it's there she is hi everybody charles hoskinson here live from warm sunny colorado or not so live i guess we're recording sam uh but it's always warm and sunny in colorado except for the times it's not uh so i very rarely do interviews and today i'm joined by sam from everpedia and uh there's some context behind this one so uh for the last five years cardano's been a project we've been around and for the last three years i'd say that we're pretty notable and noteworthy because at one time we as a project were worth more than tesla and spacex uh and at one time uh you know we had a market cap i think over 30 billion dollars but setting price aside we also published 90 papers about half of which have gone through peer review written more than a million lines of code and have a community of hundreds of thousands of members throughout the entire world we've been mentioned by the european union we've been mentioned by heads of state such as the prime minister of georgia many government bodies uh we have many university partnerships from university of athens to university of edinburgh tokyo institute of technology and most recently university of wyoming which is the first u.s institution to accept ada for a payment uh we've also been mentioned by the united states house of house of representatives and hundreds of journalists and probably thousands of podcasts well despite all these things it was of the opinion until just last month that cardano was not a notable project and therefore we don't deserve a wikipedia page to be joined by the other pages such as titcoin a deprecated project from 2015 or my video game legends of valor which came out in 1992 and i think i'm the last person alive currently playing the video game in fact the people who made the game are not even alive anymore but they deserve in some cases very extensive uh wikipedia pages but we obviously did not so our community for a long time went to twitter and other places and complained about it and it reached ahead uh last month and uh during that time sam you kind of reached out to us over uh for twitter and said hey you should come on over to the uh to the to the promised land and i said okay well we we should have a conversation about that so i figured uh i very rarely do interviews the last time i did it was for singularitynet with ben gortzel which is an artificial general intelligence project and so i figured i'd do an interview with you guys and try to figure out what your platform's all about and try to get a better understanding of um what problems you're trying to solve and how the platform has evolved since launch so um why don't we start with you introducing yourself and how the heck did you get into decentralized wikipedia as a as a product yeah thanks so much for having me on charles it's uh it's great to talk to you again uh we were on the same panel ways off uh in in hong kong and uh you know i missed the the pre-kovit days but hopefully things will get back to normal but uh yeah so i personally uh co-founder president of everpedia got into crypto around 2013 uh um originally script mining uh all the uh original coins and and all that stuff um we we started building everpedia in 2015-ish uh without crypto technology right because that was back when ethereum just like came out with like all of these uh you know you you know better better than most people uh with all of the vision but the tech wasn't exactly there in in 1718 uh era is when we decided to uh look into making everpedia a decentralized knowledge protocol um we picked eos uh to build on in that landscape um everpedia has obviously huge uh ambitions in terms of being a knowledge protocol one of the things actually uh i'll you know i'll say out in in the open is that um knowledge should be neutral right true true actual uh you know and then we'll we can get to the topic of neutrality but uh in in that space it should be uh available on any medium right and so we're always very interested in looking at other blockchain implementations of how everpedia works and so the idea with everpedia is basically what if you could actually rebuild the idea of a repository of knowledge in in basically every kind of uh decentralized way where the the actual content right the blob content the media text uh is it stored you know through ipfs or peer-to-peer channels the consensus of that is um currently it's uh incentivized uh with staking of the everpedia iq token yeah let's get to that but um let's start with what were the deficits of wikipedia what didn't you like about that platform because obviously if you're trying to do something you always ask well what is the incumbency doing that's good and what's the incumbency doing that's bad so what did you like about wikipedia what were the problems you were trying to solve back in 2015. exactly the the same uh experience that uh that uh you guys had except obviously um we just noticed that we weren't you know running a multi-billion dollar world-changing project it was the notability standard and the structure of it right because before we even came onto the blockchain right the idea was uh we had this this view that um what wikipedia did to britannica is uh wide open to be done to wikipedia because what wikipedia did is they digitized the encyclopedia with the definition of encyclopedic as the classical definition right so for example uh whatever for example that might or or needs to be on britannica or or something like that now is a community online effort and that's fantastic right that's paradigm shifting right for for the internet age 2001 um when when uh larry sanger and jimmy wales you know created uh wikipedia around 2000 and 2001 um and i mean it was new pedia before i and the history is actually quite interesting right so how it evolved um and the thing is i we didn't feel like the definition of encyclopedia was even you know even thought about uh with wikipedia and i think that definition has changed because of of the digital age and that was one of the main things uh we always had a problem with i'll tell you this that that uh i'm i'm a pretty big fan of wikipedia right like i think a lot of people think like they are a lot of what's right with the internet obviously there's a lot of things wrong that could be improved on but that they've a lot of people point to wikipedia as a lot of things that they get right that no other place does right so there's no tracking there's no uh direct advertisement and things like that um and and credit to them i'm a huge fan of all of