Mechanism Design Explained: How Blockchains Program Human Incentives

[00:00:01][Music] thank you [Music] okay uh hi everyone I'm Sam one of the founders of awive and today I'd like to talk to you about mechanism design which is was an obscure piece of literature kind of study in economics until around 10 years ago with the birth of Bitcoin and that's obviously extremely important to us as people building crypto systems but it's also not something that we were all um trained in by default I would say so this is essentially what you might think of as programming humans for engineers and this is based on um experience and learnings that we've made kind of at our weave along the way and also before our weave so for the last eight or nine years without I guess formal study so I was part of a PhD program studying and building decentralized operating systems before I got into the the crypto scene and as a consequence it was kind of necessary to learn all this stuff ourselves and I would say that most of the literature out there about mechanism design is focused very much on people inside the economic sphere by default and so really what I'm trying to do with this talk is give you a bridge from people that are very uh from the place that us Engineers are very used to sort of inhabiting to somewhere that is much more kind of Economics based and I can basically tell you along the way how things relate and also where you can go to read more and get way into the depth of how crypto economics and mechanism design work afterwards but really I would say this is the kind of missing link for engineers getting started in the crypto space on the economic side so this talk is in three parts there's an introduction which takes you through broadly what mechanism design is how you can use it and a few examples both in centralized and decentralized systems and then there's an overview of how it can go wrong and the potential pitfalls for this I mean we're talking about essentially programming human behavior and so there are a lot of moral and ethical questions that arise from this that I think we need to have a strong framework through which to uh yeah to discuss and act appropriately and then finally I'll talk to you sort of more in depth about some example systems so we'll go through our weave and also Bitcoin and see how we use these mechanism design approaches in practice to build systems that achieve a goal so the fundamentals of mechanism design are programming human behavior through the careful design of incentives in this talk I would like to posit the theory that essentially humans are just like machines that we would normally program but instead of responding to and running programs in a traditional sense so applications that are series of instructions that the computer must execute in order to be in line with you know properly functioning CPU instead what we have is people that led by goals towards well through incentive systems so this is really quite simple in a nutshell we see that humans are machines that want to achieve goals and they plan their way towards those goals through sub goals so one of these goals might be I would like to go home after this and have a stake and as a consequence as soon as I desire that my mind starts to think okay well what are the sub goals involved in reaching that position in life so we see that on the way to these uh metagalls we go through sub goals like with steak for example I need to go to a shop I need to buy the steak I also need to have a house in which to cook it right so we see that there are these sub goals and I would argue that one of the most fundamental of those is money in fact money is the unifying sub goal for almost all larger goals in life and before money we had essentially this mechanism where there was in order to have a transaction there must be a double coincidence of value so what that means in practice is that if I want to have a stake I need to go and I'm a carpenter and say I need to go and I need to find someone that wants to buy a table from me that I've made and swap it for some amount of steak this is obviously a very very inefficient system so one of the first mechanism designs that humans made was like okay well what is a global generic sub goal and that we would say is money that is everybody wants money and as a consequence we can freely exchange through money for other Goods of value so I'm a carpenter I can sell a chair to someone I get money I can take the money and I can take it to a butcher and I can buy some food and as a consequence everybody is happy and this makes economies much more efficient so from our point of view we can see okay if we can design a mechanism that gives people money as an output then we can insert ourselves on the Chain towards reaching basically any of those goals that those humans want so it doesn't matter to us whether they want a vegan meat-free burger or a steak or anything else or a chair a house doesn't matter because the sub goal in common is money so if we can define a way of giving them money as a result of playing an economic game then we can essentially push their behavior down the um the route that we're attempting to Define so this is what we do we design a mechanism that's output is money uh leading to eventually the goal of whatever that is that that individual actually desires and humans are so attached to this process of seagull generate sub goals execute towards those sub goals towards the major goal that this is an extremely reliable way of pushing behavior in a in a desirous fashion so by designing these economic games appropriately we can put selfish humans which are essentially humans that are playing what we would call the dominant strategy so the strategy that is most effective uh inside the game you've created to exhibit useful behaviors so this is the entire game this is programming humans yet a human you create a mechanism this output is typically money but it can be social reputation we'll get to that a bit later but in most cases it'll be some form of money which allows them to get towards a goal and there's a byproduct of your mechanism you have something that you find useful so example of this in practice in my native Berlin we have a bottle pickup mechanism design run