dionysuz.eth - Making Ethereum Visible

Transcript

Okay. Hello everyone. I'm Dianisis or David. Dianis is just what I go by online. And today I'm going to be talking about probably something a bit different than most staking conversations. Um, and I hope you find it exciting and useful. It's about making Ethereum visible. Um, has anyone here heard of the Delian problem by any chance? Okay, so the Delian problem was back in ancient Greece in the town of like Delos. Uh, there was like a plague going on and they went to, as they do in these ancient times, they went to the oracle of Delos and they were like, "Yo, how do we like uh not die from this plague?"

And the oracle was like, "Uh, you need to like double the size of the altar of Apollo." And they were like, "Okay, um, how the heck do we do that?" Uh, back in those times, they had straight edges and compasses. So, basically, they could draw like lines and they could make circles and angles and stuff. This was like uklidian geometry. And it turns out that it's really difficult to double the size of a perfect of of a cube. And they just couldn't do it with a straight edge and a ruler. So the intuition that I want to give you is that like the tools that we use make our reality visible and in a kind of weird way like math isn't even real in some sense. It's just a measurement of reality that help

us um achieve certain things. So there's this quote I like by uh this Spanish architect Antony Gaudi. There are no straight lines or sharp corners in nature. Um so yeah um the delion problem is basically that we're able to if we have new tools um such as the um cartisian plane you can actually map uh uh you can actually see cube roots so you can't see cube roots with just a straight edge and a ruler. So what I'm going to be talking about is like cognitive tools. Uh this is another one just to kind of drive the uh point home.

Um, at least in quantum field theory, there uh is a notion of like there being these like underlying quantum fields that like uh basically influence how particles like work in in in our in our universe. And fineman diagrams actually describe how some of these particles work. They describe something that's actually invisible. And I'll get to like how Ethereum is invisible in a sense too and how we measure Ethereum, but they describe something that's invisible and they make it visible through this diagram. And if you don't use a diagram like this, you end up having to see this whole thing as like what's called a path integral. And it's basically this really complicated math. And it's like, okay, what does it even mean? So these fine man diagrams just make it easier to

understand behavior about an invisible world. So TLDDR cognitive tools help us understand and change our world. If we want things in our world like we want to double the size of a cube because we don't want to die from a plague or we want to do something in Ethereum, we have to look at the tools that we have, right? like our rulers, our measurements, our cartisian planes, math, things like that. So, we build and measure Ethereum with cognitive tools. On the engineering side, our cognitive tools are pretty good. We use a lot of math. We use a lot of mechanism design. But on the social side, they kind of suck. And I'll talk about uh more of the social layer of Ethereum and how homestakers really embody what I'll refer to as like the spirit of Ethereum.

Um, so one more example of a tool is a miscalibrated compass. So this is really kind of a tragic condition that I want to highlight. Imagine you want to go somewhere, right? Like in the same way that the people of Delos wanted to double the size of an altar because they literally just didn't want to die and they believe what this oracle said. So let's say you're a traveler and you want to reach the illuminated city and you have a compass. So you say, "Okay, you know that the illuminated city is north. So you follow the compass north, but it turns out that your compass is actually miscalibrated and what you thought is north is actually northeast. Well, in that case, it's tragic because you are trying to approach something noble and your whole epistemic reality of what um of of what you want to do and how you

measure your world tries to take you to this illuminated path, but because your compass is miscalibrated, you end up going the wrong way. And this can happen in in the way that we measure and learn about Ethereum too. So cognitive tools make the invisible world visible, measurable, and importantly actionable. Like with these cognitive tools, we're actually able to change and engineer our world. Um, so one point I want to drive home is that Ethereum manufactures antifragile civil civilizational substrate. Um, so basically there's like two sides of what give Ethereum this hardness that we're able to build civilizational infrastructure. One of them is on the protocol side where we do like a lot of mechanism design and we

figure out okay what is like the right way to build incentives so that there is like we have these hard properties like censorship resistance uh you know security uh and and um and things like that there's another side of hardness that I want to talk to again which is like on this social side and to a large degree is controlled by stakers who ultimately control the fork choice. Um, so here's this new uh thing that I guess the ETH Panda Ops team uh put out that people were talking about this like new lab uh visualization where you can see all of the uh decentral uh all the nodes decentralized uh uh uh um around around the globe, right? And this is awesome.