that stuff and so we you know we need to recognize that and i i do recognize that and stuff like that but our original thesis with everpedia was why is it that the definition of encyclopedic is not is not uh rethought right it's not like digital space is an issue right and it's it's not like um you know it's it's not like putting something in in this you know compendium of knowledge or something would water down anything else if like uh and there's a dynamic nature to it so when you have an encyclopedia no matter how many people read it or interact with it it's the same but when you have digital content that is community curated the more fans you get the more knowledge you start accumulating exactly and this virtuous cycle right and then also the other thing is let's say something is is on there that that is not is not notable for whatever whoever gets to define notable right which is also obviously a huge problem right but let's just say the definition aside is fine uh why does it matter if something doesn't meet that criteria and and it's you know inside of of the the encyclopedia and like no one read you know few people get there like it literally is a problem that doesn't is not a problem right like like if if no one's interested they won't read it and it's not gonna you know but but if someone is uh then they will and actually that's what grew uh everpedia in 2015 16 and 17 before we moved it into a decentralized protocol that's what grew uh the entire uh platform to millions of page views a month what we actually um what we actually got page views for is we you could see you know on on like our google analytics and stuff people looking for the wikipedia article of something whether it was like uh an up-and-coming reporter influencer entrepreneur uh an up-and-coming like app or startup that you know wasn't as big as uber to have like a wikipedia page or something right um they were like looking you could see the the uh google search terms right you could see like so and so like uh wiki wikipedia like info like you could literally see so many people searching for these things and so everpedia came up right because uh we were a a well-designed um nice welcoming uh inclusive uh not you know deletionist ideal of knowledge and so we uh we had a lot of pageviews for for things that you know people really wanted to um actually read we uh branded it kind of like think of like you know why crunchbase exists right because people want to know what who invested in what startups and stuff uh if wikipedia was super liberal you would just go to the most likely via wikipedia page for for those startups but because it's so difficult and uh wikipedia is not really welcoming and embracing of of like the startup ecosystem everyone well crunchbase is a huge thing and i like crunchways too but but the fact is it would have just been absorbed by the compendium of knowledge that is wikipedia if they were more inclusive right and so you'd have a let's talk about definitions because you mentioned that the definition of encyclopedia needs to be evolved and grown from the britannica paper-based definition that wikipedia tried to adhere to so what would you say is a proper definition of an encyclopedia if you had to write it down in a paragraph or two what would you what would you say about it oh gee that's a that's a pretty hard one but my personal definition encompasses uh neutrality of topics and obviously you know wikipedia has their own definition but my personal definition is something that actually tells you all the positions cited right cited reputably about a particular topic and that actually encompasses the definition of neutrality right because one thing that's really really really important that admittedly wikipedia aside from these very clear violations most of the time and i know this is probably controversial these days with everything being partisan but in its history most of the time wikipedia has gotten right but but not really everything especially recently and so part of the thing is neutrality is really important it needs to really really uh encompass the definition of of an encyclopedia and what neutrality to me with what it means is that you what you're reading doesn't actually take a position it meta describes all of the you know spectrum of positions right and so that's what's really really important so when you for example read a uh encyclopedic article on prop 22 right which is the california proposition uh the uber lyft thing should they be employees or should they be contractors you should read it instead of you know feeling like you know who edited it like if they're on on a labor union or if they're like a upper or lyft like uh you know manager or you know uh executive you should actually be able to read it and hear both positions uh from that article and so that's what's really important that's my definition of of an encyclopedia you hear both sides and for example if you know dara jose the ceo uber for example was reading the uh the wiki for prop 22 like a really well written encyclopedic article he would probably read him like okay yeah like that that is part of that that uh uber's view is is really in there and he would uh you know contend that it's it's well written and then also on the opposite end if like the leader of like the labor unions for for like you know uh drivers and whatever was reading the same article they would be like okay okay like uh you know the no position really really well uh written or like maybe you know it's just so well inclusive of positions that's my definition of an encyclopedia i think it really has to uh really the central core of it is this idea of describing positions rather than just talking about a particular topic from a point of view okay so that that tends to work well when you're talking about like say a religion or a political event uh but what about when you're actually deriving knowledge from some sort of mechanistic version of truth like scientific knowledge uh so there's certainly for example a spectrum of opinions on global warming but uh you can take a view of well the opinions that matter or those that are scientifically derived or you can take an opinion well perhaps we need more nuance there so how do you handle those controversial things where there's a clash between a systematic way of thinking that derives at a truth that's concluded uh but versus an opinion for example prop 22 yeah right and and that's so so that obviously uh becomes pretty difficult once you once you have a definition