The Delphi Podcast Host and GP of Delphi Ventures Tom Shaughnessy sits down with Francesco George Renzi, CEO and Co-founder of Superfluid, a protocol building programmable money streams. The two discuss how streaming solves payment problems today, the potential of streaming as a DeFi primitive, using future cash flow for on-chain credit scoring, and much more.
- Delphi Podcast Summaries
- Superfluid Website
- Superfluid App
- Superfluid Docs
- Ricochet Exchange Website
01:00 • Tom
Hey everyone. Welcome back to the podcast. I’m one of your hosts, Tom Shaughnessy. I help lead Delphi Ventures. I’m one of the co-founders. Today I’m thrilled to have on Fran of Superfluid. We’re tackling streaming payments. We’re also an investor in them (as a disclosure). We’re extremely excited about that. Fran, how’s it going?
01:13 • Fran
It’s going great. Thanks for inviting me.
01:15 • Tom
Tell us a bit about yourself.
01:17 • Fran
I’ve done a lot of different things in my life. About four years ago, I think it was autumn 2017, I landed in a room in Estonia. I just moved into the country and I wanted to meet people. I went to a meetup which happened to be a crypto meetup. I walked in the room and everybody had a screen open with a countdown. I’m like, “What are all these countdowns?” It was all ICO countdowns. I don’t know if you remember that period of crypto. There was a lot of electricity in the air. It was very exciting. I had no idea what any of it was. Luckily, the person speaking at the meetup was very sophisticated and was talking about crypto economic design. I was immediately captivated. He gave me a few pointers of projects I should look at.
02:03 • Fran
Then I fell down the rabbit hole pretty much immediately. I worked on a few (not very interesting) things for about six months and then started working more deeply on the Ethereum development site. I’m not a developer myself, but I taught myself enough to code proof-of-concepts, understand solidity, understand how stuff works in order to then also look at how to design stuff. I’d say I was one of those people who immediately jumped down the rabbit hole. I didn’t find out about Bitcoin or crypto before, but, as soon as I did, I dove in.
02:39 • Tom
How did you come around with the idea for Superfluid? What were you seeing in the market that you thought, “Hey I want to fix this,” or, “Hey I want to change this,”? How did you land on the idea for Superfluid?
02:50 • Fran
Me and my current CTO, Miao, we were working on a different project called rDAI which was a derivative on top of Compound where you could split the principle from the interest and redirect the interest to a third party. It was exciting. It was one of the first DAI derivative tokens. A lot of people were excited at that time from the fact that we had lending markets at all, but what really captivated me and Meow was the fact that we could redirect interest and make it accrue in a third-party account without making any transactions. We started thinking a lot about how this was happening. How were we moving money on-chain without making any intermediary transactions? That captivated us, and, from a technical point of view, we started thinking about how we could generalize that into non-interest-related streams, so just generic streams.
03:40 • Fran
How can we use this idea that we’ve built and transform it into a protocol that can allow us to send money over time? It was more of a technical intuition. We came up with how to do it first and then we came up with what it’s useful for. Ultimately, it’s funny because if I look back I was actually criticizing money-streaming on Twitter. If you look back at my tweets in 2019 before we came up with Superfluid, I was openly saying that that stuff didn’t make sense, and that was more about how it was previously implemented. Superfluid has some characteristics that previous attempts at money-streaming don’t have which, in my opinion, make all the difference. So, while I was previously a critic of the idea, I then immediately flipped into allowing the idea and now we’re working on it every day.
04:30 • Tom
What’s your thesis on why payments today are broken? We live in a society where everything is kind of a recurring subscription or “Check how many I have at the end of the month on my card,” “I probably just forgot about all of them,” but, what’s your take on rearchitecting our world from recurring subscriptions and “Here’s a card” to something that’s programmable or something that’s streamed? Do you have a thesis around why that’s broken in the traditional world?
04:55 • Fran
Absolutely. First of all, I think it’s super exciting that we’re even talking about this now that we have a protocol that can actually solve these problems. It’s pretty amazing. It’s testimony to the fact that crypto is a land of opportunity and innovation. I wanted to mention: Silicon Valley has been talking about subscriptions for ages. There are entire VCs that have built a thesis around the fact that everything is going to become a service, everything is going to become recurring payment, everything is going to become a subscription…but they never actually went and looked into how those subscriptions work. There hasn’t been any fundamental change in the way subscriptions work since basically ever. In Europe, everything is monthly. In the U.S., most things are monthly. Salaries are biweekly, but, ultimately, it still works exactly the same as if you had to physically carry a check to your landlord.