And one thing you'll notice which is interesting that people in like cryptography or like mechanism designers always talk about is they want to disintermediate control so that control goes down to the user, right? They want to say, "Oh, we want um control to be with the user in a biometric sense, in the sense that, oh, I I I I I scan my face and that uses a private key to give my intent and you verify a signature and that's me." Or I give a fingerprint or I burn some calories and I touch a screen and say, "Okay, sign this transaction and do this thing." Right? So we like to look at systems and we say okay when when we when we try to distribute keys distribute signin keys throughout the network we want to give those keys to actual users and humans that are bounded

by like this cognitive uh human layer. But what's interesting that I'll say about this diagram is that like even though you have distribution of the signing key of the of the participants in the validator set which are you know signing out to stations proposing blocks and things like that the there actually is a cognition layer that is arguably not decentralized where for example let's just say like there is some huge campaign that goes around on Twitter that's like oh we need to 10x the gas limit like today like for some reason everyone thinks that we just have to do that and they're like really strong arguments Right? Like even though your signing key is like uh distributed and theoretically everyone has their own reasoning and it's and it's relatively pluralistic and decentralized, it's easy

to convince a large amount of the network to say okay let's actually do this change and in a sense it has some centralization. So, one thing that I want to um um hit home here is that the protocol kind of purchases human reasoning as a way to manufacture anti-fragility in terms of this like substrate we build for uh civilizational infrastructure. Uh so yeah, let's just get right into it and look at some cognitive tools in Ethereum. The first ones are like social feeds. So we used like Xarcaster uh things like that in order to measure the world in the same way we use like rule uh the Greeks used rulers to like measure sides of cubes and things right we're like okay this is the state of the world this is how things are this is how

we learn about the world and um we primarily use that through social feeds and the these have different types of algorithms that that that show you um you know about the world and with different incentives. So um the cognitive tools we have uh kind of suck but there is promise of things getting better. Uh again we have like ETH magicians e ETH research this is like uh synchronous human deliberation uh with text. Uh we have Farcaster again it's a you know text based feed for you to understand the world and you can also propagate sentiment throughout the network. Uh we use math uh podcast and then there's this really cool tool that the uh um uh e uh protocol support team has which is called forecast. Uh forecast is this new tool which is like

basically takes the all core devs process and it has um uh some new tooling built on top of it new cognitive tools actually. So one cognitive tool is like you take the abstraction of like an all core devs call right and it might be a complicated thing because you don't understand all the details of the protocol and all the mechanisms involved with it. But if you have like a TLDDR summary from an LLM especially an LLM that's tuned to your preferences in the way that you understand the world, you have a new cognitive tool for understanding how Ethereum works. Another cognitive tool is like over here on the right. This is basically a hierarchical uh like like segregation of um of of priority between EIPs, right? So, it

uses color coding for you to understand which things are relatively more important and like things like SA, B, C, and D to like kind of weight things between each other. I'll give another example. There's a project uh called deep funding uh in the uh Ethereum ecosystem where basically we're trying to like figure out what is the relative value of a bunch of different repos inside of the ecosystem. Uh so for example um we we kind of found out that like it's hard for a human being to figure out like how valuable a single thing is in isolation but it's a lot easier for a human to take two things compare them against each other and and measure them with a multiplier. So for example, it might be hard to say okay

like you know you want to produce you want all of these percentages to add up to 100% or the number one right in terms of like capital allocation on like what's a good idea and you can easily think of these repos as like EIPs in some format or things that you u you know you want to promote within the ecosystem and think are valuable. It's hard in isolation to say, "Oh, yeah, like the ETH Panda Ops like the Helm charts is like worth like 3% or something, right? It's like like how do you come up with that, right?" But it's like it's a lot easier for someone to say, "Oh, comparing these two like which one is more valuable and by how much?"

So, this is a cognitive tool, right? It allows us to measure the world better, produce better data, and allows us to actually engineer better solutions for a world that we want to go towards. Um this is another uh uh thing called polis. It's a policy selection mechanism that is similar to the rough consensus algorithm we use in all core devs. The basic idea is that you have um a bunch of statements. Um people agree on them, disagree on them or pass. You take data about all of these people and you plot them in like multi-dimensional space and you cluster them actually. Right? So people have dimensions and features and then people are all of these rows in a database. And

if you can plot them in multi-dimensional space, you can kind of have clusters and you have groups. So you could say for example that there's like I don't know six or seven kind of clustered groups in Ethereum based off of certain properties, right? And what rough consensus does and also uh this polish policy selection mechanism does is they say we're going to reject any policy that doesn't meet a certain threshold of each group within the ecosystem. Right? And it turns out that policies that use this like bridge consensus algorithm overall have better like agreeability in the group than ones that use something like majority voting.

