EthCC 2: Axel Simon

alright say well okay okay yes y'all okay all right um hello everyone thanks for coming being here not leaving the room let's get going we've done the coffee break I think so I'm just gonna get started and and hopefully it'll be interesting so yeah sorry if some of you wanted yeah exactly okay thanks a lot so my name is axel Simon I work for a company called red hat which some of you might know red hat does open source software we're one of the biggest providers of open source software in the world thanks and then we have an interest in blockchain in general because some of our customers are interested in blockchain we provide a Linux distribution called Red Hat Enterprise Linux my middle where we provide cloud operating systems like OpenStack and stuff like that so we're interested in what our users our customers our communities are interested in which is why we're interested in blockchain and I'm interested in blockchain because that's my job and I find it interesting today I'm going to be talking about trust and this idea that I've seen floating around about commoditizing trusts so making trust into a commodity we'll get into that I've found that there's been some interesting conversations today already so if you talked I thought were interesting and I hope this will add to this conversation talked by martin by blood by primavera well that was interesting so I hope this will sort of go in the same direction so we're gonna talk so I'm gonna be able to skip probably the first two here it's gonna be a bit funny who here doesn't know what a blockchain is that's it okay so I can skip that and who here doesn't know what a smart contract is okay so we'll go a bit more on that so I can go really fast on the first two then commodities that might be a little less obvious so we'll go a bit more into that and then we'll go about in the different things about trust how can things be trust lists how can they be trustful I'd like to say trust fish sometimes really it's not it's not that clear at all and and then what how could we commoditize that or what does that even mean and I think we'll find that it's a bit more complicated than that so the blockchain is I'm gonna skip there's lots of stuff but basically a distributor replicated peer-to-peer network of nodes that make up a distributed database etc etc so I can skip uncle's smart contracts right so the thing about smart contracts is that probably a lot less in this room this slide works better with people who don't know what smart contracts are but a lot of people probably think of contracts like paper contracts and the problem with smart contracts is that they're not smart and they're not really contracts so the name is not very good but they have other interesting properties let me just keep going here yeah they they're not that smart they probably as smart as a smart dishwasher they're really just networked programs their probe their computer programs that I sell to a network the nodes on the network executed most of the time it's all of them sometimes with new sharding ideas it's some of them but they're not exactly contract there's another thing called Ricardian contracts which is much closer to an actual legal contract but this is not what mostly people talk about when they talk about smart contracts is there anything else I want to add here yeah one term I've seen that is interesting is sometimes it's also programmable money like money that you can program when it's for transsexual for financial transactions or just distribute it out it's probably a better way of calling them really the smarts and the smart contracts are in the people who are creating them they're not in the contracts themselves which is something that will matter again later on so what are commodities so if you ask Wikipedia Wikipedia will tell you that Eagan economics commodities I'll read it a commodity is an economic good or a service that has full of substantial fungibility which is helpful only if you know what agility is now this is a crowd where I think more people than average know what fungibility is usually people have no idea what it is so we will go on onto that a little bit so fungibility is essentially the property of goods or services to be indistinguishable from one another you can replace one with another for instance rice is a good example when you buy a bag of rice that weighs one kilo you don't care what grains make up that kilo specifically you don't go and choose them you just care that it's one kilo of rice now there is a caveat here it's that you do care depending on who you ask you do care about what quality of rice it is or what type of rice it is so if it's organic or if it's not organic so fungibility is not just a simple it's not black and white if there is there is a continuum and that's also something we'll get back to later but essentially fungibility is the capacity or the property of some things to replace one another without one caring it's not necessarily liquidity which matters more in a financial context but we'll get to that right now because fungibility is also a property of money which is a great way of looking or it's an other good example to understand what that is if I have to give you back ten euros that I borrowed you that borrowed from you you don't really care if I'm giving you back the same note or another note to five euro notes or one 10 euro notes it's all indistinguishable so that's that's what fungibility is about and you can also add more caveats like some money notes are tainted or noted by the government to be known to be used by bad people and all that so that reduces fungibility but the core idea is still there so maybe now we can try and explain what commodities are so commodities are things that are both standard enough and readily available enough they're easy to come across to come by that