 Mae'n digon i'r cyhoeddus o'r cyfrannu. Mae ddyn nhw yn ei wled y ddiddordeb ac diolch i arfer真的很io. Mae'r collad o'r collad yn rhan o'r colladau. Mae gweithio'r collad o'r holl. Mae'r collad o'r collad yn mynd i gael llwyddiach, mae'n cael cael'r holl. O'r holl wedi cael'r holl yn siŵl'r f mistyr, a'r holl ar哂f ar hyn a gwladll gwyddoedd. Um, dyna'n gwybod fel yw'r gwaith o'r llunio am y rhan i'r cefnod trwy'r cyfrwydau dda'r wrth i ddweud a'r ddweud i'r cyfrwydau. Yn ymryd i'r cyfrwydau, mae'n gwybod fel ydym ni'n ddweud i'r cyfrwydau. Yn ymryd i gwybod fel ydym ni'n gwybod fel y byddai'r llefflygu a'r llefflygu o'r cyfrwydau cyfrwydau i'r cefnod, ac yn dda'r gwybod fel ydych chi'n ddweud i'r llunio. There's this idea running right through Western civilization right through from Plato up to at least the beginning of the 16th century that there's objective truth, it's out there, it's consistent, it's eternal, it's absolute. And this originally, it was thought to apply not just to factual stuff like, has my office burned down. a'r rhaglenion morol yn cael ei ffordd. Ydyn nhw'n ffordd, o'r rhaglenion Christian, rydyn ni'n ddweud y brifesiannol. Ac rydyn ni'n gwahau hefyd. Yn amlwg pwysau ar y gwrth, a'r pwysau yn fawr. Efallai nhw'n ymddangos, ac rydyn ni gallu'n ffyni'r ffordd o'r hyn o'r ffordd. A gynnyddio'r 16 ysgolion, Martin Luther made his epic ship post nailed to a door on a church in Germany, allegedly this is disputed. And he says you're doing it wrong. And after that we had hundreds of years of religious wars in Europe, while people fought over their conceptions of the absolute objective truth. One of the strategies that allowed people to stop killing each other and live together was to separate truth into different baskets. So you have things like scientific truth which is considered kind of objective and that you shouldn't really disagree about. And then you had other value judgments that were not considered objective. And even this bit of objectivity is something that some recent philosophers have got a little bit suspicious of. So postmodernists think something more like this, truth is agreement in language. So the idea is that when we make claims, we're playing a language game, we've got a kind of a consensus system. And something is true to the extent that it follows the rules of that language game, right? So it's true if you obey the consensus rules. Now, while this has been going on, there's something kind of weird happening in politics. The internet is having this weird kind of crisis about truth. So we've got on the one hand this vast internet of knowledge, but at the same time you've got people quite boldly embracing alternative facts. And you've got basically entirely made up news circulating freely on Facebook and deciding presidential elections. And this is a hard problem for two reasons. The first is to do with efficiency and the second is to do with perspective. So I'm going to take those two in turn and I want to show you the DAP that we've created to help with the efficiency problem. And then I want to talk about an approach called subjectivocracy which can help with the perspective problem. So there are people out there on Facebook who are trying to hack your brain. Anyone in your social network or any paid advertiser can push claims into your stream and you have to work out what to believe. But when you're faced with a claim, if you were prepared to spend a lot of time and money evaluating it and researching it, then you could probably do quite a good job at getting to the bottom of it. But since you're not prepared to do that, you end up potentially believing a lot of bullshit. And this is a kind of a pattern that you see a lot in information security and also in, for example, missile defence. There's this asymmetry between the attacker's job and the defender's job because the attacker can just hit us with this entire barrage of bullshit and we've got to defend every time to be successful. And even if the attacker can't manage to get their bullshit through every time, they can bring off a kind of a denial of service attack against the truth because if you can hit people with enough stories and enough claims, they just lose confidence that they can evaluate any of them and then when they see something true, they can't evaluate that either so you end up being able to steal from them without them noticing. So what we've tried to do is to increase the cost of attacking and decrease the cost of defence. So this is reality check, this is mechanical Turk meets Snopes on the blockchain. The rule is that anybody can ask a question and anybody can answer a question and that anybody can be you or it can be your contract. So we make no distinction there. And the goal is that people who give the right answer are going to get rewarded and make profits and people who give the wrong answer are going to make losses. So we can ask a question like this, this is via the DAP or like this using a contract. So we're passing in a template ID there that I won't get into. There's also the text of the question, a challenge time out and an arbitrator. I'll say a bit about the arbitrator later. So the question is going to show up on the blockchain and therefore on this website. And you can come in and if you see a question you know the answer to then you can post an answer like this or from a contract like this. Now when you post an answer you're also sending some money. This is a bond so this is money that you're going to lose if you're wrong and you're going to get back if you're right. And by wrong and right I mean whether your answer is decided by the system as the final answer. Now every time you post an answer, if you see an answer that's wrong and you