by the government where basically if you pick up bottles from the street you can deliver them to drop off locations and you'll get around 30 cents per bottle so this is a way of giving the uh the local population that don't have any other job a job by default but as well as that it has this useful social byproduct of creating cleaner streets and also requires the local government to pay fewer street cleaners and centrally plan less so one of the the um I guess common factors we'll notice when we're looking at mechanism designs in this lecture is that a lot of them really have this amazing attribute that they push to the collective brain of society the question of how to play the game that you've created most effectively but we'll get to that later so it doesn't have to be government run here is an example of the juicer economy so when these new scooter startups came around they they basically had this problem where okay so at the beginning of the day we distribute the scooters across the city we put them in places where they're useful to people and we make sure they're fully charged but by the middle of the day all those scooters who are almost all of those scooters are in some way that's much less useful and they have half charge and by the end of the day it's in complete disarray so the early startups went down the route and saying okay so what we need is a central planning system whereby people um can yeah we essentially hire enough people with Vans to go and pick up the scooters to charge the scooters and then to redistribute them to useful places um in the city but this is extremely extremely costly and very ineffective it's actually a an another sort of example of the traveling salesman problem right you have to work out how do I go around all of these scooters in the most efficient way to charge them and then also redistribute them so the smart school system companies decided okay well let's just not do this instead why don't we say for every scooter that you charge and you move to a point that we just determine we will give you some reward and that was it and then they found a way of marketing that opportunity to people and as a consequence people just came along and they would pick up the scooters they would move them to the right place and they would charge them and there was no Central planning involved and um and it's extremely efficient so this really works in practice and if there's only one thing you take away from this talk I would say is that you can take that same mechanism and that you can apply the same reward rules that we do in Bitcoin mining and in cryptocurrency mining of almost all kinds and you can make a mechanism that will give you a fair market value for any task like this so let's just imagine that I'm Bird right I don't know how much I should be paying people up front for this job of Distributing the scooters around the city like it's actually quite difficult to determine what an appropriate reward for that would be so what I do instead is I say okay well I'll start off and I'll say I will distribute two thousand dollars per day for people to do this task for me and then people join and as more people join they have this sort of uh competition effect which lowers the cost of um each of these movements and charges of the scooters until it reaches the opportunity cost at which point that no new people will come to the mechanism and say yes this is a profitable thing for me to do so that's fine that limits your growth but from this point as a startup you've really essentially found the optimum pricing strategy for having the service delivered for you it's extremely effective and it's possible to do in almost every situation but doesn't just have to be economic Rewards so here's a BitTorrent client BitTorrent worked with a mechanism design called optimistic Tit for Tat it's extremely simple and extremely powerful what happens is I we play a game all together that says I will give you data if you will give me data and then occasionally we will give data to each other at random what this means is the Nash equilibrium so that's if you imagine if only selfish players are playing the game at rest when they are when there is no more movement in the structure of the game then what will happen is everybody shares data all the time with everyone else and this is not just Theory again this is practice so this mechanism was responsible for the transfer at one point of a third to half of the internet's traffic it's extremely effective and this was long before Bitcoin what's really interesting about this is that actually it requires zero Global synchronization so that is I just need to remember who I like and that's it and then you can earn reputation with me and essentially gain a social ranking just by interacting with me in a nice way so giving me data when I give you data that's it there's no need for on-chain transactions or anything like this it makes it extraordinarily scalable but of course this leads us to the final example here which is Bitcoin so Bitcoin is decentralized ownerless money that is a system of money or a system of value storage at least where no one party is able to control it um or to yeah uh change its ownership by force or print more of the tokens in the network we have proof of work as a mechanism to reward people for um securing the network and also processing transactions again just another simple mechanism design which is able to do something that was previously impossible I mean from a purely technical point of view Bitcoin is just evidently a better version of gold than gold is Gold's main value is that it is get it is there isn't going to be a massive influx of more gold into the system but even in that case it's fairly easy to forge gold and indeed we don't know how much gold is out there on asteroids and other planets and so on so it's not actually that's guess whose Bitcoin exploits cryptography and Mathematics to enforce that scarcity making it at least at a technical level simply a better system that's the basics of how you programmed humans you find a way of getting between the human and the sub goal that gets them to where they want to go then you generate a mechanism which rewards them for doing a task that you want this is extremely