then it's like okay you said include all uh reasonable positions okay so now what's a reasonable position and then how much air time in this article do these reasonable positions get what about scientists versus like tin foilers right on some of these uh uh positions that's a totally uh valid question it's pretty difficult to answer and i think i think that's one of the weak points of uh wikipedia right and i think that's where they're actually staying a little archaic because they've just embraced uh this classical idea of like anything that's a classical source of stuff uh if it's if it's like scientific it needs to be covered by like these these list of sources otherwise it's not going to be actually was one of our issues with wikipedia we were cited by springer vale which is one of the largest academic publishers in the world and we had peer-reviewed conferences that we were cited at and the editors actually said those were not credible citations yeah i mean that doesn't that that doesn't make coherent sense uh even even before like you know 20 years ago when they were found but even less so now because if you think about it the internet is so well integrated that like knowledge both synthesis uh site citing knowledge as well as uh you know creating and consolidating it all happen uh on the internet through different means right from public publishing scientific papers uh news or or like perspective opinions and i think wikipedia has just like absolved themselves of trying to even come up with a new uh you know relevant opinion and being like that's the same list that we have from 2000 uh these are these are it right and they probably don't even consider uh a lot of things that that are possible today that you know as part of their founding ethos and that is exactly what what people talk about when they're like the old get disrupted and and new uh new things come out and what one of the interesting really really funny things actually um i i want you to look at if you haven't seen it but there was uh there was this author i forget their name but i i have this i have this uh new york post that i think or new york times article bookmarked i'll send it to you but uh i don't know if you've seen it but this author had their wikipedia page incorrectly edited by whoever edits wikipedia right because it's it's uh anyone can edit it and they were the author like they proved like they he has a website he's a famous author right like an award-winning author and stuff and then he's like hey just wanted to tell you like um i it says my inspiration for writing this book which is super famous book is is not what's written on my wikipedia page it was it was uh the inspiration was this other thing and then the guys told well we can't change it if if the citation from uh the new york times someone else writing about you uh says that that like it was this we're not gonna change it he's like author it's lit it's my i am the author and i'm saying i was not inspired by this how can you change it and then and then what happened was they the new york post published his like post saying this is ridiculous i am the author and like i uh this is incorrect so that wikipedia could cite that thing and then like uh change it and say okay this was an error like you know his real uh according to him now that we can cite it on on the new york uh post or new york times or whatever and that's ridiculous right this is like the exact example like this is such a well like clear and ridiculous example of like wikipedia just being very archaic with i i ran into a somewhat similar issue so i have a wikipedia page about me and one of my fans is trying to get my commercial picture that's on the io hk website as the official picture there so they contacted me and said hey can you send an email to the wikimedia foundation to release the copyright so that it's under a creative commons license so they can post it i said sure so it gives me a template to send to them i email them and then email back well how do we know you actually own the ip and i said well i paid to have the picture taken well we have no guarantee that you're on the ip get the person who took the picture to uh to send us something i said oh [ __ ] this so i right there and then take a picture with my cell phone uh and i send it to him i said here you go this is my new creative commons and the email back saying this is not the original picture please send us the original picture and i was like guys i just took this on my cell phone i'll probably send screenshots and tweet them out but i i can understand you're you're just kafka-esque uh adherence to orthodoxy that these guys have but let's follow that thread so neutrality is part of your definition sorry i had the exact same i don't know if this is like a crypto founder experience or or what but like i also have a wikipedia page and then i had the same exact issue except i think i i think i kind of outsmarted my way uh it was a really long time ago but um but the they said the exact same thing they're like hey uh you know this page is you know this thing is uh this picture is clearly uh taken really well or whatever who's the person uh that took it and i'm like here's the um i have the copyright for it right and then the exact same answer how do you know that uh that that it's it's you that took them like okay i took it with a with a tripod myself are you happy and then they never replied and kept the picture yeah yeah it's crazy it's crazy it's absolutely crazy but you know it's it is it is what it is when you're worried about liability so let's talk about neutrality you know this is key to your definition of a modern encyclopedia so part of that definition if i understand it correctly this idea of completeness so you would like that all viewpoints have been fairly represented but then there's this issue of fairness of representation so for example let's look at abortion um the people who who are pro-life would claim hey this is murder and we'd like that to be in the article the people who are pro-choice would probably have a difference of opinion so how do you square that circle of who gets to decide first if it's complete or not for the sake of neutrality and then second who gets to decide if the representations made by each side are fair and accurate you know what what does that mean in a system like this yeah it's it's really difficult because uh then you you kind of come to the idea of okay sure you can show both viewpoints in the article or something but you can never