#DappCon22 : Day 2 - Sybil Resistence In Web3, Kevin Olsen

Transcript

now we have Kevin Olsen he is engineering lead at Bitcoin Dao he will talk about civil resistance in what three let's welcome him on stage cool thank you all right cool hi everybody um so I'm not Kevin owaki um I'm Kevin Olson so uh I kind of feel like this is the sort of the uh can we get Kevin to speak at our conference meme mom we've got Kevin at home that's Kevin at home Kevin couldn't make it so I'm gonna do my best to step in um but yeah big shoes to fill so maybe to get started who knows about Bitcoin

grants how many of you used it yeah who's made money with Bitcoin grants anybody here gotten funded yes all right cool yay um what I'm going to talk about today is civil resistance so there's kind of a problem right in the space that we've solved through you know some elegant methods um but we want to kind of characterize this like what is this civil uh resistance issue why it's a problem so we'll kind of go through that um talk about how it's uh you can kind of think of it as this game right and we'll give you the characteristics of the game and then we'll propose a civil resistance money Lego

so to get started oh my slides aren't moving that's okay um if you want to play along at home you can go to wtfizzqf.com that's a site that we run you can play with qf if this was moving what you would see is a match amount at the top right there um that's like the matching pool in Bitcoin right and quadratic funding as you add funders to that you'll see that the votes are distributed not evenly but quadratically meaning many small votes actually brings a much larger match than one large vote so apologies this is a PDF and I'm not getting the animations there um you can kind of get the sense that what

happens is really big votes really large single votes uh the kind of impact is attenuated right so this is a close kind of uh sibling to quadratic voting but in quadratic funding what we're doing is we're taking in these votes are the voice credits in the system are the individual donors donations right your vote to support a particular project is matched out of the matching pool but it's summed with all of the others in the uh in the in the voting system in get going we've raised and distributed 65 million dollars and this is a network uh graph so every dot here is a donor and every line is a transaction um what's super cool about this is the

original ones were pretty sparse and now you can see this is a very dense network of donations problem so again if these were moving uh what you'd see is this donor at the top with their 10 uh one dollar donations is dragging much more funding to their project right now if I want to drag more to my project what stops me from generating 20 accounts for myself right 20 song puppets and I can steer way more funding back towards a project which may be I control and this is the problem that we see we see people putting in Grants that they control setting up these sock puppet accounts and distorting the output of our of our system

um but beyond just Bitcoin there's a really cool opportunity here which is we can start moving into more one human one vote type systems right if we had civil resistance so here's the kind of prism right so with civil resistance you can get into quadratic funding right we need that for our system you can get a quadratic voting you could maybe do things like Ubi right um and data collectives and some really interesting stuff that allows us to move away from what are essentially in the system plutocratic methods of decision making right so actually honoring true wisdom of the crowd decision making so okay I think we get a sense of like why gitcoin needs this and why other

systems need it so why is this a civil resistance game we talk about this a Bitcoin in terms of a red team and a blue team right this is kind of like I think security parlance the red team are the bad guys and the blue team are I guess us in this case or the parts of the Dow that try to uh uh not have these effects distorting the outputs of our quadratic funding rounds so who are they so there's just script kitties right there's just people who are kind of in it for the lulls who are trying to like impress their friends uh you have Petty criminals you have people who are just trying to see if they can exploit the system then you get into like rational economic actors and this is a really interesting

one right because these people aren't really doing anything wrong there's funds that they can get there's airdrops they can Farm they're doing their best for themselves but they found a way to exploit the system right they can create a bunch of sock puppets they can try and distort this and drag funds to their projects it was a lot of D5 depth if you followed this one this guy stood up like a whole defy ecosystem whatever not not our problem but maybe somebody else's uh you've got organized crime you know people with more resources potentially bigger uh goals and then you'd have like nation states right you have people potentially unlimited resources right so imagine you're building some very large uh systems you could you could very easily lose to these folks

so now how is this a civil resistance game so what are the characteristics of this game right and why why propose it is this is because we want to kind of propose a solution to it as well but it's important to think about you you can't just go after this with a single idea right so one of the things that you've seen maybe some people claiming they've solved civil resistance is you get into some pretty dystopic uh Solutions so let's let's try to avoid that by characterizing the game so the most important thing is criteria zero it's an adversarial game right so if there are funds to be captured someone's going to try right someone is going to try and exploit that they could be a rational actor or they can be as I said you know actual

actual bad guys actual criminals so okay it's adversarial we can't get around that step one here is the need to be privacy preserving right you can't well you could right you can see these biometric approaches you can build a centralized thing here but ultimately what you want to do is make sure that people are have their own data right we love giving our data away for free right don't we well we all love just giving our data away to Google and Facebook um I I think we're here because we don't

