Surprise AMA 03/12/2019

00:00 to 1:23:35

let's see if this is working first time I've ever done this on my iPad and let's see if I can see the chat live okay all messages all right send me some messages guys let's see if I can see them there we go yeah I can see them all right like this a little bit better for you guys orientation is locked okay okay all right apologies this is the first time I've ever done a lot of live video on my my iPad yeah there we go all right can everybody hear me all right can everybody see me all right okay good so I just came in from Australia I'm currently in Hong Kong although I got something cool from Australia so got a little Joey here and I haven't named the kangaroo yet but you know you always have to pick up one of these when you come from Australia they lost my luggage so you might see me in the same clothing for a little while although I'm in a city of tailors so I might break down a bison or clothing anyway I just wanted to do an AMA I first to fill you guys in on a lot of the things that have happened over the last month all good news basically cardano's mostly on schedule the rest of all was a big success and there's very very strong possibility like greater than 90% chance that we'll be shipping and the Shelly test net this month it'll be a rust command-line interface and I needed to talk to Jeremy Roni about moving the state pool task force over to the rust test net so look for that but there should be some messages and things coming out very soon vincit and his team did a very good job and incredibly admirable job and they really came together if they work for an entire month straight basically seven days a week with very little distractions in a small room you know Vincent was a tough taskmaster but we got an enormous amount of work done in a very short period of time so we're quite happy with the end result there so the other thing is 1.5 is also unscheduled we just cut released candidate five and we ship 1.5 to the test that so we're so we're doing pretty good with respect to getting 1.5 to the main net so that should ship this month as well so 1.5 is the first Shelley related release and it's the last release that we're gonna do that involves the legacy code base that was written by Sarah Kell so after 1.5 the next releases are going to be basically to swap out the back end and core code so that it's a hundred percent new code so what's going to end up happening is there's going to be a merger of the rust the Haskell wallet back end and Daedalus the Haskell wallet back end team is rapidly prototyping the connection points between the rust code and that's being done by Vincent Han Casas guys and that's being done by the Matias Ben courts people so both of those will work quite well together within the next few weeks or so and that we can begin running experimentation so anyway after March we're going to be in a really good position to to start swapping out all the old legacy code and hopefully by the summit we should be able to show a lot there and we'll start getting a huge amount of data from the Shelly test net about state pool registration delegation and once we wire the Haskell wallet back end we should be able to Wally a wire on Daedalus to that and then you guys will be able to play around with the state pool center and all these other things so it's all moving the right direction ledger supports coming soon I was told by Niko's Nico that the soonest it could come with within the next one to two weeks more realistically it probably will ship next month because it depends on ledger actually updating some things on their end but the software is basically done they've been testing things like crazy but things look really good so ledger support coming either this month or early next month test net coming this month one point five most likely coming this month as well it's a very active month a lot to do and the summit is the next month so we'll have lots of cool workshops and announcements and a whole bunch of things to discuss and hopefully a lot of people there too to see you know basically we're at okay so that's the that's the update for me but now let's go ahead and take a look at the live chat and let's see and I'm a little jet-lagged and a little tired but I wanted to get this out of the way okay with all due respect Daedalus sucks slow is dripping molasses good for you alright can you talk about the 10 watt full node and how that would work and where we can buy it so basically it's open source hardware so think Rock PI or Raspberry Pi it's real cheap they're usually under $100 and they do require a little bit of configuration and work to get them running correctly so there's a community member named Marcus who's volunteered to take upon that task and he'll be conducting a workshop at the I which case summit explaining how to actually configure these things and what we'll probably do at the back end of that is have a tutorial online and a reference model so basically kind of like a preassembled kit that you can buy or show you which components to buy and how to put them together for a reference table so you have to come to the summit if you actually want a hands-on tutorial on how to assemble them and put them together and surely after the summit we'll probably have some sort of online tutorial okay can you say something about these employment in Ethiopia okay so that's a good question so I'm flying to Ethiopia not immediately after Hong Kong I'm going to Athens first for a private summit with our engineers and scientists for a related project then after FM's I'm flying to Ethiopia and there I'm