escape true fairness because then like what do you call it in the first sentence like is it you call it abortion is like the murder of like you know an an unborn you know child or fetus or again right like what are you gonna say right so um it's at some point uh just like a lot of uh human you know philosophical concepts it becomes just an intuitive grapple um but what's what's really uh interesting and stuff that we're thinking about at everpedia is like is it better to design the system where these categories of pages and views are disjoint in there's there's multiple encyclopedias but using the same protocol this is actually something uh larry uh singer suggested and something that he's working on in his view of knowledge called like the encyclosphere or or like a greater uh protocol a greater wiki so to speak right where it's like instead of trying to pack everything into one unified article you create a protocol right like an interchangeable protocol so you know uh pro-life people have their own uh and it's kind of what's kind of happened with like conservapedia wikipedia and things like that but without a clear uh overarching protocol for this right that actually uh defines some some communicative uh rules but with everpedia we so far have had a a really welcoming community overall that's allowed um both uh you know both sides and and without without too many issues uh and and so well one of the cool things is people say like if you're pissing off both sides you're doing something right because you're getting both things right but at everpedia actually what's really cool is both sides like us which is extremely difficult like we do get a lot of uh retweets fans and and and people on on the conservative spectrum saying this sounds fair in in this in this environment like i appreciate that and then obviously uh progressive and liberal sides to uh a lot of our content they're like uh you know this is this is reputable the sources make sense and stuff like so as hard as that does sound as difficult in today's environment we actually get that quite often which i'm pretty proud of right well larry is actually coming to do a seminar for us at every week we do a technical seminar uh for our engineers and uh on the ninth he's going to be presenting the encyclosphere idea and we'll have a great chance to kind of dig into that and the bullet points and checks and balances there and so forth um so that's fun and he sent me a book of course he's i guess a trained philosopher so it's going to be fun to uh to see what he has to say i mean i i also studied philosophy i was i don't have a phd but at ucla i was a neuroscience and philosophy double major but uh so me and mary have gone so much back and forth he's shaped a lot of my beliefs on neutrality so yeah and it's clear that we don't have it because people are getting the platform but it's not innately clear what to do about it you know and there's always this tendency to build a meta system above the existing system to solve it we even do this in mathematics uh for example we had a lot of problems with set theory we started seeing paradoxes occur like russell's paradox for example we said oh we'll just create some sort of structure above sets like categories or something and that'll that'll just solve everything you know so so sometimes you build a bigger system to resolve the problems of the smaller system and then of course inductively the bigger system will still have some form of problem with it that you have to then build a even larger system anyway setting all that aside so we have this baseline definition this concept of neutrality but one thing that hasn't been discussed is knowledge categorization there seems to be an implicit difference from knowledge that's discoverable from knowledge that's not and what i mean by that is let's say the world has amnesia and every copy of the bible is destroyed there is no reliable way that we could just somehow through inspiration reproduce exodus or genesis or these things whereas if we just have that amnesia and forget everything we knew about gravity through a series of experiments and interactions with the physical world over time we would rediscover many of the truths that we'd learn about gravity we've seen this with a lot of knowledge throughout human history and we have the ancient world and then the dark ages happened and then the renaissance occurs and we kind of rediscover a lot of the things that the romans knew about of the greeks knew about uh for example the minoans even had um geothermal heating for their and indoor plumbing in 1500 bc took it till like the 16th 17th century before these things in a.d uh became prevalent again so how do you categorize knowledge well you know first in general in what you've chosen to do with everpedia do you have kind of a catechism of how you separate things into different buckets and say well this is this type of knowledge and this is this type of knowledge and based upon that category you have different ways of curating it or incentivizing the curation of it yeah that's a that's a really deep question especially because it kind of grapples at the idea of uh what we can even know right and what we can discover through empirical science and what we uh need to preserve and then you get into the question of okay well who gets to you know preserve it and say what it is um but one of the things that at everpedia is and this is pretty controversial i'll be the first to say this is that since this is a crypto protocol there's this there's an iq token you know cross chain implementations could even have different kinds of tokens you can you can create minted tokens on other stuff you can bring uh market dynamics to an incent game theory incentivization to categorizing and upkeeping different types of of knowledge and again that is a very controversial idea right so so bringing monetary value and value into any of this stuff immediately gives an intuitive like knee-jerk reflex of oh my god there's money in here now that's uh it's like a you know like a market of like misinformation and stuff like that um we actually kind of this is everypedia is a big brand experiment right like like a lot of these things like bitcoin and stuff like that um and so we we try to use game theory and and like uh seeing how different types of pages and content from like the scientific ones to the historical ones to actually even to even like recent or like like trending or or news related things that become prominent uh right how they're up