powerful so powerful that it's actually more it can overpower the moral Frameworks that most people live by in ways that are dangerous and I guess fundamentally disconcerting the first time you would see them I would say um this is not Theory again this is practice so this is Maurice Maurice was a governor in the 1920s 1930s in Canada Maurice was responsible for the children's homes at the time as well as a psychiatric ward where he's noticed that the federal government would give him two dollars a day for the care of a single child in these homes but he also noticed that he got seven dollars a day for care of mental health patients in their appropriate places then he realized well okay so I could just take these 20 000 children that are supposed to be in children's homes and I can put them in Mental Hospitals and I could get three and a half times the reward and he did this in practice it was real it caused 20 000 children to be misdiagnosed with mental health issues that they didn't have who were then forced to take medication for those um for those non-existing conditions with all of the effects on their lives that you might imagine this is pretty terrible in fact this is foundationally terrible and this is a result of a mechanism design that was not deliberately and carefully thought through but in decentralized protocols it's far far worse so at least the Canadian government could come along and they could say okay this mechanism had a bug let's change it and then the world moves on the the damage is done but you can at least change what happens in the future with decentralized protocols reinforce games that cannot be terminated or updated that is once the mechanism designer puts it out there into the network there is no way that they can then go and change it they have to get the ascent of all of the miners in the network or all the participants and that's not a predetermined um outcome like again in practice so in 2010-2011 Satoshi Nakamoto commits the one megabyte block limit to the Bitcoin uh git repo doesn't really say much about it just says enforce one megabyte block them in okay fine it wasn't a big deal back then when there were almost no transactions in the network but then fast forward to 2017 when Bitcoin is arguably having its first shot at Mass mainstream adoption and there are thousands of miners out there and they are highly economically incentivized by the network like the the mechanism design is really working it's enforcing scarcity and it's rewarding people for accepting transactions looks great but then what happened well we start hitting this one megabyte limit and that means that the users of the network have to start bidding for their places in blocks so they're trying to bid higher than each other to make sure that their transactions are included in a block first or sooner or even at all and in the peak of the 2017-2018 cycle the transaction fee inside Bitcoin got to 30 dollars per transaction now if you're a miner of this network this is actually great news you've got 12 of these happening every second uh so uh in the average block time you're getting pretty significant amount of money relative to the block reward just from these transactions and so you now have no incentive to change the protocol such that the number of transactions that can be included inside a block increases because if people are paying you thirty dollars per transaction now and they were going to and they're going to pay you say a cent something like that when you increase the block size then you have to have 3 000 times as many transactions in order to get the same reward as a miner so you simply don't have an incentive to do so and this one simple error in Bitcoin led to the fragmentation of the community and arguably a missed shot at mainstream adoption but this is actually the nice side of it gets much worse okay Bitcoin uses proof of work that is every time you have a block we guess a random number this random number is hashed with a challenge that's created around the block and we see whether it is preceded the output of this is preceded by a certain number of zeros so if it's below a certain number this is how a proof of work works and this simple mechanism is very effective at ensuring that an amount of work goes into producing a computational output it's extremely simple it works very well but of course it burns enormous amounts of electricity producing very little useful work just that we happen to know if you put this and this together it produces a hash that has a certain number of zeros kind of like a mathematical Oddity nothing more really there are many other systems that could have been used we could at least have put that computation to something valuable like I don't know folding proteins right protein folding has the same structure of very hard to find the right way to fit the uh the three-dimensional shapes together but once you do find the way to do it then it's very fast to verify in the same way that if you take so it's hard to find these mathematical Oddities but it's very very easy to verify them so you just hash it together it's a single hash you compare the value um yeah there's a few operations in x86 so we didn't do this instead we just burned electricity doing nothing again massive massive incentive design flaw but I think that this could just be the beginning of how these mistakes go in the future if we're not careful and that's why I've put so much stress in this lecture I think as an industry it is our responsibility to make sure that we're doing this in an appropriate and moral way in AI safety one of the key things that we look at is how do you make sure that you're going to be able to turn off a generally artificially intelligent machine if you make one if it starts to do things that are against human safety and good well the last thing you do is you put it quite literally perhaps the last thing you do is you put it inside a decentralized network where humans are selfishly incentivized to keep it running to keep it powered on if we do this with an artificially intelligent entity essentially that we we don't know how to control and that isn't working out our interests the effects of that could be absolutely so I would just argue that we should um yeah we should