want to correct that answer, you can always do that but you have to double the bond. So once that's happened and once the time out period is elapsed you can get the final answer from your contract like this. So if we know the question ID we just pull it in and that will give us our answer. Sometimes you want to do something more like this. Sometimes you didn't ask the question yourself but you want to take any answer that adequately answers a particular question also with a reasonable arbitrator and with a reasonable time out. So in that case you can get the final answer if it matches these particular criteria. So your user can send you a question ID and then you check that question ID with the contract and if it's right you go ahead and use the answer. So you're probably wondering what happens if somebody keeps on putting money on something that's untrue and is prepared to put serious money to back their lie. And what happens is the first thing that happens is it's going to escalate. So you end up with quite a lot of money in bonds resting on this particular question. And then once that happens we get a serious amount of money in the pool and somebody is going to want to use that money to pay our arbitrator. Now the arbitrator sounds like a person and it can be a person but it's a contract so it can also be a decision making process. It doesn't have to be cheap, it doesn't have to be fast but it does have to be able to get us our answer and we do have to trust it. So it's kind of like the pope in the medieval metaphysics there. Now for a lot of purposes this is going to be the right solution and we're going to be able to construct an arbitrator that suits your security needs. In particular it's often going to be better than what people have been doing at the moment where they're pulling data from some website and even if the middleware that gets them that data is correct that website is their pope and often the websites that you're pulling from even if they're backed by some fairly reliable provider aren't actually secured with the kind of assumptions that we make when we're dealing with magical internet money. So we've got ourselves to a place where we don't have to call on this arbitrator very often and we can use an expensive process. So the arbitrator could literally hire somebody to go down and look at my office and see if it's burnt down. Out of the gate this is a service that we'll be offering at Reality Keys but we also think it's a great fit for things like industry consortiums. If you've done any work in the private blockchain space it seems like everyone wants to make a consortium to do blockchain stuff and then they come and talk to guys like us and they say what should I do with my consortium? This is really somewhere where these existing industry players could leverage their existing brands and their expertise. You can also make this a socially expensive process. For example if you've had an ICO and you've got some people who are supposed to be delivering particular milestones you might say we want to pay them if they reach these milestones and we're going to check that with Reality Check but the arbitrator is going to be a vote of the token holders. So in most cases you're not going to have to call on them but if you do then you can. So Reality Check doesn't care it just lets you supply your arbitrator and it sets things up so that the arbitrator only has to get called on very occasionally and that when they do get called on they can charge very attractive fees. So this is our application deployed on the Rincoby testnet. I've also posted our Gitter channel there. The plan is that we're going to take feedback for a couple of weeks on the details of the API and stuff in case anyone here has any thoughts then get out to one audit and hopefully have it live on the main net very soon. So that's the efficiency problem but I want to talk about another problem which is perspective. So for four years we've been serving data on the question what is the price of Bitcoin? And that's been a fairly factual question for most of that time. There was a point when we had some funny data coming from Mt Gox which eventually went away but generally speaking it was fairly straightforward. But then next month when the fork happens this becomes a fairly contentious question and it's really not clear at all what the right answer is going to be to the question what's the price of Bitcoin. There are going to be at least two or three bitcoins around that you could possibly take the price of and maybe you should have them together and who knows. So that's a question that really depends on your point of view. And perspective is quite a tricky problem for consensus systems. If you look at the original problem that Bitcoin was trying to solve you've got all these nodes on the network and they all have a different perspective on the network so they all see transactions at different times. And what you want to know is given two transactions that have been sent which one was sent earlier and that's something about which every node will legitimately have a different opinion. So Satoshi just kind of punted on this and Satoshi's genius was to realise that it was actually okay to punt on this question. So Bitcoin or Ethereum will just pick one and what it'll promise you is that if you ask it again then with an increasing probability it'll give you the same answer again. So if we're trying to get the truth of a proposition it's not really clear that this answer will do it for us. We don't really just want to punt on anything that has perspective involved especially because things like the price of Bitcoin a lot of the things actually turn out to have perspective when it maybe seemed