How to Fix DeFi with Dan Elitzer

Bangless Nation, it's time for State of the Nation. The topic today is DeFi broken. That's the question we raise. It's from a post that was written that caught our attention. Why DeFi is broken and how to fix it. Part one Oracle Free Protocols. Huh. How is DeFi broken, David? And how do Oracle Free Protocols fix it? Actually, that's gonna be the topic of the day. So don't give me the answer. But why are we exploring this topic? Give us some context.

Yeah, the guests on today's show, bank longtime bankless listeners will uh become completely familiar with Dan Elitzer, who has written a number of articles that have, I think, all turned into bankless episodes because he has the uncanny ability. He has got he's got a great hit rate. Uh so when Dan writes an article, uh, I pay attention. He doesn't write too many. Uh he he conserves his bullets for that when uh he shoots one, it really means something. And he shot a bullet. And so we are going to go and explore why Dan thinks DeFi is broken and what DeFi needs in order to really uh not lose $3.8 billion like it did last bull market.

Oh, oopsies.

Yeah, yeah, 3.8. Uh, that was a number. Uh, so that's the topic of today's show.

So is DeFi broken is the question today. I I really feel like um man, it's been a while since we've done an entire podcast dedicated towards DeFi. And yet what's ironic here, David, is that's where we started. That's why we're here. Yeah. The reason bankless started is because we were like, oh, you know,

we're not the industry isn't talking about DeFi enough. Like the thing that actually makes us fully bankless. And yet I feel like while our protocols have uh experienced tremendous success and DeFi now has clear product market fit.

There's something missing. Right. Like we're we're we're are we working on a new narrative? Everyone's talking about okay, this is the next bull run, that'll kind of fix things. It still feels like we've got some things to build in deeper in the in the stack here at the kind of the core primitive layer. So I think that's on on deck for today. Uh yeah, why else are we talking about this? What what's more context?

I I think this episode will pull in some themes that we've been talking around on Bankless since day one. Uh the protocol, when I was reading Dan, uh Dan's article, the protocol sync thesis came came to mind. Um, the other image that came to mind is you know, that um that meme of like apes evolving into men, and then like there's like it's like the Darwin picture. There's like seven of them. Uh, but then there's the joke of like the last one is like, wait, we've gone too far, turn back.

Uh

I feel like that's what DeFi has been for like the last like 18 months. Like we skipped a step. And I think that's what Dan is saying. It's like, wait a second, we've gone too far. We skipped a step. We have to go back.

yep.

And we have to go back to progress forward. And so uh the words primitives are going to be said in this podcast. I'm going to bring up the protocol synthesis, which we'll define later in the show. Uh, and uh risk dependencies and oracles. These are all gonna be themes uh about why Dan thinks that perhaps we should kind of.

kind of go back to the drawing board, at least in a couple certain respects, just to be able to actually progress forward in the world of DeFi innovation. Uh so that that's the the uh context I would leave bankless listeners to go.

I think the theme is back to the basics. This is this is gonna sound a little bit like a 2020, circa bankless 2020 type of episode where we're talking about these fundamentals. Remember that? I can't wait to dive into this. Guys, we're gonna get right to Dan in just a second, but before we do, we wanna thank the sponsors that made this episode possible.

Bankless Nation, I would love to introduce you to Dan Elitzer, a partner at Nascent, a VC firm that specializes in deep crypto economic primitives. You may remember Dan as the guy who wrote, I don't know, about the threat of liquid staking derivatives back to when we he called them the death of Ethereum, as in delegated ether. Uh, now we call them LSDs. Uh, and that was three years before proof of stake was even a thing. Uh, maybe you remember him from writing about superfluid collateral in 2019, which was a huge precursor to what became yield farming, which he also wrote about yield farming called aquapophonic yield farming in DeFi.

Uh, I think in June of 2020, right before DeFi summer, and then also wrote UniChain is inevitable, is inevitable, a thesis on the inevitability of app chains. All of these things uh have more or less come true with uh varying degrees of success. I would say very strong degrees of success. Uh, and all of them have also become Bankless podcast episodes. And Dan has also recently released an episode why Def uh an article, why DeFi is broken and how to fix it. And so here we have Dan on the show today. Dan, welcome back to the Bankless for I think like the fourth or fifth time. Uh it's good to have you back, my man.

Hey guys, always great talking with you.

Thanks for helping us produce all that content too.

I feel like you've owned a earned a co host

here or something.

Hey, any anytime I'm always happy to chat.

But uh on that note, like what when you how do you decide when you are going to write an article? Because you you like I said in the intro, you conserve. You don't, you don't, you know, it's not you're not like Dan Ellis every single week. You're not shooting your shot left and right. Uh what does it take for Dan to actually put pen down on paper?