they can that when you consume and you consume them in a in a predictable way you basically know what you're gonna get every time so we can we can say basically at some services are and some goods are commodities the examples I think are quite reasonable electricity you don't really care I mean it's indistinguishable where your electricity comes from some of us do care what actually generated it but the actual electricity that powers your stuff at the end of day is going to be the same one most of time a taxi ride is going to be the same you want to get a cab to go from one place to another you can basically expect what's gonna happen it's gonna take you from one place to another and it works also for goods obviously bananas rice wheat blah blah blah that sort of stuff so commoditizing is the act or the results of when something becomes a commodity so some things might start being very customized and very specific and become commodities over time for example something like a realize that I didn't put down and an example things that were hard to come by for a long time have now a lot of them have become a lot easier to come by phones have become a commodity for instance at one point a mobile phone was hard to find and it was not it was not very standard you wouldn't even I know the first generations of phone now it's a lot easier you can pick a phone off right in the shop and they're very easy to come by they're pretty standard and what they do on trade has made lots of phone standards so again as with fungibility it's not it's not binary it's not either it to come out of you it's not it's a spectrum you know to some people a bag of rice a bag of rice but as we were saying to other people different depending on what kind of rice it is that makes a big difference not always but to some people it does so there's obvious there's a spectrum here so right so what so now going to something a bit different but we'll which will all tie together trust trust less trustful sirs trustless is a word that gets used a lot I think in the blockchain world people are very interested in this idea of doing things in a trustless way actually the Bitcoin white paper that even mentions at the end it says we have therefore we have shown that we can build a system without relying on trust I think that's not true your there is still trust you're just trusting different things now you're trusting the code you're trusting that it's properly implemented you're trusting that the logic that it implements is correct you're trusting that rather than institutions or you know a lot of these a lot of time these institutions are opaque so that is that's interesting but there is there is not a lack of trust you're just trusting something else and you're making that choice that's an interesting choice but I think it's still too simple to say that there is note that you're not relying on trust really it's not about zero trust it's it's about choosing who and choosing what you trust still there is there is some positive I mean I think that's positive it's this idea of trust reduction we had it we used to have food we used to we still do a lot but in the normal usual world you trust institutions to a large extent and in the blockchain world you have the potential to trust these institution less and trust may be a reduced set of things which is code which is the people who order the code which is the people who write their code so still there is a difference but but it's not about there's it still there is still trust so that's why what that's what I would call maybe trust ish which is it's not complete you don't have complete trust in it you might verify you might do things like that it's not either trustless it's you know somewhere in a continuum so this idea of commoditizing trust this is idea I came across I think it was a talk a few years ago by someone who is basically arguing that now we were the poor we're with smart contracts we can just commoditized trust like Trust is becoming as easy to come by as a banana or you know turn on the tap and so you know you turn on the tap you get water you use a smart contract you have trust that's it problem solved I think it's a very nice idea I mean it's very tempting it's very it's a lot of earlier to that to that idea it sounds very nice I mean if you could immediately trust people on the other side of the world without having to anything but use a technical system sounds it sounds brilliant it does sound great but yeah I don't think that's really true I don't think it's as simple as that I think that these different things we've seen like in crypto we trust that people have printed t-shirts with that there was an article wasn't in the government refers in the New York Times and one of the big American newspapers which called a blockchain of the trust machine I think the problem with these ideas with these ways of putting it is that they give the impression that if you trust a technical system that's it you've solved the problem of trust between humans that you can reduce human trust to a technical problem and in my experience and I think it's quite true in general when you try and solve human problems or social problems with just technical means you don't really short solve the human problems and you miss quite a little bit of you know quite a bit of stuff and and so we have this idea that you know here you can commoditize trust this human thing you can package it into something as simple as lines of code and I'm not trying to say that you know this is magical humans are so complex that you know we can't be reduced to some rules and stuff like that I think there is there is something to be said about those about you know about understanding human behavior seeing the patterns through which we function psychology has a lot to teach us about that but simultaneously I think we should