want to do that that's why I'm here so I think for a design solution you need to make sure that's privacy preserving and that people are have data sovereignty right that they control their data and they control where it goes um step two right is avoiding plutocracy right so this is the token voting that we're all pretty familiar with um it kind of works in this adversely environment where people with the most tokens probably have the most investment in the project maybe you can kind of draw that out but it's a kind of economic gatekeeping right and you're really trying to Source the wisdom of the crowd where many people who don't have a lot of funding can steer things like Bitcoin grants you need to have something other than this right so you can't just gate on tokens I think is a long-term solution to the to

the game collusion this one's a harder problem right so what stops me from bribing you to use your identity in a voting system um there's some technologies you have Macy right where you can make it hard to prove that the person who was bribed did the thing that the briber wanted them to do But ultimately I think some of the the better uh proposals here kind of come from the the dsoc paper right that vitalik and Glenn and fuja wrote that one's interesting because it kind of sidesteps the province is all coordinated actors are colluding and let's let's just sort of take that as the the premise but in general collusion is a very difficult thing to to look at

from just an individualistic uh approach so step four you need to be able to build in systemic defensibility so I think this is a good one where if you think about the way we secure chains uh it's economically incentivized or or maybe the risk reward is better to use your mining Hardware to mine than to attempt a 51 attack right so that you can build a system that incentivizes systemic defensibility and if you want to drill another so this is the get coin grants round one completely wide open anybody can show up end users you have the red mixed in with the the good actors

quadratic funding grants and 25k and matching but where are we now we actually have a number of kind of places so this is where we're sort of defining the Battleground right we've got Grant reviews now right so grants go through a review process um we've got reviewers right that's kind of the blue team there then you've got the ability for those grants to get through the approval process and you have passport which is a product I'll get into a little bit and so other d-apps in that ecosystem other identity providers that help us to kind of sort out who's a human and then we use pairwise qf which is another way for us to attenuate um forms of uh collusion and to add things like the trust bonus which we

added a few rounds back and ultimately that is the gate onto the grant so we've kind of got our points of control here and other systems would need similar ways to sort of yeah intervene so step five is building Network effects what I like about this one um actually I think if you look at this if we've all read impact networks I think by erlichmann um I think we're kind of somewhere in the stage when the stage two phrase right we have lots of projects that are attempting to solve uh civil humanness proof of personhood type problems but they're all kind of disconnected they haven't really been integrated so we're in the phase now we're trying to kind of pull that together into a hub and start providing some way of

aggregating these signals and to build some kind of community around that but then we also want to make sure this is a multi-hub approach and then ultimately you get into this kind of core and periphery where you get the real sort of effects of that Network all contributing uh without a single point of control or single point of failure I think kind of piggybacking on that same thing you have modularity and forkability right so anybody should be able to iterate on this permissionlessly from the edges um kind of goes into the open source ethos video eyes makes any problem shallow um I think that that's an important design criteria if you're not interested in that then I think this may be a bad Community for that too um and then obviously decentralization right so getting into this idea where

there's not one project that controls this or one community that controls that and lastly that I think the important thing here is that this is an evolutionary game right as the red team gets more sophisticated we need to make sure we have tools and that we're thinking about evolving our system as well right that we need to make sure we're we're working to to solve this right um so where we are now like the feedback cycle for Bitcoin is kind of quarterly right we run our Bitcoin rounds we look at what happened where the signals distorted or did we find uh fraud at the end you can do like an end analysis but then we need to also be doing this daily uh groups with the

FDD inside the git coin Dao which does a real-time kind of monitoring of this and then again instant feedback so you have to have these signals kind of nested within your uh your system so then lastly here's the proposal all right um what will grants around look like in around six thousand Milwaukee joke uh nice um is it needs to be anti-fragile right to many types of attackers you have your script kitties all the way up to nation states you have to be ready to be resilient against all of those things uh and you do privacy preserving right we can't be um harvesting this data we can't be concentrating it I don't really want to run a system that's a Honeypot to docks

people right so we need to make sure that we're designing this openly we'd be avoiding plutocracy right if we're designing systems they're supposed to be Democratic and surfacing the wisdom of the crowd um we can't just fall back on token gating we need to avoid collusion like I mentioned there's a open design space there and some really interesting things to do um this system has to be antipragile because it's cheaper to defend an attack that goes back to that systemic defensibility right that we the cost of attacking the system is too high to be economically feasible the network of suffix of Open Source enabling The Innovation preventing capture and then lastly that we need to make sure that we can't just say this is

a one and done thing right there has to be a place for this to continue to evolve so what have we built we built passport um the docs are there if you want to play with this or if you're using it and get coin grants around 15 which is running right now you can go to passport.getcoin.com what this is it's an aggregator right so you can sign in through a variety of systems you can attest to your activity on chain and we use that as a way to then you can feed that passport back into git coin and get your trust bonus score so this is part of one of those blocks there where we're attenuating the signals of Bad actors or people who can't prove their personhood uh by um looking at that uniqueness of all of

these signals what we've chosen to do here is to make sure there's no pii in any of these stamps so we're using the did and VC approach um check it out it's all open source there's an SDK feel free to play with it um and that's it thank you [Applause]