think very carefully before we turn these machines on and we should make sure that our mechanisms are elegant and morally appropriate and in fact there is already a paper out there that studies the Bitcoin blockchain and comes up with some reasonably good arguments that people have already experimented with putting AI inside uh the Bitcoin blockchain but I'll leave a reference to that at the end okay so that's how you do it that's why you should be careful why you do it or how you do it and this is what it looks like when you actually go through the process so when you're designing an economic game it goes through three stages you need to choose a goal you need to choose a reward mechanism how you're going to make sure that the players in the game have some kind of output that pushes them towards one of their sub goals or goals in their life that they're attempting to achieve then you need to create a function the deportions that reward to people in a way that will keep them satisfied while also achieving the hopefully pro-social goal that you've set for the network so let's look at Bitcoin the goal of bitcoin's um mechanism is to maximize the network security and avoid double spend one interesting thing about cryptocurrency history is that cryptocurrencies have been around since the early 1990s they all had this basic structure of some kind of centralized database and then keys that were held on the person's computer or in whatever device that they were using and you would essentially sign a transaction send it to the database which would update the ownership records of who had how many tokens of which kind and that would be it but the problem here of course is that this centralized party the database could be messed with and no one would really have a record of it so some of the early advancements were like okay well what if we use hashes to create work chains that showed that actually this couldn't have been altered but or at least couldn't have been altered without the record also being altered so that that's kind of fine um but of course it's not a fundamental answer to the question and so then Along Came Satoshi Nakamoto with the idea of a blockchain in 2009 that says okay fine what if we replicate that database across all the machines in the network and then we create a mechanism design so an incentive a game really where one of the helpful byproducts is that their incentivized to accept transactions and to follow a version of Truth follow a version of history in the network where there aren't two transactions that take the same coins and spend them twice that's a fundamental idea okay so the second thing tokens are then distributed relatively to the security contribution of each Miner inside the network this is a pretty simple way of pushing those miners to contribute as much security as they can for the tokens and if we think back to that example they gave with this generalizable mechanism mechanism of Mining and how you can use it in centralized startups to just push down the price of a of a service essentially towards its base level this is exactly what's happening here so we're saying I'm going to give you 12.5 Bitcoins and you just give me as much work as you can to satisfy that 12.5 Bitcoins and then the price of the Bitcoin determines how much work is going to go into satisfying that choose a reward function to match it proof of work block acceptance and Fork avoidance and recovery so this idea that we're putting extra security we're producing blocks which have transactions in them and then we're also creating the incentivized to take the most worked on chain uh otherwise there's essentially this incentive just to ignore everyone else's blocks and only mine your own and then of course everyone in the network is running their own blockchain and of course that doesn't hold any value so it's very important also that the incentives push people to accept each other's blocks okay now let's think about our weave which is I would say so later generation in the same kind of field of decentralized networks and mechanism design so our goal the thing that we want to get out of this uh pro-social byproduct is to secure data forever the idea here is to build the new Library of Alexandria where everything from documents PDFs web pages applications really anything can be stored and perpetuated for future generations to come so that we don't forget what has happened in the past and we can replicate ideas and essentially memes in the non-cat gift sense but also those as well into the future so The Helpful byproduct we're attempting to achieve is the permanent storage of data what's the reward mechanism going to be users contribute funds to an endowment when they add data and then miners are rewarded from the endowment over time seems pretty simple so of course if you want to store data forever you have to have some economically sustainable mechanism of making sure that the data is or rather the reward is available indefinitely and so for this we do something that's actually quite simple but it sounds a little strange to begin with so when you put data into the rweave network you pay for 200 years worth of storage of that data up front this you would think is expensive but storage is so cheap that actually this is like half a cent per megabyte so if you're thinking about web pages PDFs applications this kind of thing it's extremely affordable then over time that 200 years worth of storage as the cost of storage declines you gain essentially a kind of interest on that principle in the form of storage purchasing power so over the last 50 years the cost of storage has declined at a rate of around 30.5 percent every year on average the RW Network assumes that the cost of storage will decline at some rate above 0.5 percent on average and if it does at the end of each given year you end up with more storage purchasing power at the end than you did at the beginning and consequently you get this sort of compounding effect where the available reward for storing a piece of data actually increases over time relative to its burden to you rather than decreases so it's essentially like an endowment for University scholarships something of that ilk okay fine so we've got our mechanisms here and now we've got