05:44 • Fran
Why is it that card payments still follow the same schedule as old paper-based money transfer systems? Why is it that you use the service all month and you pay it at the end of the month? Why is it that you work every day but you get paid every so often, right? Realistically, this is purely a habit. It’s just a tradition. It’s something we got stuck with. Again, I trace it back to the fact that you really don’t want to see your landlord too often, right? So giving them a check once a month is often enough, right? You don’t want to see them more than that, and that kind of carried on into everything else, right? It was the friction of the payment method that we started with that carried on into modernity even though now we could very easily automate that and make it a more frequent payment.
06:31 • Fran
Unfortunately, I think a string of payments are definitely a good technical solution, but they’re not going to help us to change people’s traditions and habits. That’s something that we have to do ourselves. That’s something that everybody looking at this technology and its opportunity should think really hard about, “Why are they waiting two weeks or a month to get paid by their employer?” Why is that the case? Why are they accepting that? And, ultimately, they shouldn’t. If you think about it, every delay in payment carries a risk. If you were working two weeks or a month for your employer and they’re paying you at the end of the month, effectively you’re giving them a free-of-interest loan, right? You’re giving them a loan. They’re not paying you for that loan. At the end of the month, they could not pay you. You’re shouldering all of the risks of that transaction and the employer ultimately could be default before and not pay you at the end of the month.
07:20 • Fran
If you were paid more often, the risk would be reduced. Reducing the risk in service provision, reducing the risk in employment, Reducing the risk of every delayed payment is a huge opportunity because, ultimately, risk makes things more expensive and reducing that risk also increases the liquidity and capital efficiency. Ultimately, we believe streaming payments are strictly a better way of paying. Although, if you look at it from a micro point of view, they may seem they’re worse for one of the two parties in a payment. They are a net improvement if you look at it from a global point of view.
07:56 • Tom
That’s really interesting. The capital efficiency point makes a lot of sense because people are getting their money much sooner. I don’t have to wait two weeks. They can get it streamed. Do you think that culturally in a Web2 and traditional environment, will people welcome this? It seems to me in a Web3 sense it makes more sense. Somebody is part of a DAO, they’re interacting, they’re earning tokens, and potentially they can earn it in a streamed manner so that they can get rewarded and see that they’re actually earning something for their work. Do you think that Superfluid will be more prominent in a Web3 sense or do you think it’ll be also as prominent in a Web2 sense? Also, just to play devil’s advocate, I mean we love Superfluid -we use it to stream salaries for our employees at Delphi as well. Obviously, we’re a user of it… but I’d love your take on the dichotomy between the Web2 and the Web3 user Superfluid.
08:43 • Fran
I think you’re onto something. Web3 is always changing things radically. People inWeb3 are used to disruptive behaviors, and this is all this is, ultimately. Sure, there’s the technical side, but getting people used to this is going to be the hard part. But, in Web3, people adopt new technologies because we’re able to question realities that we got used to. You were mentioning the interest case right. In DeFi/Web3, we’re used to getting interest that compounds every block. That’s just the fact, right? You wouldn’t accept an interest that doesn’t compound every block because now that’s a standard. If you look at how the banks work, they still give you a statement at the end of the month, and it’s a monthly compounding. If you look at people in traditional finance, they’re happy with that.
09:30 • Fran
People in Web3 would never accept that. It’s a matter of showing people that things can be better, and, once they get used to that, they’ll never want to go back. If you’re getting paid every second, if you’re able to set up your financial life to get paid every second, invest every second, pay for your expenses every second, ultimately, it’s so much better for you. It’s so much more convenient that you will never want to go back to a monthly paycheck. That’s the way we push it. I think that’s the way to do it. Rather than trying to convince people, just improve things dramatically – get them hooked so they never want to go back.
10:08 • Tom
You’re totally right. Web3 people are accustomed to having streamed payments because of how we interact with DeFi and Web3 where we get interest per block, we yield farm, we earn per block, and we see everything happen very quickly. Do you think that Superfluid will be used for DAO-to-DAO value exchange as well? Or, is that less important? I’d assume it’s as important because of all the flows that the end-DAOs have to pay out to say their contributors or to their expenses, so it makes sense to have just-in-time streams as well. Curious to hear what you think there.
10:41 • Fran
All sorts of payments will be streamed in the future. As of now, it’s still a bit of a mix. At the moment, it’s mostly companies or DAOs paying their contributors. We haven’t seen much DAO-to-DAO payments, but that is more because there aren’t that many DAO-to-DAO payments yet, and when there are these kinds of payments, they are often settled in governance tokens rather than money. I think that might be one of the reasons. Going forward, this is definitely going to become the case. We’re speaking to some DAOs who have a multi-layer structure. They have a big DAO and then they have subgroups or units or whatever they want to call them that handle specific things and have a budget. For those kinds of DAOs, having internal strains makes a lot of sense.
