, not this kind of shadow equity. That’s insane.

That’s interesting. Okay. So to clarify it it’s true that at some point tokens did become seen as equity, as a way to get a share of a crypto project. Because of the governance power that they have– it’s like this expectation that at some point token holders can vote to get some sort of dividend and part of revenue on the fact that you can participate in governance, make token holders have some sort of right to the project. So in those ways, it does make them look like stock, but the distinction you’re making is that the thing that tokens should control is not the entire project. In the case of Uniswap it’s obviously just not Uniswap Labs, the company, it’s actually the protocol, the whatever kind of revenue and liquidity that’s on chain. That’s the thing that’s directly linked to the token. Nothing else. Is that right?

Yeah. And I think that their equity is really where all the value lies in any of these companies. It’s whatever is out there in the zeitgeist today for Uni is that’s what the UNI token is for. It’s like the fee switch and BSL license and governance, but Uni’s gonna make more hits, right? They’re gonna make V4 or V3.1 or whatever. Right? And I think that’s the value, the talent and the team and the management, because protocols can be released every day. I mean, I think Sushi did the same thing. It’s just rolled out a copy of V2.

So even to you, even if like all of Uniswap activity happens on chain, to you, the equity is where the value is because the team is what’s valuable in the case of Uniswap, especially in web3 where everything is open source and in the case of Uniswap you’re not allowed to just fork it anymore. And you know, they have this business license, but in most cases, stuff is open source and you can just copy it. So the teams, the actual talent becomes even more important.

I think for sushi, the power was the team as well. It was the team behind it, because initially it was just this V24. And then it evolved. The team was hyper talented and focused on delivering a bunch of different verticals in the DeFi stack. And for another instance is like Yearn, right? There are a million Yearn forks, but the team is what’s valuable. Not the protocol.

Yeah. I think that’s totally right. So how do you think, what’s next? how do you think this space will evolve? What’s the future hold for tokens? Do you think they’ll become valued less as it is? This becomes more apparent or maybe there’s a change in regulation where tokens do become more like equity? I don’t know. Where do you think DAOs as a model are going.

Yeah. I would love to see some regulation pass that allows tokens to be equity. I think it’s very unlikely, but that would be the ideal scenario. I think that’s what everybody’s hoping for so long kind of this pipeline. And I saw this actually the opposite way for a long time, but so long, I’ll say this pipeline of VC to IPO has been, like, we only allow sophisticated investors. And these sophisticated investors’ qualifications for being a sophisticated investor are you’re rich. And they have controlled and extracted the value from this pipeline until acquisition or IPO. And I always thought that that was just a value extraction, but actually now that I’m in this system, having seen tokens, actually, it’s value extraction, but it’s also accountability.

So tokens, a lot of these teams kind like release tokens and then they’re gone, right? There are founders of projects that are gone in six months because they’ve already kind of like exited their position by minting tokens and having them on the open market. I think there’s like a lack of accountability when it comes to. And there’s this very firm accountability when it comes to equity. So I think most of the people…I think the question is, how do we evolve in this space? 

I think the way that we evolve is that we have this really great set of ethos in DeFi, which is like kind of somewhat free and open source, allowing other people to use the platforms, trying to help people who don’t have access to sophisticated financial tools. I think those are great ethos that the space carries forward, regardless of whether they’re equity companies or token companies. And I think most everybody who is talented is kind of going to migrate into like an equity only company.

Okay. So to you, where is there space for tokens, like one showed up a project, have a token.

Totally. Yeah. I think that there are definitely a space for tokens. I think it is like a fundamental building block in this space as well. So I think like when we saw the ICO boom in 2017, we saw a lot of these utility tokens that were really forced and governance tokens were the next step I think in terms of the value of something and governance tokens had this real value to them whereas the utility tokens were kind of not making a whole lot of sense in the system. The tokens definitely have a place. And so we use them as an atom to build new things and not necessarily governance tokens specifically, for instance, on Astaria we have an ERC-20, 40, we have an ERC-46, 26, which is an ERC-20 compatible vault token that was developed by, I forget if it was Faye or Rari, but one of the two, and we’ll use this the way you might use– an integer or a float or a char in any software development, we kind of use it as like a building block. I don’t know that there’s a huge future for governance tokens. I think people will still use them. I think there will be a hybrid model that infers ownership over a thing that’s on chain that you can actually govern, say interest rates or more token issuance or fee switches, that sort of stuff. And that’s where, the token model really makes sense. I think governance over a thing that we can control.

Makes sense. Can you explain a bit more what you meant about using tokens as a building block?

Yeah, sure. I think the most fundamental level, most everything on chain is gonna be a token of some sort. ERC20 or ERC721, because it’s almost like a memory pointer. And so we call them the DeFi primitives because we use primitives in a programming language. This is my two cents. I don’t know if that makes any sense. We’re gonna be swapping tokens. If we’re swapping, we’re not gonna be swapping differing standards. That was like one really clear thing that they did in Ethereum really early. I mean, ERC 20. It was like the 20th EIP. It was like to write a standard for the token interface. So we didn’t wind up with a billion different token types. I mean, we still have quite a few, but they all have to conform to the base set of C 20, otherwise they’re irrelevant. Right.

Right. So you mean tokens are useful as financial assets and they can be fungible, they can be non fungible, like NFTs. Maybe they can represent, I don’t know, like things in the physical world, or?

Joseph Delong is the former CTO of SushiSwap, one of the biggest decentralized exchanges, and the founder of Astaria, a recently created NFT lending protocol. Joseph has been heavily involved in building and managing DAOs. With the recent controversial pro...

🎙Ex SushiSwap CTO Joseph Delong on Why DAOs Fail