kept and how much you know iq is staked on them for example or or if there's uh to where influx of new editors are going you know just there's there's a lot of types of knowledge both philosophically categorized like you're saying empirical uh historic or in in the internet age uh breaking right just like when people are trying to discover you know for the first time empirically like what is true or not like you know all of these uh events like the protests in the in the us and stuff like that um like what's actually happening right like like is is uh is everything burning or is is is there uh unrest here or is it peaceful and all that stuff right um so we actually are experimenting with market dynamics as you know and some people will will raise an eyebrow there but we that that's what uh part of everpedia's grand experiment is and we're seeing different kinds of uh stuff and that and so far it's been pretty good it's been pretty soon yeah i mean we pay scientists to do research and you know we pay authors to write books and these things so it's not necessarily the world's first thing to create a marketplace but when you create a market create winners and losers so now that you've been running the experiment for a little bit what have you noticed your marketplace values and there's a lot of activity and incentives and what have you noticed is not so valued and actually is losing in the marketplace yeah that's a great question uh actually we've been surprised that things that uh you know that you would think um you know are should be potentially not as valued so much because it's not as clickbaity like basic science or historical things are actually valued a lot of our uh most edited things are uh upkeeping of our uh encyclopedic and scientific content so like we actually started when we started everpedia right we uh we forked wikipedia right because creative commons all of ever pds content is also creative commons it's all free knowledge open uh decentralized and so a lot of the things that people really like to do is actually upkeep uh the original uh wikipedia content but in a better you know more diverse light and and things like that and you would think intuitively like okay well maybe if there's iq tokens involved and things like that the more interesting stuff is uh is like breaking news like people want to get page views or eyes on on a lot of this stuff and that's not actually necessarily true i think um i think there's an intellectual uh curiosity that once there is a market you can earn iq tokens or cryptocurrency or something of value uh for genuinely pursuing your your intellectual interest as long as other people in the system you know vote that this is this is actually adding value you you can get rewarded and it's actually very very cool like one of the things we realized is um our science pages and stuff uh are edited as much as um you know the influencers and up and coming people and cryptocurrencies and and stuff like that now obviously like you know just in terms of numbers uh the more actual page views coming from search engines and stuff at this current time is you know influencers and and new cryptocurrencies and new entrepreneurs because most people currently still go to wikipedia to learn about uh you know quantum mechanics rather than i've repeated but in the coming months and a few years and stuff i think a lot of that will tip a little bit right okay so there's a lot of questions we go for here you know there's definitely market related questions and then there's questions involving the transitivity of knowledge curation so let's start with the mark because that that's probably more familiar to most of our listeners um you have this token and it's called an iq token yeah and is this a participation token in that to edit an article you must have it or is it uh like a different model like steam where you you can create something but if it gets well curated you earn it i mean how do they how it's like a modified updated um obviously in our uh opinion improved version of steam and so when you uh for example create an article you have to stake some iq and then people that have iq staked can can vote on whether to change like revert the state of of your article which is literally just the meaning of that is reverting your edit right or approving i can create an article i have to be vested in the system um a minimum amount so actually that's technically yes in terms of programmatic but actually what we do uh so you can right now go to everpedia you can sign in with a twitter facebook email and it's what we wanted to do actually is in order to start we wanted to get rid of that huge barrier to okay i have to go somehow get some iq tokens so i have to first go to coinbase because i don't have cryptocurrency right and then buy bitcoin and then like move it to some exchange for iq or something like that what we actually do is we give you the minimum amount right some constant c right in the protocol of whatever right and so you stake that and the idea is um if you do a good job right if you do something that adds value by definition of the people who have stake and are you know also doing stuff then you get a small amount of the inflation of that uh reward period and then you can actually just it uh the front end programmatically continues to stake the new iq you've earned to be able to do do something else again right to edit again and stuff like that and so actually what it becomes is if you're doing a good job again as defined by people who have stake right again the definitions are pretty important um it's totally free in fact not only is it free but it earns you value right but so as defined by the value you're creating but if if you're attacking the system as defined by like you're taking up resources you're just putting stuff that other people think that uh you know is is uh is like a waste of the resources of the system then you're not earning the value you're just basically you have to continue to get iq to uh make these edits which then can be overturned but then you know you can keep you know proposing them or changing that okay so is the relationship for everpedia to the everpedia protocol similar to the relationship between steemit and to steam and that there's kind of like um the website on off ramp yeah there's some similarities uh obviously steam you know the idea is anyone can post uh anything um and in ad repeated you can as well you can make articles but but the idea is that they should be cited um they should be uh