11:27 • Fran
It’s a way of budgeting whereby you say, “I want to give this business unit a certain amount of money.” This effectively, the moment they open the stream, it will start flowing and it will keep flowing until it stopped. So, it’s a way of making recurring payments that don’t require more than one transaction. This is very good for DAOs because DAOs have very high transaction costs. They need a lot of people to coordinate to get around the table and vote. In order to streamline that, having only one decision to be made that will initiate the payment that will continue potentially forever is a very big improvement. That’s the reason we’re very excited to see DAOs adopt this for things like budgeting, and then smaller DAOs to adopt this for things like contributor payments. DAO-to-DAO payments I’m a hundred percent sure will come.
12:20 • Fran
We are starting to see those DAos as guilds. I think as we start seeing more DAOs coordinate around specific functions we will start seeing them work for different DAOs and getting paid by different DAOs. Ultimately, they’ll want to get paid every second. I’m pretty sure that will happen.
12:42 • Tom
That’s really cool. I totally agree with you on making that happen. Let’s walk through an example. If I have your app up – app.superfluid.finance for those that want to follow along (you can try it out on Polygon or some test networks) – how do you get started with Superfluid? If I have an amount of money in my wallet, how do I set up these streams? How many transaction fees do I have to pay to get this started? Worst case scenario, let’s say I want to cancel something. How do I go about doing all that?
13:11 • Fran
It’s worth going back a bit and seeing what Superfluid is. Superfluid, at its core, is a token standard which defines the flavor of ERC-20 which has an extra functionality which is streaming. Superfluid tokens are streamable. Going back to your example, you go to our dashboard. If you want to send a stream, the first thing you need is tokens that are streamable. The most likely situation is that you don’t have any in your wallet. If you want to stream USDC, what you would do first is wrap that into the Superfluid version. We call them super tokens. Once you do this upgrade from normal USDC to super-USDC, then you basically are in a situation where you can stream. At that point, if you know who you want to pay and how much you want to pay them, you can simply open a stream.
14:03 • Fran
That’s one transaction. You open a stream specifying who the recipient is and how much you want to pay them per second. Once that transaction is confirmed, basically the payment will go on until you stop it, or until you run out of funds. Imagine you’re paying all your stuff. You upgrade a bunch of tokens to super tokens and start all the streams you need. As long as you keep topping up that account with super tokens everybody will get paid and you will never need to touch the streams again. The day you want to fire someone or maybe you want to give them a raise, all you have to do is upgrade or delete that stream. As soon as the transaction is confirmed, the stream stops and basically their account stops increasing and your account stops decreasing. It’s as simple as that. One account can have any number of outgoing streams. An account can also have any number of incoming streams. Incoming and outgoing streams can co-exist which means if you’re a business that has subscription revenue and you’re also paying salaries, the money that comes in to pay your subscriptions can simultaneously be used to pay for your employees. The money flows in every second and flows out every second in a constant manner. That’s where the capital efficiency comes into play.
15:19 • Fran
Generally, if you think of a normal business, you’ll get paid throughout the month and at the end of the month you’ll pay all your payroll. If some of your invoices don’t get paid, you might have a capital problem. In the case of Superfluid, you get paid every second and you pay every second. If somebody stops their stream, you know immediately and you don’t have to wait until the end of the month to adjust it. When things start getting very exciting with Superfluid is when people have both incoming and outgoing streams because you start seeing this extreme capital efficiency kick in and you never need to necessarily top up your account because you are receiving funds in that account at the same time as you’re spending them. You don’t need to actively top up the account in the same way.
16:00 • Tom
Can any token be wrapped as a super token? Like, if I have a native governance token for a DAO or a project that I want to stream to my contributors for their work, it doesn’t have to be USDC, right? Can anybody go and wrap the token that they want on Superfluid?
16:16 • Fran
Pretty much any ERC-20 works. We have some limitations on rebasing tokens, but they can be overcome. You can also issue your token directly using Superfluid. That’s something a few projects have experimented with. If you know you’re going to use streaming functionalities and you’re not afraid to use something cutting edge, you can basically issue a token directly through Superfluid. In that case, you never have to upgrade or downgrade. 100% of the token supply is streamable, and that’s really the best experience. We’re working with a few projects so that when they bridge from mainnet Polygon their token is immediately streamable. That’s really the best UX for us, but obviously a lot of tokens have already been issued, so the wrapping / unwrapping part is necessary. Over time, we think as more people use these functionalities they’ll get used to it. As they use more apps in the Superfluid ecosystem they’ll have less of a need to upgrade and downgrade so often.
17:15 • Fran
If you get paid in Superfluid you can also make payments in Superfluid and you don’t have to downgrade or upgrade the tokens at all because the money can just flow through your account.
