Backfeed

Transcript

hi I'll tell you about backside and proof of value so let's just start we have 15 minutes and very simple plan so I'll talk about the centralized decision-making or the centralized governance sit more specifically about the back feed protocol for that purpose and finally some example real example okay so let me try to tell you about what back fit is with respect to what the theorem is which you already quite familiar with so if a liam or if your blockchain is basically a machine to

machine concerns protocol by that I mean that the concede it concerns us that let many nodes distributed around the globe connect and agree an exchange value this consensus is basically triggered by algorithmic measurable quantities a break a grading the quantifiable data such as proof of work proof of stake etc in essence that makes this distributed computer which allows all those nodes in the network share a state together with respect to that backseat protocol is basically human to human concerns protocol so it's again it's consensus

protocol it's you know in a way from the same family but now this consensus is based and triggered by information that is only quantifiable or measured measurable by human human beings that makes a distant us governance or instead us decision-making and we'll look into it more in more details and that basically makes a shared decision in some some of different words worse theorem in order to emphasize the difference between these two kind of protocols that comes to they really address different questions so the theory of auction addressed the question of let have many many nodes

this with around the globe and let's ask how can they agree about this state like I I do an action I transfer some ethers to my friend how Oh everyone else will do we'll know about that the Blackfoot protocol is a addressing completely different question you can imagine minimum people are gathered in the same room they all see the same data the same information when I say something everybody hear me so there is no question about what the state is we all see the state but still these millions of people can ask the question how do we share or distribute value or allocate resources shared resources together in a

decentralized way without any hierarchy or at least rigid hierarchy or or centralized Authority the most powerful thing it happens when you put them together when you put them together just get what we call the centralized collaborative organization this is the true the central network that we have in mind so it's distributed in terms of geographically distributed it also has completely ceased real governance what this kind of machinery includes it includes the Sandra's reputation system a distributive value distribution collaborative human curation like decide

what is better what is worse and quality decision-making now let me emphasize one thing reputation systems are being have been in the space or you know in a broader space or for decades the point is that every single reputation system so far is crucially relying on the fact that it sits on a centralized topology any reputation system ranking system that you will take as for today and put it on this central network will miserably fail due to simple attacks and I will argue today that we have built the first Cybil resilient reputation system for the centrist networks which will be

crucial for any possible distant wise interaction between your movie it has many nice features it has the emerging coordination which means in the reclamation there is no need for agents to directly coordinate with each other if the coordination happens spontaneously just like and coordinate indirectly and make seemingly intelligent structures it has economic incentive for good behavior economic resilience has scalability if you remember from Greg's talk via compositionality and factorization it has a driving force into micro alignment which means every subset of the network is being driven into alignment and homogeneity well as there is macro

diversity and we get a diversity of themit at the network level or network of networks with this kind of machinery together with the blockchain we can now imagine many many gigantic networks whether it's the centers Insurance network of millions of people who ensure each other without any centralized Authority the central startups what it look like if millions of people will actually work on a project together with some centrist collaborative platform in Turners journalism of millions of free freelance journalists just post content edit others content etc it's a reserved ENB search engine and even multiplayer games and it's just the just random list

so I won't get into the deep details of the protocol is it it won't fit in that time but let me just give some some of the crux of it so how how it becomes super easy resilient the point is that and then identity in the sandra's network is nothing but the accumulation or the history of one's action you are what you do and if you don't do anything from the network perspective you are nothing so stable by definition is nothing and nothing means that he has no reputation and cannot affect Network it doesn't always be seated by making a valuable contribution when I say

valuable I mean by as perceived by already reputable agents and then valuable condition is also rewarded with tokens tokens and reputation the crucial point is that when one is making a variation of contribution of another he's actually putting his reputation on stake it's kind of proof of stake of reputation or another way to think about it as a prediction market for reputation of for tokens and then if the evaluation was good or bad we'll make him gain or lose his reputation now what good and bad is so good but is defined by a line with the community rather than this line of the community the consequence of that

is that it creates a feedback loop towards alignment those who were the majority of alignment before will define the excess of alignment then they will gain reputation from that step in the next step they will even have will even composed a larger majority and that's what they finalize the access of I might even strongly more strongly which means creates a feedback loop of alignment so what happens about the misaligned agents the misaligned edge agents has the incentive to fork the network just as you for code on github you can fork a network in new reputation system a new token system and by that you get micro alignment and macro diversity that

results in if you follow this process this is just exactly how evolution works and if you just follow the process of the process you get the fractal network of networks that's how the program dynamics look like in a very abstract way you can make a contribution from that configuration it can get rewarded with tokens and reputation once you have reputation you can make evaluations and then with this evaluation now you can evaluate a more contribution and that it goes on like that sliding with more details again not too technical but you can present the poor protocol in the diagrammatic

fashion which are physicists are very used to so these are this is a dictionary for Network collaborated contribution contributing evaluating a rotation it looks like that something like that so this is a network and then there is an agent who has all of the reputation of that network and then he's making some contribution and since he has all the rotational network he also he also has all the tokens of that network and now there is a new contributor he's making a new contribution now the first agent can evaluate that contribution and since he has all of the reputation as a result now the second agent received some tokens and reputation now he can evaluate back the first contribution and

maybe the first contributors token rotation will grow and finally there can be a third könig yeutter right now but maybe he's already the system so he is he's part of a new network and he owns all the reputation there but maybe the second agent evaluated his contribution but he's still not the majority but now also the the first agent evaluated it and now he's also getting reputation in the first Network and his process basically can go on and on and you can draw it you can make diagrams there is concrete calculation behind it and you can prove it it's fully a super resilient resilient let me quickly show you three proof of

concept that we've built so far and there are some more in the pipeline so basically this is a decentralized chess game so thousands of people can play together one hand against the machine so you can suggest the move yeah let's see what is better okay that move and then everyone can see that move and make evaluation of that move and your evaluation is of course weighted by reputation if you make a good move as judged by the community your passion will increase if you make a bad move your reputation will be decreased at the end of the day or at the end of

the turn the move with most reputation or most the most potent move for will be chosen and will be played let's see yeah yeah so that's especially how it happens so that's just a proof of concept let me show you another one oh so that's that's more specific like a the centrist application you can also check it out in chess make dot IO another another application we've made more more towards the center of gravity organization is designing a Chrome extension for slack

so that teams on slack can collaborate in the sanitized way so they can submit contribution they can evaluate each other contribute contribution they can gain reputation tokens see statistics and so on so I'll just skip the rest of that the last application I want to show you is the most important one this is a project together with etherium and the idea here is it basically that we want to make a proof of concept and more deeper proof concept of these protocols and where and for that we're looking for a wide participation of the community

this is called curate and the idea and curate is really to curate the content of the web so it's basically the centralized library you can add links you can add links to the library you can search tags by tags for example so let's just with Def Con let's see what we get we get some result about uber make it Dow etc you can you can vote on the tagging whether tagging are relevant or not you can vote about the quality of the links and for each time you're making a vote each time you're making a

tag basically putting your reputation on stake if the rest of the community agrees with you your reputation will be increased if the rest of community disagree with you your patient will be it will decrease and gradually we're getting a humanely humanely curated database for the web specifically here for this project central space but more generally anything you can just basically can just make a better search engine by that oh I see the guy that has made this oh this demo gave two stars for colony which means it gonna lose a lot of reputation become because colony are really awesome so that's our you add

links etc so any end by that and just saying Keep Calm nothing is under control and basically that the I would say gateway for real governance of the central Network thank you