the community should have one you know round of of like vetting the the edit state and and stuff like that so yeah it's similar to steam and that that's touches back to what i was uh talking about before is like one of the things you know we've thought about is um again then that just kind of creates the same definitions of okay well if the iq token holders end up becoming as calcified in ideology as wikipedia right then you kind of eventually end up with the same uh systemic issues except now it's uh now there's money involved right and then uh um one of the things we're thinking of is well is there some way we can create different everpedia universes but all governed by the same protocol uh kind of like what i was saying before where uh there is like conservapedia in wikipedia but they don't even talk to each other and there's no overarching interoperability between them there's no unit of account or anything but one of the things we're thinking of is what if you could for example uh color your iq tokens like conservative iq uh liberal but you know other stuff right maybe like math math iq right and so you can uh like a stack exchange type of approach exactly right but then it's still the same protocol exactly that's a better uh way and so so for example like a decentralized stack exchange kind of encyclopedia right and so that's one thing we're looking at right now so there's this plutocratic curation that exists inside the system so i guess the hypothesis is that the desire for the token to survive and thrive will override the personal preferences and biases of the whales in the system uh for steering the protocol in a particular direction is that a fair way of stating it that that's a fair way of saying yeah as long as those things are are balanced right like but if one thing and so the idea is like you know as the importance of everypd as a brand and where people go uh goes up if the token price doesn't proportionately go up then you have then then someone might want to sacrifice some value because the of the token to influence the the knowledge there right because it's a good deal right but as long as they're both growing the idea is you're right what you're saying is hopefully going to hold true it's kind of like the point where people talk about uh well if the bitcoin block uh you know rewards go to near zero unless the transaction fees go up right um like proportionally right then the hash rate either has to go down and then that lowers the security right the security has to kind of uh match the bitcoin dollar price in some way right and so yeah that assumption is uh similar in in everpedia's uh protocol which is not a given right it's not a given for the bitcoin security structure of the the fees and blocker awards and it might be we've actually seen deficits on many cases but they're usually short-term and self-correcting so there's a kpi question that of well how do you know the system is actually doing what it's intended to doing you know what metrics do you guys follow to indicate that you're converging to neutrality and plutocratic curation is actually resulting in a lack of censorship as opposed to what we see on wikipedia which is commercial censorship yeah so so that's a two-part answer right one is um the kpis one of the things is that obviously we have our own opinions but uh we really uh so far have had very positive um you know uh responses and and anything that uh people like like i said both sides are actually happy but both sides and like the spectrum is is quite happy so that from just an intuitive perspective is quite good the more the more interesting part that i think is important is that's where the blockchain and decentralization aspect comes in is that i think that if we build the system correctly it should be impossible uh for us you know as a group or as a website right to co-opt the protocol itself similar to how uh bitcoin is doing really well you can't you know coinbase can't cut off access to bitcoin right because it's just it's one one front-end interface uh ford one wallet one service provider and so the idea is the way we've built everpedia right where the content is is stored um on ipfs uh the hashes are updated you know the stakes and voting is all all the back end logic so to speak is that on chain the most recent uh hash history for each article is stored inside you know smart contracts and so you can actually see all of the history of all of the state changes um and in that respect i think we've designed the system uh similar to bitcoin right so if if i you know get possessed to become evil and want to spread uh misinformation or something like that uh which hopefully is never going to happen but if if that happens you don't even have to trust me right anyone still you're talking about knowledge preservation and it's my mine is more of a question of of quality because we've taken the position with this definition that not all knowledge is created equally there's a difference between highly biased junk stuff and highly sourced neutral stuff that's universally valuable to the you know almost utilitarian sense to as many people as possible so the question is what kpis would indicate where on the spectrum you fall and month month do you guys see progress towards that right side of of neutral balanced information like how do you guys track certain things for example just an analogy we created auroras for cardano and it's a consensus protocol and it's it's one thing for me to say as always super decentralized and secure and so forth it's another thing entirely to say okay we have metrics of participation so we have the number of registered stake pool operators and we can look at the trend are more unique people participating or less or the same and these rewards enough to actually create sustainable businesses or businesses going out of business all the time and it's you know declining so we track that as a kpi of whether the protocol is working or not do you have you know analogously equivalent uh kpis for are you converting to more neutral articles and both volume and quality yeah so so one of the most clear ones uh that we keep track of is as you know we originally forked wikipedia so we have all five million english wikipedia articles and then one of our kpis we actually have uh over one and a half give or take uh unique uh everypedia articles that have uh that are only on everpedia right now and and a subset of those um and i can get you know the community if they're interested in stuff and card out the