17:27 • Tom
That’s really cool. The upgrade / downgrade process is pretty interesting. For those that want liquidity or want to exchange their super tokens, they would have to downgrade them and then exchange them for the asset that they’re looking for, or is there a way to exchange super tokens to other super tokens?
17:43 • Fran
We’re working on a way to exchange super tokens for super tokens. You can also use a DCA app. If you’re interested in trying a new app, (I really suggest this) go to ricochet.exchange. It’s a really cool application built on the Superfluid Protocol. Basically, what it allows you to do is send a stream of tokens which will be transformed into a different token, so you are dollar-cost-averaging into different assets. Their most active markets I think are USDC to Ether, USDC to Bitcoin, DAI to Ether, and DAI to Bitcoin. You can go from stables to Ether or Bitcoin. You’ve got liquidity mining going on. Basically, what you’re doing there as a user is one action: you’re opening a stream. That’s it. Now, after that, the contract basically accumulates funds for about 20 minutes and every 20 minutes makes a purchase of that token to you. After they make the purchase, they use a different functionality of Superfluid to basically massively redistribute the tokens to everyone who has been streaming.
18:47 • Fran
This is another functionality of Superfluid that’s a bit harder to explain. We don’t talk about it too often, but basically it allows you to send distributions of tokens to any number of recipients. The way Ricochet works, even if it has hundreds of subscribers, it will always have a fixed cost to distribute the assets after it purchases them. So, it’s a very scalable system. As I was saying, they’re still early stages, but we already have some very active markets. I think they’re consistently draining the liquidity on certain markets on the Polygon because they’re not as well arbitraged as the ones on Mainnet. So, they’re starting to get some volume. I think it’s a very exciting project if anybody wants to have a look.
19:26 • Tom
We’ll definitely link to it in the show notes. To zoom out, I think people might be interested in the secret sauce here, right? Everybody in crypto pays insane amounts of fees to use Ethereum. Granted, we have L2’s coming, and some are already live. What is the secret technical sauce for making streaming happening without having to pay transaction fees over and over again? You mentioned super tokens and the standards. Where is the technical magic here that happens that enables all of this?
19:55 • Fran
I’ve tried to explain it a bunch of times and sometimes it doesn’t come out right. Basically, what happens with Superfluid is that your balance is not settled by transactions. Rather, transactions settle what we call agreements. If you think of transfers, basically what you’re saying is, “I want to send $10 to Tom. Take out $10 from my account and add $10 to Tom’s account.” That’s a simple action that we can describe, and that’s essentially what the transfer function does. That’s what transfers have looked like since we’ve used ledgers. Even at the time of paper ledgers, that’s basically what a transfer looks like. Now, when you send the stream using Superfluid, the transaction is not what’s moving the money. The transaction is simply giving that similar instruction to the contract, and the contract will then execute that transfer over time without the user needing to do anything else.
20:56 • Fran
The instruction you’re giving to the smart contract is, “Transfer $1 every second to Tom.” Once you’ve sent that information, you don’t need to send another transaction to the chain until that information changes. The moment I want to stop sending $1 per second to Tom, I have to send another transaction. Until that moment, the contract has all the information it needs to calculate what my balance is, what your balance is, and what the global state is. Until I change what the state looks like, I don’t need to touch the blockchain. So, for anything where we’re able to very simply describe what the future state should look like, we are able to settle agreements rather than settling single transfers. I don’t know if that made sense, but it’s always a tough one to explain.
21:50 • Tom
It’s a highly technical thing that looks like magic to people on the outside. So, instead of doing an L1 transfer where you’re signing a transaction and paying a fee, you’re streaming within a smart contract. You only edit that if you want to stop it or modify it.
22:05 • Fran
Exactly. I think the essence of the fact is that nothing changes. When you say, “I want to transfer this amount every second,” then until that intention changes, you shouldn’t need another transaction. That’s basically how Superfluid is built. As long as your contractual commitment stays the same, then there’s no need to touch the blockchain. The moment you want to change your current commitments, then you need to touch the blockchain. Effectively, by innovating at the token standard layer, we’re able to change the rules of settlement. Generally, settlement has only been happening in one specific way, but by rewriting the token contract itself, we’re able to basically reinvent what settlement means. So, in the case of Superfluid, settlement is effectively happening every block. We’re accounting for streams over seconds and we’re settling the amounts every block. This is only possible because of programmability building blockchains, but also because we’re, let’s say, “crazy enough,” to go and change what the token standard itself looks like.
23:11 • Tom
That’s really cool. We spoke earlier about paying salaries or incentivizing people with tokens in the stream manner, but it sounds like the possibilities within DeFi are pretty cool, right? I’m thinking about people getting liquidated on a DeFi app when they’re taking out leverage and being able to stream collateral from your own pool of capital to continue top up so that you don’t get liquidated. That’s off the top of my head, but there’s a lot of examples here where I think you could work with DeFi really well.