to work out how are we going to adjudicate how that reward is distributed so this is just another piece of mechanism design proof of access so a normal blockchain works by including the hash of the previous Block in the production of the next block and it goes on over and over again and when there's a fork you get two blocks that have the same hash of the previous block and they essentially uh yeah they they fight on the network for the one that can accumulate the most work and then because of these Fork avoidance mechanisms as described earlier they all sort of um they said well they can join on the same block and they move forward from there so it has this Fork avoidance mechanism too so with are we what we did is we took that same fundamental structure of a blockchain and we modified it so instead of just including the hash of the last block in the production of the next block you also include a random proof of access to a previous piece of data in the network in the production of that next block so you can't take part in the production of new blocks unless you can prove that you have access to old blocks and we can stochastically so statistically check the data in the network at random and as long as people can't predict which parts of the network are going to be essentially recalled for any given block there's no better way to play this game the dominant strategy in this game is to replicate as much of that data as you can so this mechanism lays the groundwork for a network that provides the possibility of business models that you just couldn't pursue previously there's simple stuff like let's make sure if we store if we make a contract together that we can store an immutable copy of that contract uh embedded in the blockchain so it has this kind of time stamping that's extremely effective and reliable as well as the permanent ability to recall that contract so we can tell it's essentially the shared record of the truth between us but you can also use it for academic publishing right so when you want to contribute to science what you're trying to do is just push a piece of information to as many people as possible well just like the Internet connects people over large distances the r weave connects people over long distances of time and essentially by conjoining these things together you can get a larger spread of your academic article than you could have otherwise but you can also use it for journalism right when you want to publish a piece of information you want to get wide circulation and you want people to know that it was you that published it and so essentially in this way our weave is acting like a um a ledger of every piece of information that was ever asserted by someone over time and these are just the simple use cases that you can use with our leave it goes much deeper than this on top of the iweave network you can store basically any piece of data and that means that you can also Store applications and these are essentially web apps that are accessible in a browser so you go to a Gateway stroke a transaction ID and it'll render for you that web page and that web page could be an application it could be just a normal piece of information a Wiki page really anything and the applications that you store in the system have a set of extremely interesting and different properties than applications that you find on the centralized web so if you're using a perm web app it cannot change the developer can send a new version to the network which can use all of the same data if they like or data even from other applications but they can't take back your access to the first application and this means that you essentially preserve what we call consumer Integrity that is if you find a perm web app that you like today you can make sure or you can be sure that that application in exactly the same state is going to be available at any point in the future just to make this kind of clear why this is valuable let's think about Gmail so we all made our Gmail account say a decade ago right we all signed up we pressed through the we pressed through the screen that said make sure you read all these terms and services as fast as possible no one cared and suddenly we had this identity which was associated with an account that was held by Google I had no idea 10 years ago what Google was going to do with my Identity or with the service they offered me what they were going to do with my data how they were going to use it for advertising and indeed the service that I would receive and then you're stuck you're locked in now we can't really move away from Google for Gmail well there's an email system on the are we've called weavemail weavemail is available as a protocol and as an application so there's a general specification for if you would like to send a Webmail message it must conform to these properties but then it's available in any weavemail client there's also an application so you can be sure when you go to this transaction ID that you're going to get that same interface back every day in the future simply forever way past the end of your death and a consequence of this is that you can be absolutely sure that if it's not now mining your data if it's not now showing you adverts it just won't do so in the future this also changes the incentives for the developer now when they want to issue you an update they have to give you an update that is actually of a higher quality than the previous version this we think is extremely powerful if it doesn't stop there let's think about online social media Networks so social media networks generally go through this pattern of at the beginning everyone loves them because they have a small community that is aligned around the same kind of ideals but then over time the community gets larger and larger and splits into different factions that have different ideals and then you have this question of what content should be allowed on the network or not and it becomes the job of the developer to adjudicate this and this as a developer this is just the worst thing in the world you don't want anything to do with this is a purely social problem to which there is no good answer if you censor too much the Free Speech people are going to hate you and if you don't Center enough