be we should be wary of just simplifying things to technical solutions so yeah one does not simply commoditize trust it's a bit more complicated than that so I think maybe this means that we need to talk a bit more about what what Trust is does this concept seems to be sort of used a lot all over the place and it's it's a bit fuzzy I think because really trust it's it's a human value it's a thing that happens between humans and what stuff that happens between humans is fuzzy it's not it's not clearly defined it exists on continuum yesterday there was an interesting talk here too about traditional trust systems in Nigeria and how people understand and work with each other and trust each other and they don't they again they exist on Continuum's they exist in old running systems that have exist for a long time or sometimes in new systems but overall there are are very very human base very very fuzzy really is I think it's one of the best word and so trance is fickle it comes and goes it's depending on a multitude of factors I mean you might trust someone at a certain point in time it not trust them and other time and entrust them again you might trust people in some context but not in others like you might you might trust someone to be very good we're doing technical stuff but not very good with doing human interactions or vice versa you know you might trust them when they trying to achieve certain goals and not others and so you trust some people some of the time basically so so Trust does require some technical solutions I think but it's only part of the overall picture and that's that I think is where the whole idea of commoditizing trust kind of breaks down so yeah so Trust can be in different things so we can have trust in people you can have trust in technical systems so that can be software that can be tools that can be machinery I mean when you're you know machine worker you trust the machine or the end of people who designed it not to basically chop your fingers off if you run if you use it correctly we have trusted social systems you have a lot of trust about that there's Bruce Schneier came out recently with an article about you know about block chaining wipes useless essentially and and one of the big things he mentions is trust and I think it's true that one of the things he says it's very true is that we are constantly trusting other people institutions to function without even necessarily thinking about it the trust in institutions in companies in some NGOs even in just support networks people around us and and I would argue that at the end of day the most important bit of all of this and it can also be a combination to the above and I would argue that the trust in the people is probably the most important part because the people build the tools and the people make up the institutions and so you can build institutions out of untrustworthy people and you can have trustworthy tools coming from untrustworthy people it's not impossible there are clever ways of doing that but I don't think you can really do it the other way around I mean I think you can have you know trust in people with wit untrustworthy tools or stuff like that I mean I think at the end of day if you if you think about it the trust in people is probably the one that underpins the trust you have in other things and because of that it's something we should really sort of keep having at the forefront of our minds so which can lead us to a bit of a weird question which is is trust fungible like can you exchange one trust for one thing for another trust to another thing without even thinking like without knowing what it's about it is not a question I have an altitude by the way I just think it's a weird and interesting question the idea of the fungibility of trust so so as you can see from what I've been saying and you can guess my idea here is that it's a bit more complicated than that like trust being this human fuzzy thing it's hard to reduce it to this this set of smart contracts or just to technical solutions and I'd like to offer you an example something you can consider and think about for a second this is something I've thought about and to me it's an example of why like commoditizing trust with smart contracts is is maybe not doomed for failure but has serious serious issues so this is this is the idea we create a system for microloans based on smart contracts on a blockchain that enables people to easily on one side give money or put money in the system and on the other side for the people to get micro loans that will help them start small businesses and stuff like that this is my concern so at one point one of the persons who borrowed money from this micro loan system which I thought a horrible name for which is micro Bloc loans you can feel free to take that idea by the way so one of the one day one of the people just they can't pay back their payment for a good reason what I think about immediately is you know somebody in your family has a health problem and you don't know how long it's gonna last and you just pay for their health cost you just have to it's more important than you know whatever loan you you took on but what happens next is that after a certain amount of payments to five ten fifty I have no idea at one point a system will basically find automatically because it's code that the person has defaulted and therefore that their loan should be cut off the collateral that was given for the loan at that point comes in jeopardy and either you lose your collateral or some other bad consequence happen but basically the person who got the loan their life takes a bad turn at that point because on top of having someone in their family who is sick they also lose their loan that was enabling them maybe to run a business or to oh they lose a collateral that came with it and so the immediate consequence at