Well, hey, I'm trying to try to ramp up the frequency here a little bit, which is one of the reasons why this was part one. I wanted to kind of publicly commit myself, but you know, you can't wait six months for a part two. Uh, but really when I write, it tends to be because I'm seeing a trend and I'm seeing enough teams, enough people talking about similar themes. And then I'm I try to

pull it together and express it in a way where I I have the

For great fortune to be exposed to a lot of really talented builders and be talking to them all the time. And so I see some things and some patterns a little bit before I think

a lot of other people have the opportunity to. And so when I'm when I'm seeing these patterns emerging,

I then try to take it and condense it and articulate it in a way that I think will help it click for more people the way it started to click for me. And to be clear,

I don't know where things are necessarily going, right? I'm not claiming that I that I can see these things. I'm just seeing possible futures and trying to make the case for that future

as clearly as I can if I think it's a future that is is worth considering, worth discussing.

Well, we definitely want to get into what that trend is, but first can we just establish what the problem is? Uh the trend is here to fix things, uh, but does that imply that something is broken? What what about the current state of crypto? What about the current state of DeFi is broken?

Well, like you, I mean, I I really got into this space with first Bitcoin thinking of it in what was actually a DeFi context before the term DeFi even existed, but saying, hey, how do we actually

Make financial services available permissionlessly for the world. And I think we've made incredible progress with this experiment, but we've also had huge problems. And I think when you look at that headline number, you know, $3.8 billion being stolen in these hacks and these exploits just in 2022, right? Not even the full bull run, right? That was just last year. Um, and

that is just, I think, a couple orders of magnitude higher than an acceptable rate if we really want this to be the future of the kind of global financial system. And we can't expect

Be to be able to recommend to our friends, our relatives, entire countries and communities that they should be using DeFi when there is still this much money being stolen through vulnerabilities. And I think it is a

tractable problem. I do think that teams need to do a better job taking high quality audits and security practices seriously. But I also think that just doing more of the same.

Is not going to get us to that orders of magnitude improvement that we need to have in order to really be able to advocate normal, kind of non-tech forward people like ourselves be using this and being comfortable putting their life savings into these systems.

So Dan, are you are you positing though that um the big scalability, the the big thing we're missing in DeFi, the big reason it's broken is

hacks.

stolen money three pounds million dollars. Is that the the the real reason that DeFi is kind of broken and not taking off uh you know as quickly as we might hope?

I don't think it's the only reason. Certainly, I think there's a lot of other things we need to work on too. But I think that if we don't make meaningful improvement on the security front,

it's never going to be able to be mainstream without improving security. And if we manage to improve on UX, uh, improve on the way that we onboard people into the ecosystem, the way people uh manage recovery, the way they understand protocols and interact with them, if we don't

Uh, if we manage to fix all those other things without addressing security, we're actually doing a disservice to the world because that's just gonna mean it's easier for more people to lose more money.

Yeah, the worst user experience is like the one we got with SBF and FTX where it's like

Exactly. Exactly.

it's

fast, but oops, doesn't have any of my money.

Right. So it's not just about like smart contract vulnerabilities, right? It also is finding ways to better uh like self-police, hopefully within the within the industry around things like you know, scams and uh bad bad actors, both at the kind of company level, at the project level, at the like, you know, Twitter spam, telegram spam and scams, WhatsApp scams, all these things. Like we need to make progress. The number one thing holding back the industry is retail users in particular getting wrecked.

Right. Because then people don't trust it. They don't want to want to use these things. They they get a bad impression. They get uh it makes sense that regulators would want to come in and regulate when they're seeing retail users getting wrecked, because that is their responsibility, right? As they see it, is that they're supposed to help keep people safe. And if people are not safe,

then they're going to feel like they need to step in, even if they really don't understand this thing and don't want to be touching it in the first place.

So that $3.8 billion number that you cited, uh, you're not saying that is because of one specific part of DeFi. It is a global problem. There are many different types of hacks. Absolutely. Perhaps that's why there's part two coming, maybe more parts after that. But let's focus in on what the content is around part one. Uh, what what is broken about the current state of DeFi that is leaky in the uh in the Capitol? Like why why is DeFi uh vulnerable and and why does the current flavor of DeFi that we have not suitable to scale up to the masses?

Yeah. So I think that.

The major problem here that you know oracles fit in the headline, but it's not fundamentally about Oracles, right? It's about thinking with a layered approach to how protocols are being designed. Uh, one of the things that we've spent a lot of time focused on at nascent recently has been around security and really understanding the whole security process, the security stack,

and

working with both portfolio companies and looking at other code bases and understanding

how hard it really is to go in and understand the surface area of a contract, all the things that could potentially go wrong with it.

And so

our view is that the right way to do this is to start with a primitive.

And a primitive

Is a contract or set of contracts with zero external dependencies. And an Oracle is a type of external dependency. Governance and upgradability are massive external dependencies. And so if you have to reason about all the possible values that could be changed or different ways in which external data can be brought onto chain in both an accurate and timely manner, and how even if slight deviations in that accuracy or that timeliness can cause really bad consequences for the protocol, for users of the protocol, that the more we can strip that out at the lowest level, the better.

And there are absolutely very good reasons to be using oracles, to have governance, to have forms of upgradability. But the point is, we want to be very explicit