23:37 • Fran
The way we think of it is really as a primitive. If you think of your ERC-20 token, until now, it had one single function which was transfer. That’s all you could do with money. Imagine that everywhere you’re using transfer you could now also use stream. So, as you said, you could post collateral in a stream, but you could also borrow in a stream. You could also lend in a stream. You could also repay your loan in a stream. Maybe you can even exchange in a stream, right? You can add liquidity in a stream. You could yield farm in a stream, right? You can do anything in a stream. Some things might make more or less sense, but what’s really interesting is not so much rebuilding exactly the same products, but it’s more, “What are the unique things that we couldn’t do with transfers but that we can do with streams?”
24:23 • Fran
That’s where things start to get really exciting. The way I see it, streams are a way to make payments in the future. That’s something we could never really do with transfers, right? Transfers happen the moment you signed a transaction. Streams happen from the block after you sign until you stop them, right? So, it’s a payment in the future, and this could open up a bunch of different opportunities. From a very basic level, you can think of your relationship with applications being continuous rather than spot. There’s a lot of financial crazy stuff that we’re working on internally. We’re probably going to be releasing some research next year with some financial products that we are working on. There’s some very interesting things you can do with streams. I’m really excited to start unveiling some of this stuff and getting people excited about the possibilities.
25:12 • Tom
I’m amped. Definitely it takes some time for the holidays, but you guys have spent so much time building a primitive that now you get to the fun part where you could start releasing your own ideas for how it should be used. I’d love to talk about the traction you’ve seen. Are there any use cases that you’re seeing that you didn’t even envision? Like, somebody comes to Superfluid and they say, “Hey Fran,I want to do this or that, and you guys are the standard that I’m going to use for it.” Has any of that risen up yet?
25:38 • Fran
Traction-wise, at the moment, it’s mostly salaries from contributors and people doing DCAs. We’re starting to ramp up on the subscription side. If anybody in the audience is interested in that, there’s a couple of projects built on Superfluid that are looking at the subscription side. I think that’s going to be a very big opportunity. At the same time, there’s some products that we absolutely had not thought of, especially on the finance side because nobody in our team has a deep financial education. So, on that side, we’ve seen some very interesting things coming from some of our community members. On the payment side, also, you notice in DeFi that you have people that come from very different backgrounds. Some people have some experiences in traditional finance or some problems that they’ve had with traditional finance, and then they see something and they’re like, “Wow! This solves everything. This is the solution I was waiting for.”
26:33 • Fran
There was one person that we spoke with last spring who had a lot of experience with affiliate markets. They worked for a long time in gray area businesses. Gray area businesses, as you might know, have a very hard time getting bank accounts. They have a very hard time getting access to any financial services. Affiliate markets are one of the ways they have to promote themselves because they’re also gray area services, but they’re also extremely slow. They have 90 day payment terms. They do a lot of unnecessary KYC AML sometimes, and effectively create a lot of friction because they don’t really want to pay out. What he came up with was a very cool system whereby they use Superfluid for subscriptions. So, if you imagine you want to subscribe to a service, you use Superfluid to subscribe for the service, but you pay every second rather than paying once a month.
27:28 • Fran
Whoever referred you gets a percentage of that stream streamed directly to them. So, what this does is it immediately cuts that 90 day payout that traditional referral marketing has to a zero second payout. You get paid immediately. As soon as the business gets paid, you get paid as well as a referrer. This is already very impactful for anyone that’s ever done referral marketing. Where it gets really exciting, especially for me, was when we realized that that income that the referrer is getting could also be tokenized. By tokenizing that referrer income, that person who referred people can then sell that future income on a marketplace to anyone who’s willing to take on that risk. At that point, what you have is an NFT that represents future cash flows that you can trade anywhere in DeFi right? And, obviously, it’s all open standards, right? It’s a traditional NFT.
28:24 • Fran
You can see it in MetaMask. You can transfer it and stream transfers with it. It was one of the first very unique applications that we saw. I thought it was very interesting how, by using the NFT standard, they were able to create a transferrable future cash flow. Basically I think this was built about two weeks after we released. It was one of the really first hackathons that we were sponsoring, and we found a very exciting project.
28:52 • Tom
That’s awesome. It’s funny you mentioned no one on your team has deep financial experience because you’re rebuilding how we interact with cash flows from the ground up which is insane.