Like majority voting usually ends up looking something like this where just a huge group of of actors in your ecosystem don't actually agree with the policy and that causes all kinds of social issues. Uh open agora is one other uh cognitive tool that I worked on with some uh people in the open source space. Uh it's a tool that like kind of like utilizes polus and tries to think about how we can um you know we were actually thinking about this uh when uh there was all this like eof debacle going on like a very sensitive issue because people worked on it for a long time and we wanted to see like okay how can we like measure how people feel about this and like whether or not it's a good or bad idea for Ethereum. So, you know, this is one way that you can like

actually try to have a view into making something that is, you know, kind of invisible in some sense like, okay, how do people feel about EF or how do people feel about Fossil and and making that visible to to to a certain actor. So, an actor always has a lens that they look at Ethereum in the world and the policies and and and and they try to measure it and figure out like what's the reality so you can act on it. So this basically uses like polus you can get an idea of like okay how much consensus are there for different issues and I didn't show the part where there's like a um a voting mechanism like agree disagree pass but it's like the same thing um and there's a little bit of a loaded slide uh but uh because I just wanted to get information in there but

the basic idea is that you know when we measure our our our world like Ethereum when we try to understand Ethereum we all come from different backgrounds and we all have different like technical understandings, different philosophies and we can't just use one tool for like everyone because that's just not everyone thinks the same. Some people like to read, some people like to watch videos, some people understand diagrams. I mean everyone understands the world differently. So what's important is that you give the person who's trying to like measure the world their own tools for understanding like how things are working, right? So to give you an example like one thing that you could do is just an example I thought of recently is like there's recently this like

trustless manifesto thing that was going around on Twitter right that Vitalik and uh some other people uh made and uh you could say okay I I agree with this right you could say I agree with the trustless manifesto like this is something that I'm I'm aligned with right so you could for example take an LLM that knows your preferences knows that you align with the trustless manifesto and interprets some set of uh things to measure the and it can tell you, okay, these are the impacts of this proposal based off of the values that you have. Another way to think about this is like I'm not sure if people have like are familiar with like the issuance uh debate like uh I think under like Ellison like wrote some stuff about like issuance and like for me like a lot of the uh um the the formulas are

like I just don't like understand the math and all that very well and I'm like I just don't really like get like how changing different like parameters and the functions like actually change things that are important to me. But like the thing is is like you don't actually depending on the lens of of what actor you are in the ecosystem, you don't need to know the details of that math. You just need to know how this impacts you. Like for example, it like if you are a staker, you could say, okay, you're just going to earn less money, right? I mean, there's going to be other re there's more nuance to that, but but you can see the outcomes of something without knowing all of the details, right? And LLM are actually really good at that. like you can say okay here's this like like this hard

engineering thing that I don't understand like what happens like what is the result of changing this like complicated thing that I can like actually understand um and um yeah so the TLDDR here is that you know Ethereum cognitive tools like uh kind of suck in my opinion but they're getting better with things like forecast uh experiments like polus and experiments like deep funding Um I think that this talk in some sense is a little bit philosophical but that is because Ethereum is a little bit philosophical like it's just new technology right and we have to accept that we're doing something new that hasn't really been

done before so we have to we we have to invent these new things actually like it's it's really a new precedent so I think that we need to try to like invent algebra in some sense right in the same way that algebra and like daycart coming up with like the cartisian plane allow would have allowed the ancient Greek uh you know people to see cube roots right like we we need different tools to be able to see uh the reality better and the reason for that is because it'll allow us to measure our the the world better and and engineer better solutions faster. So, um I I think that uh one thing maybe I didn't hit on enough is that like you know really the most aligned actor in this ecosystem in my

opinion is the home node operator the home staker um because they have the most alignment with um the the the hardness of the protocol because they they earn ETH and I think it's just really important anyone that stakes that it's really like a um it's it's it's really like a role to be proud of I think and and it's really impactful in the ecosystem. So, I hope that we can get more stakers on Ethereum and more people that stake involved in the process of of of of controlling the network, right? Because you run a certain binary that that has the protocol rules. And um you know, I think that that that it's very important to keep Ethereum as this like antifragile uh civilizational infrastructure. So, if you want to hack on this, I'll be at ETH

global over the weekend working on cognitive tools and we can always chitchat. I'm Dianis's X on all offchain places in Dianicis everywhere onchain. And then I also, as a shameless plug, I uh make an open source tool called Kitty Node where it's like easy to like run an Ethereum node with it. So yeah, uh I don't know if there's time for questions, but that's my TED talk. We do have about time for maybe one quick question.

I see one here from the audience, but looks like I'm I'm not seeing one. Will you be around a little bit? That >> I will be around. Yeah. >> Excellent. >> Okay. Thank you guys. >> Thank you.