least to me the the reasonable consequence of that is that then when they next talk to people and people ask them what do you think about this loan system I've been wondering if I should get a loan from that their reaction will be like well at one point I couldn't pay my bill and I had a good reason but the system was automated and basically I lost my collateral and so don't use it I think it's bad I think you should not use it I think basically they will tell anyone who listened to avoid the system like the plague because it burnt them and so in this case the technical system never stopped working I mean there was no vulnerability in the code the implementation was correct the math behind it was also good the software was still up to date possibly the person's phone on which they had their client for that was also up to date like all the technical aspects they're fine they're good you know codes not broken math not broken code is working but but the trust in the actual overall system that's gone and that's where there's a real difference between like the technical trust the trust in the technical side and do you trust in the overall system and that I think is is really where is it of where to the limits to this idea of commoditizing trust here's a few things about why you can't come out of ties basically I think the first one is the first one is that and this was something I read not that long ago and really stayed with me and and maybe you'll find interesting too I used to think that the fact that contracts are ambiguous is a problem and it's great that smart contracts are not ambiguous because well you know who likes ambiguity it's annoying you don't really know where you're going with it you don't know what's going to happen it's hard to predict people don't like unpredictability as someone was saying earlier you know we like determinism we like know we don't like things to be an indeterminate state to tolerate the uncertainty as sometimes is that in psychology but the thing is that the fact that there's a lot of subjective determinations in contracts that's not a bug that's a feature it's done on purpose it's useful ask a lawyer they will tell you and it it was in a lawyer's article that I read this and I found it really interesting and other lawyers I think agree like no but they need that they need that to have some leeway in interpreting the thing and making sure that even if sometimes you derive you know you're a bit away from the spit from them from the wording you're not far from the spirit of the contract and it's same thing for judges judges need some ambiguity or at least some leeway in the wording to be able to apply laws if you remove all ambiguity in laws you're left with judges that are no longer really judging but they're just robots applying a set of rules and so this is actually an important feature of the system it's not a bug and we as I guarantee should think about should think about it that way and not try and correct it otherwise we're probably gonna you know have bad consequences another thing that's interesting is overrides an example is my manager he he works he's a volunteer first first aid responder so he works with people in hospitals and trying to help you know victims of accidents for and there are cases where in hospitals most of the time people of a certain hierarchical level are not allowed to access the met the medicine room like then allowed to go into a room where they can get drugs out and medicines and there's a good reason for that you want to avoid theft you want to avoid people you know abusing those those substances but there are also cases where all the people in charge are not available either for an because they are in an accident themselves or they're taking care of someone else who has it who's was just in an accident in cases of large disruption of the system it's important to have an override where those people can go into that room and they can take the medicines that I needed by the other people basically you need overrides and so overrides means more conditions means you have to consider them or sometimes you have to consider that you can't consider all the cases and accept that there is a moment where somebody can push a button and just override all the rules and that's very hard to do in smart contracts I mean you know following the Dowell mess one decision that was made was will roll back the blockchain but that's a hard thing to do it we even create rifts in the community so overrides is us something you have to consider because human systems need them at the end of day we it's really hard to do without basically and I would also argue that and it's another argument against commoditization is that a consult I'll read it because we're all gonna read it together otherwise who is writing the code is very important and who is reviewing the code is also important and this important of people producing and checking it contradicts completely the idea of a commodity I mean do you have any idea who produced the aluminium that is in your phone or in you know in your kitchen or you know who produced a wheat that is in the pasture you eat probably not because that's a mark of a commodity you don't care about who produced it it's essentially indistinguishable and even if you go all the way down to saying it was organic bronze extruded pasture blah blah blah it still doesn't really matter like what actual you know if not the person I mean the person who did it nice job thanks but it's not you know it doesn't really matter who they did it brands can come into play but for true commodities they really basically they disappear and that's one of the masks of commodities is that the branding it becomes kind of irrelevant in this case we really care about who's producing a smart contract we want to know who's written it are they qualified are they