29:03 • Fran
It’s exactly why we are doing it. I think the more you’re into something, the harder it is to look at it from the outside. That’s the reason you see a lot of traditional finance people that look at DeFi and say, “This doesn’t make sense.” Well it does. It works. It obviously makes sense, but they can’t make sense of it because they’ve been too deep into other ways of thinking and understanding certain things. We tried to explain to a lot of people that we talked to, “With Superfluid, settlement is continuous,” and they’re like, “What do you mean? Settlement doesn’t work like that.” And we say, “It does. It’s exactly how we’re paying salaries to our team.” They’re like, “No, you’re wrong. Settlement happens later.” And I say, “No, it happens every second.” They can’t get it because I guess you get used to how things are supposed to work and then you can’t look at them from the outside.
29:55 • Tom
You’re totally right. Being on the outside makes it easier. The hackathon project you brought up is interesting. Basically, you have the smart contract streaming money, you take that, you discount it based on its future expected value, apply a discount, etc., and you could sell that. That seems really cool, not only from the optionality for users or DAOs like, “Hey you can sit here and wait for your stream or you can get it upfront or you could leverage it maybe in some other way.” I feel like that has the potential to take off pretty dramatically not only with salaries but within other use cases of DeFi too.
30:29 • Fran
Absolutely. I think the potential for securitized cash flows is pretty much unlimited. Speaking with people in traditional finance, they’re all excited about this. One of the reasons they’re excited is because DeFi generally manages to take this stuff and make it reduce the friction a lot. So, in traditional finance you have things like cash flow financing. You have things like payday loans, but they’re generally very expensive. For example, cash flow financing really only works for very big businesses because the cost to do due diligence, the costs on the regulatory side, the costs to make it work for a financial institution… is very high. They only do this for companies that have millions of dollars of recurring revenue. Now, imagine in DeFi we can bring this down to very small businesses, right? We can bring this down to micro enterprises and allow them to borrow from their future earnings.
31:24 • Fran
That’s huge. We can do that because in DeFi we have transparency. We have auditability. What that means is, if you’re observing the chain, you can see how much money I’m making. You can see who’s paying me. You can see how long they’ve been paying me. You can see all the details, which means you can run an algorithm on the data that’s on-chain for every account and immediately decide which ones are worthy credit or not. You can very easily access the information you need to build a credit score on-chain. The moment you have a way of scoring credits, you can much more cheaply provide credit. So far, nobody’s even managed to provide uncollateralized lending to DAOs or even to businesses in the space. So, as you said, it’s something we’re extremely excited about. I think future cash flow-backed loans will be the first step in providing credit on-chain because it’s not fully uncollateralized. It feels like it’s collateralized, although it’s collateralized by something that doesn’t exist yet.
32:23 • Tom
One concern for somebody that is going to buy a future cash flow is the credit worthiness of the other side paying. Is there any way to DD or check on-chain or some kind if the sender of the stream has enough money to pay that in the future? If I’m getting my salary every month, and I wanted to discount that back and sell that, obviously it’s contingent on the person sending the stream to have enough money to pay that salary. Is there any way to do that due diligence for that person? I know we’re digging into this specific hackathon example. It’s just super interesting to see what could be built on you guys.
32:59 • Fran
The short answer is no. There isn’t really a way to do that because the protocol is built for complete capital efficiency. You can stream money having basically nothing in your account. You can literally have a balance of $0 and stream a million dollars because the protocol doesn’t care about your balance. It only cares that your balance isn’t negative. As long as you have more than zero in your account, you can still stream. If you have a million dollars of income and a million dollars of expenses, your streams will keep flowing, right? So, it’s very hard to say that, “That’s not a legitimate user that can afford the payment to the business,” because they have income. How far up the chain do you go when doing due diligence? That becomes the question. I think it’s very hard to really assess that. You can take different measures.
33:50 • Fran
You can say, “How long has this person been streaming?” If they started streaming a minute ago and now they’re asking for credit maybe we should not consider that stream, right? Because maybe they just created it a minute ago to try and change the metrics. “Has that stream been open for a month, and do they have an average 10 subscriptions to different products? Well then, maybe they’re a legitimate user, and we should consider that stream when calculating a rating.” This is all stuff that people who do machine learning can figure out in a few minutes. Chain analysis companies can repurpose their software for something useful and start doing credit checks instead of other things.
34:26 • Tom
I totally agree. It’s interesting. It’s cool to see what can be built on you guys and what people think of. When you’re building your primitive, you just have no idea how many and the types of apps built on top. One thing we didn’t talk too much about is Super Apps. Can you talk to us a bit about what a super app is and how that leverages Superfluid?
34:44 • Fran
Super Apps are a concept that we introduced when building the framework. When we were about to release, we realized that to actually build an application that uses streams, you need to have more control over the life cycle of a payment. If you look at a normal application in DeFi, a normal smart contract, it only has to deal with users when users transfer the money, right? So, if you think of Uniswap, the way it works is you send money to Uniswap, Uniwap sends you money back, and then you go your separate ways and you don’t have an ongoing relationship with the application. In the case of Superfluid apps, you open a stream through the app and then your balance is connected to that app’s balance, right? Until you close that stream you are effectively in a relationship with that app.