it's censor enough everyone else is going to hate you this is a no-win situation if only are weave with permo web apps we've taken approach which says Okay so the applications in the network are truly permanent and ownerless once you launch them so once they're sent to the network anyone can access their transaction ID and the owner doesn't have any further rights than anyone else in the network to adjudicate how that application works so let's think about decent a forums for example it's a forum Network built on top of the rweave where the developer has launched it they paid less than a Cent and users pay a tiny fraction of a cent when they want to post a piece of information to it and now it simply ownerless it's run by code so in the same way that in ethereum code is law for smart contracts in our weave web applications their code is law for adjudication of content model moderation and that means that if you start a community there you can be confident with how that Community is going to be moderated in the future fundamentally changes the game these things together we think enable an ecosystem of Open Source web applications so one of the questions we were thinking about when we were getting started building the permal web is why is it that in the desktop space and in the service space in different ways but it's true open source applications have taken so applications that are built by people that don't necessarily have a clear business objective for doing so have been produced and are used by very large numbers of people in fact open source software fundamentally runs the web today with Apache and so on and Linux of course why didn't that happen on the web and we think the answer to that is pretty simple it's because the web application incentive model is broken so let's think it through like you build a really fun little side project on a weekend you launch it you put it on a seven dollar a month AWS server everything's great even on hacking news whoa suddenly you have tens of thousands of people using it now your seven dollar a month server is not enough okay so you're starting to spend eighty dollars two hundred dollars you're suddenly realizing that your usage costs are scaling with the number of users yet the users want something that is free and this gives you one of two difficult options you either have to take VC funding very early on or you have to monetize very early on so VC funding is fine it's certainly one way of doing it but this pushes towards an exceptionally exceptionally high growth model which for something that is just a weakened project might not actually be what you're going for and if you monetize early you know that this is going to kill growth and particularly with applications that require Network effects to start going so perhaps social applications this might remove the possibility of your application being adopted in practice at all from web apps have a fundamentally different model you pay one cent per megabyte less than a cent per megabyte when you upload the application in the first place and after that you are done with it that's over your that's your contribution done and then users when they use these applications um they can essentially pay tiny tiny fractions of a cent in order to store data in the network as a as a result of an interaction with the app and as such they're essentially paying their own way and the costs are not accruing with you yet the cost is so low that the user doesn't even realize it's happening really they just think they're logging in with a wallet of some kind and then they have you know a very small amount of money topped up on it and once in every two or three years for something like Twitter then their two dollar Top-Up Runs Out is extremely simple changes the incentives entirely and it works exceptionally well and we think that it enables you to build an open source web ecosystem today in fact we don't think we have already seen it in practice there are already 200 applications built on top of the r we've that are going down exactly this route with growing communities around them so the broad idea with what we're going for with our weave is a global immutable record of knowledge and applications that a applications and history that are cryptographically verifiable and unchangeable over time okay thank you all for listening uh if you'd like to get more into the economics of mechanism design from a sort of different standpoint with a large amount of mathematics Focus then I would thoroughly recommend Tim rough Garden's lecture Series algorithmic game theory if you want to come work for our weave these are fundamentally the uh yeah required reading required listening actually in this case in order to to work with us and I would say they are the the most accessible um economics focused introduction to mechanism design out there you could also look at Trent Trent from Ocean protocols towards the practice of token engineering so what I've outlined for you here I hope is how mechanism design works what you can do with it in practice what the pitfalls are and then Trent's piece takes you through okay so if you were to actually think about this from like a tool chain you know what are the applications you should be using to try and experiment with this and so on then I would thoroughly recommend this blog post by Trent and then of course there's this artificial artificial intelligence on a blockchain paper that I mentioned earlier okay thank you very much awesome thank you Sam that was great I don't know if I I can't see you guys right now but I'm sure people are cheering in this like um so we've got a bunch of questions from from the audience and I'm gonna sort of moderate them so you guys feel free to keep them coming in the slack but um first one is Paul cogwell Cowgill sorry um and his question is with the event of AI and modern Computing done some of the arguments against Central planning based on computational throughput go away yeah I mean that's very interesting I think there's definitely an argument to be made that essentially what we're doing with uh modern Ai and perhaps if we get the artificial general intelligence is we are centralizing the crowd right so if if we're trying to make one of these mechanisms that rewards