exact number but subset of those uh that were deleted from wikipedia uh cardonos was was one of them thankfully you clearly and well deserved sorted that out but like for example a subset of that was literally uh stuff that wikipedia purged for for lack of a of a better term and then a lot of the other ones was uh new articles and knowledge on everpeda that wasn't even made on wikipedia right just because of the platform uh community and and uh um guidelines and structure of how that we've actually been able to foster right and so those are like hard numbers um now how to analyze the content in those it becomes extremely difficult right because then you don't have a wikipedia comparison at least right then then you're literally uh just just it's just everpedia right and so it becomes really interesting um for our wikipedia forks a lot of them are uh up kept slightly different in my opinion better more visual kept by a diverse group of editors also upkept by by wikipedia there's a lot of wikipedians that actually cross edit and like i said we like a lot of these talented and very smart uh editors so like i think um i think it's really cool that there's a lot of crossovers so we actually have some good kpis over a million uh and a half-ish of uh of you know just knowledge on everpedia and some of those unfortunately removed from wikipedia which i obviously really uh disapprove of personally i mean that's pretty impressive it took wikipedia almost 20 years to get to a scale of 4 million articles and you guys uh you guys have achieved about 20 of that if not more in just a few years so that's a great growth curve uh but that's a qual quantitative metric what about qualitative do you have a way of when you actually go on a case-by-case article by article basis like for example one of the articles we have umbrage with on wikipedia and we're trying to get changed is the proof of stake article it's it actually hurts dan larimer as much as it hurts me because the only representation of proof of stake is like nxt and these things back from 2012 and 2013. i mean grand and eos and cardano and polkadot and all these other things and there's like probably 50 proof-of-stake scientific papers that have been written and any time we try to do anything to improve the article uh it's it's removed and they say iohk spam and other people suffer the same fate so i would argue that while that article exists it's part of that set of existing existing articles the quality of the article is quite low so do you have a way of if i was to arbitrarily pick at random an article out of the set of the everpedia articles looking at that and then being able to say yes this meets the direction we're going and this this neutrality and high quality information uh and you know we can measure it maybe even give it a scoring between zero and a hundred it's like a hundred is perfect zero is like junk you know can you actually are you at a stage where you have those capabilities yeah that's a great question so first of all totally agree about i don't know what it is about the these like crypto articles being literally like archaic from the crypto middle ages right like now there's avalanche cardano uh you these are like set like i mean goons here is like a like a cornell professor right and like it just none of these like breaking technologies are even you know proportionally discussed which is absolutely crazy so totally agree also frustrating um so that's a great question and something we've we were working on directly when when larry was uh at everpedia and then larry's actually working more directly full time on that exact idea of rating knowledge one of the cool things is we can also have uh value marketplace kpis built into it but as of right now the only things we track are how much quicker our stuff gets updated than wikipedia or sometimes slower most of the times quicker um and and the length as well as the diversity of citations but no we don't we don't have like a like a rating system um as of now in fact i would love to uh either collaborate with uh you know really scientific community like cardano's or or anyone that wants to you know join in our like kind of mission of knowledge and stuff like that to develop something like that like i said we're working on it with larry uh when he was at everpedia but um there's no rating system as of now but there's a wealth of uh ways to model this this kind of stuff because it's all on chain it's all open uh open source um now there's market there's value you can also model as well right like how much stake was put on or that and it's it's really exciting but but no we don't but i would love to uh you know and then the problem is that there's a blurred epistemology line between uh of subjectivity and objectivity here of you know the value of this knowledge that's very contextual to user um so the minute you start rating it you run into the same issues that like politifact and facebook for example when they say hey this is not true versus true exact first thing we ran into actually well because because the the point is like what do we have a unit of account of value in the system which wikipedia doesn't right like like but but even then how is like more stake proof of something uh and what is it proof of right is there more proof of like uh people fighting over what you know what can be written on the you know donald trump wiki or or something like that or is it just you know what i mean and that that was the first thing we ran into but at least now there is a stake and for example you can also include uh page views as a you know like which front ends are getting uh more uh you know viewing and so one of the things i forgot to talk about actually that makes everpedia as a protocol really interesting is for example you can create any front end uh and pull a subset of all all knowledge and then like for example only show uh you know you can make elephant pedia right only show elephants in my content about uh elephants not everything right not politics or science yeah actually let's talk about the metadata component of your data so how much uh tagging and discussion around what these articles are it happens to exist a lot of that's hidden from the users and wikipedia but what does every pedia stance like if i wanted to spend an afternoon and learn everything i could ever learn about i don't know some chinese herb pick your favorite one and so i enter that search term and obviously it's going to find that based on the search ability of the data