35:31 • Fran
What this means is that the app needs to be able to account for every cent that it’s receiving until you close that stream. When you close that stream, it has to know that you closed that stream. It needs a way to effectively manage the full life cycle of that financial relationship. Super Apps are basically a framework to build applications that know when they’re receiving streams of money and can react to those streams changing. In a kind of less technical way, we can say that any contract can receive streams, but they won’t know they’re receiving streams unless they are Super Apps. They will simply see their balance go up, but they won’t understand why it’s happening. They won’t know when the stream stops. If you are a Super App, you can know that a stream started, you can know that the stream is updated, and you can act when the stream is closed. This allows you to build proper accounting which is ultimately the basis for any financial product.
36:26 • Tom
Yeah. It just sounds like building a very capital efficient app from the get-go… that might be an oversimplification though. *Laughing*
36:31 • Fran
That’s it. The simplest app we built, (which was the first solidity smart contract I wrote actually because I was super excited to build on Superfluid, so I taught myself solidity and I built this app)… it does one simple thing. It takes all the incoming streams and it sends them to the owner of the contract. Imagine it has a million incoming streams, and it packages all of them into one stream that goes to the owner of the contract. Basically, it’s a super app, because, if one of those incoming streams is closed, the outgoing stream will be updated to reflect the total income. Now, why is it interesting? Well, that smart contract can manage any amount of cash in streams, right? So, it could have a million coming in and a million going out, and that super app always has a zero balance.
37:22 • Fran
It’s managing millions of dollars in volume without having any balance whatsoever, right? That’s something that I don’t think you could ever see in a traditional smart contract where it’s all about TVL and value locked and storing somebody else’s funds. Basically, it’s a smart contract that’s moving money without ever storing it. That’s something that we got very excited about. That contract is also transferrable, so I can basically take the ownership of that smart contract and transfer it to you, and suddenly you’re the one receiving all of those streams. It formed the basis for a lot of other contracts that were built on Superfluid at hackathons.
38:00 • Tom
That’s awesome. One thing I wanted to talk a bit about was: we’re in this transition where we have a lot of L2’s, either live or launching. There’s Starkware, Arbitrum, there’s Polygon, there’s Immutable built on Starkware, there’s a bunch. There are other L1’s ones like Solana that are faster and cheaper. I’m curious your take on how Superfluid interacts with L2’s or if you can leverage them or you’re built on them which I believe you are in some cases… but more importantly, just the human aspect, right? Do you think if people are able to send money for free on an L2, will they care as much about the streaming aspect? I know this is very different because we’re debating transaction costs versus seeing a balance automatically accrue every second on my computer, which is obviously huge, but I would love your take on the L2 side. I know there were a lot of questions there, but take it anyway that makes sense.
38:54 • Fran
First of all, we love the EVM. We think Superfluid will probably live on every EVM out there. We are gearing up to be able to deploy on everything that runs on Ethereum Virtual Machine. We’re very excited about roll-ups. I can’t tell you how excited we are about it because they’re just going to change everything. Having cheaper transactions, having cheaper DeFi, all this stuff that we’re talking about…realistically it just simply can’t happen on mainnet, right? It just can’t happen. It’s too expensive. But, the moment you have very cheap transactions then everything we’re talking about starts not only making sense, but being able to, again, enable use cases which are much closer to us than just whales leveraging their millions of dollars on mainnet. It’s very exciting for us to deploy on L2s. What I can say about the cost: streams are not really about saving gas costs.
39:50 • Fran
It’s something that a lot of people obviously think about it. It’s kind of the first thing people think about it because people come from very painful times on mainnet, but streams are more about saving mental transaction costs. When you make a payment, you have to do something, right? You have to go to your wallet, sign a transaction, think about it, and also pay gas costs. When you open a stream, you think about it once, and then that stream goes on forever. We started paying some of our team in streams, we set up that stream and I haven’t had to touch it or think about it ever since. That’s a non-custodial payment that’s happening every second without any interaction from my side. Some of our employees are then taking that salary as they receive it every second and investing it in Bitcoin and Ethereum with streams without having to touch them every second right.
40:40 • Fran
They set it up once and then that stream goes on and buys Bitcoin and Ethereum for them automatically on Ricochet. Being able to automate future payments – there’s a lot more to it than simply saving on gas costs. It’s about de-risking transactions. It’s about paying for things in the future but deciding on them now. It’s about having transparent on-chain information regarding cash flows so that we can build financial products. Honestly, I mean the cheaper gas costs become the more excited I get because we can start making cheaper streams. We can start using streams for more use cases. Overall, when there is more space, people will play bigger games. I think things got a lot more exciting with very cheap transactions.