people for using their brains to essentially optimize some problem yeah if we can spin up computers that are capable of doing this for us then definitely like yeah it's just another way of addressing the same thing right um how computationally efficient those algorithms are right now and will be for the next say 50 years Israel question like for AI startups it seems to be and I think there was a great a16z piece on this recently that AI started suspending inordinate amounts of money on computation so it doesn't necessarily have the same same scaling mechanics there's definitely an interesting argument to be made the kind of the wisdom of the crowds could actually just become centralized in computers again cool okay so next one is from James and um his question is people sometimes talk about useful proof of work but if the work we're useful or valuable in some new way you can imagine people solely paying people to complete this work or doing the work for their own sake such as GPU rendering or ml model training wouldn't this disrupt bitcoin's mechanism doesn't the work need to be useless so partially I mean well there's one way of thinking about it that's like that but another is that as long as access to basically solutions to those problems already so like a rainbow table essentially as long as that this is equal then no no it it still works the competition mechanism is still functional but of course you need that that access to the rainbow table to be uniform between people otherwise you get this uniform Rewards although you could argue then that there's another way of looking at this that says well in that way you know there's there's uniform access to um electricity at certain prices inside the Bitcoin Network as it is now so I think the problem is essentially just another analogy for that interesting question though all right next one is from Paulina and her question is in the context of coronavirus how would you build an incentive mechanism for people to stay home instead of travel around especially that one incentives are most often non-monetary and two there's no single party that would be interested to provide those incentives appeal yeah that's a very interesting question I mean you like a basic way is you work out some kind of task where the human has to be in their home and then you work out well yeah I can't work out what the concrete example of this is but I can take you through the process of doing it abstractly right you work out some kind of task where the human has to be in the home you work out some way of verifying that that task has actually happened and then you put some reward mechanism so maybe like a bunch of dye that you release to people for submitting proofs that they are in their house or something like this to the to the ethereum network yeah interesting question I I think there's another part there and I actually wanted to edit my slide deck today to add this but but I didn't about um the mechanism design of panic buying right this is fascinating so mechanism designs are not actually always just designed mechanisms can be emergent and this is one of those which is essentially that if you're going to Panic by then I should also Panic by in fact I should front run your Panic buying I should make that make sure that I'm buying before you are and if you think about this in practice this is recursive all the way back so they're extremely dangerous mechanisms at play in pandemics same same with this question of masks and this is particularly interesting uh to us at are we because we're we're kind of fascinated by information availability over time so there is evidence to say that at least a medical mask is better than zero right you are better off wearing one but in society we are all much much better off if our medical professionals are wearing masks so now there's this strange information game at Play where some people in the the media like to tell us that no there is no real point wearing a mask which is kind of true but not actually completely true but it is in the best interest of everybody if we believe that to be the case regardless of whether it's true or not so yeah there's a lot of interesting mechanism design that goes on in pandemics or at least mechanisms not so much design right okay great next question from James how do we do better than terms of service when most people cannot or do not want to read through the code of the applications they're using immutable is good but what about what it guarantees how an app is going to behave so that's a really interesting question I think that you can gain reputations around pieces of code that are held at certain transaction IDs is essentially how it goes so you're right nobody or very few people are going to go and read The Weave mail source code but they could and a few people will and they'll say this transaction ID has these properties and then others will will read about this and there's a kind of I guess proof of authority mechanism in place where if I see enough code reviewers that have reviewed this one specific transaction ID that is immutable and permanently available um yeah then I guess I can trust that by proxy that I think is the way that we do this it's it's fundamentally different though from the terms of service approach where almost nobody's reading them and they're changing all of the time as well um so that I think is where the core differentiator comes right okay another one from Paul what are the best practices for determining the potential pathologies of a given mechanism ahead of time for example using a tool like Gauntlet or deploy orowion iterate based on how it ends up being used how can you take into account behavioral economics in addition to just Game Theory yeah this is interesting uh people are starting to build tool kits now that are trying to work this out like CAD CAD is an example of this which is pretty good and and Trent mcgonoghue goes into much more detail about this in the blog post that I pointed to so so that's something but I would say there's a lot of work that needs to be done in this area like if we're being sensible about it we need to get our AI to try and essentially work out the dominant strategies given a mechanism and if that AI is smart enough then then