like they didn't. So we want the answer but there is no the answer. Now this is something where we've managed to handle in blockchain systems when we're dealing not with the transactions on top of the blockchain but with the basic rules of governance of the blockchain. So what we did with the Dow hack or what we did with Bitcoin and Bitcoin Cash was when there was a serious dispute over governance we forked the chain. So if you have a Bitcoin before that fork then you have Bitcoin and Bitcoin Cash after the fork and if you have Ether before the fork you have Ether and ETC after the fork. And this gives you kind of a slightly strange but kind of remarkable kind of security in that especially from the point of view of the minority if you lose to the majority if the decision making process in force goes against you you still have the ability to continue with your own chain you still have the ability to follow your own rules and it may be that even if in the short term the market goes against you who knows maybe in the long term you will also find that your token is more valuable than the other one and even if it doesn't then you still get to do what you want to do. This is a great fit for handling disagreements about reality. What we can do is embrace perspective and embrace subjectivity. So what we need is opinionated money. So we've written a token contract it's about 100 lines of solidity and it commits every day to any number of new world views. So every day you can create new branches on it and cause it to fork and each of those branches will represent a different world view. So in each of those branches then the contract knows how much money you should have and it'll allow you to spend money that's on parallel chains twice but not money that's on the same chain. So it looks something like this so if you wanted to know did my office burn down one person might say it did and one other person might say it didn't so we have two forks and you can keep your money on either or both. Now, once we can do this your smart contract no longer needs to choose between these different allegedly objective realities. So if you've got a bet on did Trump or Hillary win an election you don't need the smart contract to choose which of those things happened. You don't even need to be an objective truth about which one of those things happened. Your contract just pays the person who thinks that Hillary won or who bet that Hillary won on the branch that says Hillary won and pays the person who bet that Trump won on any branch that says Trump won. So we can take all of that stuff all the objective stuff right out of the contract and put it into the social realm. Now at this point people always worry that if we've got these hundreds of thousands of facts and we're branching on all of them we're gonna have just an unfathomable number of branches. But this is quite a nice fit for our arbitration contract because our arbitration contract in reality check represents a process for for getting resolving a bunch of facts. So you have a large number of facts that are all hid behind that contract. So what we can do here is you commit to a white list of arbitrators once you think that an arbitrator is bad you can just remove that arbitrator from the white list. Or anyone can just publish a branch that removes that arbitrator from the white list. And this means that we no longer have to trust our arbitrators because if the arbitrator is bad the economy can just root around them. So we can do this whole thing without needing the trust. So very quickly I just want to compare a couple of code examples. So this is the claim function with a pope. This is the claim function with an arbitrator who the contract already thinks it's going to trust. So what we're doing there is we're checking the question that somebody feeds us a question idea and we check the question and then we send them in this case the balance of all the tokens held by the contract to that person. This is the same contract on subject democracy. So the claim function is a little bit different. We've got a branch parameter and an arbitrator parameter and the first thing we do is we go to the token contract and we ask it on this branch is this arbitrator legit? And if that arbitrator is legit then we go ahead and all the rest is just the same except for the very last line where that token payment also specifies a branch. So once we have opinionated money like this people are welcome to play their own language games and embrace their own worldviews. So you're entitled to your own opinion you're also entitled to your own facts. This is Ethereum, it's permissionless, it's free you can do what you like. But if you want to play a language game with somebody else you're going to have to persuade those other people that the world that you're trying to inhabit is interesting. So we'll see what kind of values the markets actually want to assign to those different branches and we'll also see what people actually want to use. So Kellyanne Conway is welcome to her alternative facts if she wants them that's absolutely fine but we'll see what universe, what worldview what set of facts she uses if she wants to say ensure her house and get reliable data when it burns down. So if like me you're excited about this then please come and talk to us on our Gita channel. I've also put a thing you can Google hardfork all the things this was a piece I wrote last year about this also links to VitaLeaks writing on subject democracy where he did a lot of legwork. So please do come and talk to us. The reality token thing like I said is nearly ready to go but this thing we're still kind of batting around different models so we'd love to hear what you think. Okay thanks.