41:26 • Tom
I totally agree. I think the L2 unlocks a lot for you guys that you can do. It makes everything cheaper. It makes everything easier. I’m pretty excited about that. Fran, to close out, I’d like to talk to founders about their journey. I would love to learn…you know you’re building something that’s really hard for people early on to conceptualize. Obviously, people get it now because they can use it. I’d love to hear: what was the biggest hurdle in creating / founding / building Superfluid? I know you probably can’t share anything internal, but what was the biggest roadblock, and how’d you get past it? What advice can you give other founders building in crypto?
42:00 • Fran
It’s quite funny, actually. One of our biggest hurdles at the beginning was convincing cry[tp VCs and investors that there was a future for on-chain crypto payments. In hindsight, it’s incredible. When we were first talking to investors back in 2020, and we were talking about on-chain payments, they thought we were completely crazy. They thought this would never happen. They thought Ethereum was only for DeFi. They thought Ethereum was only for speculation. Basically, the narrative was completely not there for payments, right? So, crypto investors thought we were crazy. When the two most degen people in the room think you’re crazy, you start thinking you’re crazy as well because they’re supposed to have a very high risk appetite, and they don’t believe that what you’re saying makes any sense. So, that was very hard, honestly. I remember I was at ETH Denver 2020, and I was pitching people, and everybody thought what I said made no sense.
42:58 • Fran
That was very hard at the time. Funnily enough, we found some people who believed in us who were not crypto investors because, in the normal world, payments are huge. They saw the value in that at the time. They saw the value of what we were building and the innovation that we were bringing beyond looking at the narrative of the time, right? Sometimes, what I often suggest to a group of the founders is, “Yeah, sure, the flavor of the month is metaverse / NFTs, but you can think beyond that, right? What’s the flavor tomorrow? Are you solving a problem not a popular problem?” That’s probably a good thing to try and solve for. It’ll have less competition. There’ll be fewer people trying it out right now. Sometimes, doing things that aren’t necessarily the fashion of the month can be rewarding longer-term. Sometimes, I think crypto is a bit too focused on hype cycles, right? Hype cycles are good for investors, but they’re not good for builders because builders get distracted by hype cycles and development cycles last longer than hype cycles.
44:05 • Fran
If you start building everything that’s fashionable, you probably finish your product by the time it’s not fashionable anymore. You really should try and be more radical in your decisions and not follow hype too much. That also goes on a financial side where I always suggest that people do not try to build and trade at the same time because you’re going to fail at both. If you’re a builder, just build and HODL. If you’re a trader, then do your thing and live on Twitter. If you’re a builder, you should just build. I think too many people get distracted by too many shiny new toys in DeFi and leave aside the beauty of solving hard problems in scalable ways.
44:45 • Tom
Totally. Fran, it’s been a pleasure having you on to share the story. We’ve been waiting a while and I am super amped that we were able to do it. We’re proud to be investors in you and Superfluid. We think that as a primitive you guys will unlock tons of new use cases. I mean, not only the ones we talked about, but just excited to see what rolls out. For those looking to get involved, Superfluid has a great doc outline at docs.superfluid.finance. You could also check out the app. I mentioned we’ll link all this in the show notes as well. Thank you so much for coming on.
45:13 • Fran
Thanks a lot. What I would say is anybody who is interested, join us on Discord. That’s discord.superfood.finance which is the best single place to join our community and join in the fun of designing new financial stuff. It’s going to be exciting where we go from here.
45:27 • Tom
I agree. Thanks again.
45:30 • Closing
Hey everyone. Thanks for listening to the podcast. If you enjoyed it, please support the show by hitting subscribe on iTunes, writing a review, or sharing this episode on Twitter and LinkedIn. Stay tuned for our next episode out soon.
(00:00:00) – Introduction.
(00:01:17) – Francesco’s background.
(00:02:40) – The inspiration for starting Superfluid.
(00:04:31) – Problems with payments today / Why streaming payments are better.
(00:07:57) – Potential pushback from Web2.
(00:10:22) – Superfluid usage in DAOs.
(00:12:47) – Getting started with Superfluid.
(00:17:27) – Upgrading/downgrading tokens.
(00:19:40) – How gas cost savings are achieved.
(00:23:11) – Uses for streaming beyond payments.
(00:29:58) – Future cash flow and on-chain credit scoring.
(00:34:36) – What are Super Apps?
(00:38:02) – Superfluid and L2s.
(00:41:34) – Biggest challenge building Superfluid / Advice to founders.
(00:44:45) – Closing thoughts.