that could be really valuable to us but what I've seen at the moment the limit here is like we'll just try the Bitcoin mining dominant strategy or we'll try the ethereum mining dominance strategy inside whatever game we're presented with which is extremely dangerous because just like we saw with the example of Maurice right um the Canadian government didn't realize that they were building a game that was fundamentally different to ones that they'd used before right so they'd had this we will give people X for doing um for um looking after a person for some time uh yeah this this had been in place for a very long time it was when they added another one that actually the perverse incentive arose by the presence of those two games in the same space and if they would have used one of these systems that said well just apply the dominant strategy from some other known game inside this new game they would never have picked that up so unfortunately the answer is the abstractions are very messy for this and we don't have good tooling for working out how to how to do this in practice um but hopefully hopefully it'll get better and at least people are working on it now next one is from Nori and the question is for our weave how come you don't have a delete function seems like you could do something where the owner could choose to delete something and perhaps the owner could be refunded half their endowment this wouldn't take away anything from the uncensorable nature of the network since the only only the owner could choose to delete it uh no it definitely would I mean then you get a five dollar wrench right and you force me to delete it or you find my computer and you delete it for me um no it fundamentally changes the the mechanism in practice I mean that we could go down there's a very interesting rabbit holder um it was also computational issues that arise from holding that kind of state but in general I would just say yeah five dollar wrench is the problem there what can you unpack the Five Dollar Rent problem sorry of course right I think this came from an XKCD comic uh and basically it goes something along the lines of some person is um boasting about how strong their encryption is and then some other person comes along with a five dollar wrench and threatens to hit them with it and now suddenly your amazing encryption is zero got it I I was hearing Five Dollar Rent I didn't hear a wrench but that's even more okay good all right next one from Abhishek what is the economic role of the game designer in creating and sustaining a game that's an interesting question it's almost like asking the question what is the mechanism for the mechanism designers I guess um normally normally the maximum mechanism designers design some special place into the mechanism themselves where they gain some kind of reward and then it's up to the participants to decide whether the mechanism that they've designed for themselves or the component of the mechanism that they've designed for themselves is fair and of course at the moment we see this constant battle in the crypto space with like founding teams that are all like oh yeah no we should bake ourselves into the protocol as receiving a reward for maintaining a version of a client for the network which I personally think is fundamentally a terrible idea because the reason Protocols are cool is because they have decentralized clients right the single application is not the protocol they're very different things and you kill that completely if you just give one group all of the reward for doing that and we see it playing out all over the place that this is just a very bad route to go down can tear communities apart okay so next one from Sarah what is the difference between designing reward mechanism versus the reward function they seem very similar yeah they are conceptually related one is like what am I going to give you and the other is how am I going to give it to you how am I going to adjudicate which bit of the thing are you going to get but often they come pretty sandwiched together that's true all right I think this is going to be the last one and it's from the Cyber code twins I do wonder how is the gdpr friendly given the own your own data movement who owns the data if it's immutable on the chain uh that's a fascinating question um to which I am not actually the best person to speak to I mean it's a it's the same question as like okay so I put a tweet out onto to Twitter and then lots and lots of bots who have accessed that tweet do you really own it do you own all of the copies of that tweet it's not clear um but in the in the case of Are We gdpr is actually very simple so it's you would be surprised but it's actually better than the centralized web that's because you can just tell me okay everyone here's a transaction ID I would like you to delete it I'm this person I own that transaction um and here's why I would like you to delete it then it's up to the miners to decide whether they do so or not but of course if those miners are in the gdpr compliant country they will delete it so one of the things about the our weave network is the permanently embeds and incentive to recall a piece of data does not enforce any individual person to store that piece of data so then it's just up to what each participant in the network deems as a reasonable way to act and it's the same way as if all of those miners are tiny instances of AWS like they are all data hosts and as a consequence they are essentially um well just as responsible as AWS or Google should be so it's up to them to to make the mechanisms but we give them the underlying tools that they need to communicate that people want things deleted they want their access rights changed um yeah so they can deal with this and the same for illicit content on the network so that's what I would say to that when I get oh and and the final thing yeah and we don't think that every single piece of information in the world should be public and permanent it's very clear that they shouldn't um but we would say that the information that should be permanently stored is very poorly served by the available data storage mechanisms outside of our weave and that's where we fit in essentially great okay thank you so much Sam of course [Music]