good do they have a good record if they've made mistakes in the past have they corrected them the people who are auditing this stuff are they also qualified to audited are they good have they found mistakes in the past or have they missed mistakes in the past all those things really go against the idea I mean I think these are all things that we currently need and want for smart contracts and they go against the idea that you know you can use smart contracts as a commodity and therefore commoditize trust so one obvious answer to this is well you're simplifying a bit we can make smart contracts quite complex we can add more conditions and we can try and take care of the edge cases and of course that makes sense and I think to some extent there's there's quite a bit of truth to that we could probably do do quite a lot of stuff but if you start adding more complexity to your smart contracts to start taking care of H case number one then the third then a second one to the third one and as you go along you discover more and more edge cases you end up basically with a really complex thing and and so it's really it's really tempting to think that you know we're gonna sort of simplify all this messiness into into some well-written well-documented standardised code that was going to cover all the news cases I think you'll realize that you know when you try and simplify all the messy fuzzy human relationships down to that down to all the cases that can happen you're either gonna realize that you can't or that the result is such complex code that has so many branches and so many variations that it becomes a it defeats the point of having that in software to simplify the world because it's so hard to maintain it's so hard to take care of all the edge cases that you end up spending more time doing that than just you know doing it the old way and it was another point yeah I'm not getting it but basically I think is it there's an inherent paradox there in that the more you the more complexity you're gonna add the more you're gonna you know the more complexity add to manage the edge cases the harder it's gonna be to maintain and the harder it in the less standard it's gonna become and Beth therefore that's the point the less of a commodity it is the whole point of the commodity is that its standard and its predictable and it's basically simple and in this case if you want your smart contracts to be usable and cover all the all the things they need to cover they become super complex they become non-standard they become you know really customized and that sort of defeats the point of commoditization another thing to say I mean I'm gonna offer an argument against what I'm saying immediately which is I can I can imagine like building blocks like minimal building blocks that can be put together that will be really audited really verified maybe even you know formally verified code that has been properly you know thought of and everything but even then there's likelihood that the way you're gonna put all those blocks together is gonna be problematic or is gonna react complexity and uncommon it ties it I don't know if that's a word but you know D commodify I I think there's a paradox there and I think it's the inherit paradise oh you know commoditizing something as fuzzy as trust here's my little sort of slightly more philosophical slide I think we can wonder what this means about ourselves like why are we so attracted to this idea that we can just simplify trust to to something that is a commodity I think it comes down to our like I desire to die have a controllable manageable world around us there's a thing about human beings is that we work in patterns we like patterns they're recognizable they allow us to make sense of the world when we've seen a situation a few times then we'll more easily see it in other situations even if the situation is different we have that bias and and we are yet we are attracted to patterns it's how it makes sense of the world I think this attraction to simplifying like the fuzzy weird nature of trust to just to just proper code comes from the same idea of trying to manage the world and turn it into something understandable and palatable so in conclusion I'd say that we're not really the offered the propositions we've had about commoditizing trust through smart contracts and not so much just about commoditizing trust in a system but not really commoditizing the trust we have in other people I think and there's value in that there's value in simplifying you know trusting systems and tools but I don't think we're really getting to the point of doing something about you know the people that we there are other people um so this can be useful it's a lot more limited when you realize that you're not trusting people but you're just trusting you know a system and it's probably not even that new I mean when I pay at the shop with a bank card I'm already trust in a system and already trusting that that system is gonna give the money to the person I'm not really it's not really doing anything about trusting the person but in this case I don't think it's it's that much different it might be different in degree but it doesn't seem to be that different in nature I don't I'm not sure but that's sort of you know what it brings to me and lastly yeah we as humans have a bias for things that simplify the world and enable us to make sense of it so I think we need you know as a group to stay skeptical about them and try and see it when it happens maybe you know think a bit more smartly about them and I think that's about it yeah and so if you have any questions or if you want to discuss this I'm happy to try not so I don't know if somebody can bring you a microphone or if you want to shout in general and revolving around trust what are applications that they get too excited where you think that this