so i do i tag it as the author can the community curate and tag it is it discoverable even if it's not explicitly tagged there's an implicit way through ai or another means to actually discover this on the knowledge graph so how do you guys handle the metadata problem of knowledge discovery and curation that's a huge issue with these systems yeah so we do have very basic rudimentary tagging um anyone can can add a tag um but again one of the things with with tagging and discovery is we actually uh want to make the the base protocol as simple and uh unopinionated kind of like the idea of bitcoin right as as possible and actually leave the the tagging uh as well as the displaying as well as the discovery algorithms uh to as many you know front ends and stuff like that as possible so like on everpedia.org which is the main uh front end there's a few uh more there's some some front ends in uh china actually um that pull the content in their own way but um uh we we just have the state of the articles as well as their you know associated tags that come inside the whole ipfs uh blog file um but what we would like to see and what we're kind of seeing and doing ourselves is uh every front end or every group can uh tag their uh these articles partially internally to the protocol and then a lot of it externally and then like publish these links um so that there's kind of a front-end uh tags that people can integrate for their products or something if they want per hash or state of an article but also they come within protocol uh discovery tags but they're they're very basic right now so like for example it's like cryptocurrency or or like uh medicine i don't i don't think there's chinese medicine or or uh all of these things yet but yeah we do think about discovery and knowledge uh a lot as well but we want to keep it unopinionated right like again we on the protocol level we just want to have arbitrary uh tagging by editors okay and so and so those tags the metadata is also stored or a hash of that is stored on blockchain yeah yeah per actually uh per state of the article so like since what state was it tagged like you know with cryptocurrency and then was it removed was another tag added at this current state uh what are what's the what's all the metadata associated with it exactly okay all right so what about localization so obviously the world speaks english pretty well but there are certainly many places billions of people that don't and you know we live in a polygon society so what provisions are made for other languages one of our frustrations was with the cardano article there was a german article in a dutch article and a japanese article and those were all notable but then apparently in english we're not notable so you know i'm japan notable but not english notable so how is this resolved in everpedia do you suffer similar problems or do you have a better way of resolving um localization also equivalent quality across these articles because in some cases when you read the japanese article it's completely different from the english article even though it's the identical subject right right and that's it's a whole can of worms so i think i actually i think i became uh swedish notable first because everpedia's hq is in stockholm so i i think i i personally became a swedish wikipedia notable first but um so there's two things one is um everpedia is actually used in a bunch of places where it's difficult to access uh english or uh the local wikipedia um or or it's unreliable i don't know you know to make any kind of uh you know like political uh opinion whether it's like turkey china or anything like that i know we're used there um and and so uh people can can access it through um front ends there that are in internal right so like certain countries uh stop access to things like wikipedia or outside sites uh in terms of language that's a very difficult question one is uh we're used in a lot of places right there's actually uh german everpedia itself is a pretty um old school i think it was our first big uh community but um right now our huge biggest communities are korean uh everpedia chinese a lot of it is because of the um the prominent you know position in crypto the project serves um in terms of language what we would like to see and i'll be the first to say and honest that we haven't really made too much progress is that we'd like to have one object right uh that's that's about the particular topic and and uh kind of like all of the cross-language knowledge kind of synthesized into that right and so wikipedia doesn't do that right like you said the the like japanese article on cardano or whatever else is totally different than the english one and they're all disjointed on wikidata which is this uh schematic data structure uh you know back end of wikipedia um those two articles are connected this is the same thing right it'll link to the different languages but um it's not well done at all and so i i think that it can be much better done we haven't personally made too much uh you know uh progress there but we have really vibrant communities everywhere and then so i think with our more tech focused method we'll be able to make a bigger impact there but as as of now i'll say i'm of the opinion one object one topic uh one you know place where humanity puts its uh sum total and then you know you could it can break up into different colored versions but um but i'm of the one object one topic model um that's what we're striving for yeah okay well i wish we had another hour sam but unfortunately all good things must come to an end so if a person wants to join or become part of their pedia project where should they go and what's the best way of getting started yeah i mean we're super active on twitter telegram if you want to join our communities but if you want to start editing it's just as easy as everpedia.org um and you can sign up with anything like i said twitter email facebook or if you have a crypto wallet um it is currently built uh only on eos um like i said we are very very open to uh collaborating um in fact one of the things i was looking at the other day was uh ethereum l2s uh which would be super interesting right um but yeah so i would love to talk and get to know a lot of the cardano community um and it'd be fantastic to start the conversation together so thank you so much for having me on and hopefully see some of you guys on the everpedia telegram and on the side thank you so much sam this was a lot of fun have a wonderful day thanks you too