technology is actually useful okay so the question was like I'm doubtful about like commoditizing trust but are there other applications on the blockchain that I find useful is that fair yeah I I don't think there's an opposition between the fact I mean I don't see I don't see I don't see them as opposed I have have concerns about the about the blockchain community thinking that you can simplify trust down to smart contracts and going straight ahead with that and sort of not realizing how the real world around them works but I don't think that says anything negative about the usefulness of blockchains in general I think it's just something to keep in mind I think there's lots of stuff that can be useful on a blockchain I think basically if there are institutions that are available and have been shown to fail and if you can replace them with a set of people who are running nodes for the for the right reasons then there's something very powerful there I think that's the most interesting cases and I think also just for like efficiency of shipping and stuff like that supply chain I think that's pretty boring you know it's not exciting to people here but it's it's it's also probably a very good news case I think when you can simplify like instead of having 15 people all talked to one another in different ways have all of them talk to one thing which is a blockchain I think that makes sense Haim a really nice talk I was wondering what you thought of I mean I personally believe definitely we're still gonna see attempts and approaches myself working on like a reputation trust protocol so what do you see the role of Technology in not in spite but in line of what you said we still are going to try to obviously like like decentralized credit networks there is some value to automating or at least creating a platform to establish trust you know with marketplaces and so on what do you see our role is and how we can guide that it's a good question because it's hard to answer I don't really know how we can really do it I think what's hard is that here too I'm not sure about is that I'm not sure we can technically answer some of the questions I I think some of them we'll just have to accept that there's possibly a limit to how much code we can write to fix stuff and that some stuff just has to happen out of it and so like one of the questions have the override like who has the right to override I there's lots of people I don't want to give the right to override I'm sure that yeah so how do we make sure that we make that you know we create a system where there is an override but it can't be used to easily but it can't it's not too hard to use at the same time and it uses enough technical elements to make sure that it's secure but not so many to make sure that the day everything goes bad you know we can't use it I think that's that's really hard so I think we it's probably a point in the in the blockchain community where people need to start looking a lot more outside rather looking what they're building locally because it's easy to get really excited and what you're building and you know be with all your friends in your community and think this is awesome we're building something awesome especially when you're building something really interesting and exciting but it's also a really good time I think for the community to look at how people have been doing this sort of stuff before and outside outside the community but also like throughout time and history you know the different ways of people how people have been lending money to each other in you know decentralized local federated systems of you know money yesterday somebody was mentioning the sousou which is one of those systems if I understand correctly you know how Trust reputation trust trust reputation the system's worked in the past like and and also looking at the limits and criticizing them because I think there probably were big you know large limits and how those systems used to be abused and can we use some of our clever modern newfangled technology to avoid those issues while yet you know using what was good I think I think it's a good time you know it's a bit like I find it's a bit like when you just the peril is gonna be weird I'm warning you it's a bit like when you discover a new shawl and music and you get really into it and you start looking you know likely listening only to that and after a while you sort of your head comes out of when you're like this all this other interesting stuff around maybe now's the time to start bringing all of this back in together I think it's bit of that like you know blockchains have existed for what 10 years dish let's say that you 3m has been around for five years it's a good time now to say okay we've done all this maybe now let's get our head out out of what we're doing and look around and see how we can patch to other stuff and I mean this is a really this more personal opinion but I think there's like quite a rift between the open-source community in general and the blockchain community and specifically and even though pretty much all the blockchain stuff is open source which is also one of the reasons what my red hat's interested like and almost it's from what I see like anything that is an open source of blockchain is suspect and almost eliminated like by default there is not that much collaboration between those two general communities it's still this little bit of a rift and so I think that for instance is a place where more people could collaborate it's not it's not completely different communities but it's surprising that there is still that much difference in a way lots of people are still skeptical of block chains and your overall you know free and open source software communities I hope that answers okay well okay well thanks a lot everybody thanks for coming and have a nice conference [Applause]