meeting with several ministries were graduating the class of 23 developers and I'll have a chance to meet them in person and sit down with them and talk about their future with AI which K or other entities and then we're also going to be chatting about phase 2 of our Ethiopia strategy which is actually preparing a pilot related to something in their cultural industry preferably a supply chain product in the coffee industry so there's going to be lots of ministry meetings a little bit of politics a lot of handshaking and pictures and then we'll roll up our sleeves and get to the work of planning phase 2 so we have our business analyst Tom Flynn flying out and several people from Cardinal Enterprise crystals the head of Croton Enterprise and Carlos and a few others who are developers on Curto Enterprise are coming as well so we can begin the process of collecting some business and technical requirements so there's a lot to do if it's a coffee pilot we're gonna have to survey basically an ideal location for a pilot to be conducted and built a consortium to run that project and that will be done with local partners as well as the girls that we trained so I'll have an announcement probably in Ethiopia or shortly thereafter and we'll write a blog post and we'll do a bunch of stuff and we'll probably sign some sort of MoU either there or after sometime after but soon how far along is card out of regarding the solution for interoperability actually we're quite far along because interoperability is more of a pragmatic approach rather than it than a strictly scientific approach so what we've been able to accomplish in terms of the science side of interoperability has been the construction of two classes of protocols one class of protocols based on NEPA pouce and that's basically interoperability with proof-of-work side chains so we have a whole family of papers that Genesis has been developing that basically technique of pals from theory to practicality so we're going to go from static the variable difficulty and we're also going to considerably reduce proof sizes so that they're even smaller than the fly client proof so we have a lot of stuff to publish on that we have a lot of papers we haven't pushed out publicly yet because we're still cleaning up some proofs and doing some other minor things here and there but we're quite happy with the direction Uthoff pals have taken and we think that they're almost in a production ready design meaning that they're ready for implementation on the proof of stake side of things the good news is we have a pretty good construction for that it's based on a different primitive called the TMS and after Shelley is released in parallel with gogans development we will begin integrating that directly into Cardinal eight because it gives us access to side chains for Cardinal so for example like if you wanted to have EVM or awasum compatibility it'll also give us the ability to have very very efficient like clients in fact I think what more efficient than the ones that are coming out of the proof-of-work side so that's something this year exactly when that gets integrated is contingent on when Shelly ships and also what fundamental ledger requirements are are needed to integrate that but the cryptographic primitives aren't too involved we also want to inform of quantum resistance this year for card ah know what that probably will end up looking like is something like a closing checkpoint that's done at the end of each epoch and that would be probably a good trade-off the problem with a lot of post quantum schemes is that they are are not very space efficient so if you do it per block it would add a lot of bloat and the return for that is not so significant but an epic closing procedure would probably be more reasonable and also allow parallel downloading of epochs so right around the time we start putting a mechanism like that and it's probably one will put the sidechains mechanism in but anyway it's leaving the lab and it's entering engineering and you know we're bottle necked on shelley's so we're waiting for that to come out the new code base to come out and then we can start rapidly deploying features every four to six weeks and we'll just build our way up there okay oh and the other two sides of interoperability is pragmatism of accepting different protocols so lightning and inter legend of the two that we're currently working with and we're probably going to talk to the a young guy so I have a meeting tomorrow with the CEO hang on and we'll see what they have there's also feel protocol as well so we'll have we've been having discussions with them so it's just a matter of deciding what family of interoperability protocols to support and we'll just see if that's sufficient for the market okay what else do we have here everybody's asking about the Samsung wallet are you guys chatting with them CAD crypto it's not clear how to get on to Samsung at the moment they haven't announced anything about how that works and it seems to be in the family will definitely have conversations through our network to find a way to get into that ecosystem I'm a Samsung user myself and the s10 is a great piece of hardware and Knox is a really good ecosystem so I think that some point they'll probably have an open API or some sort of path to allow people to get on but at the moment it looks like it's just a theory of support so we'll see where that goes that said the demo that I did was for a tangent car which is an NFC device and so basically it's like a super super low-cost Hardware wallet they cost less than a dollar manufacturer and private keys are held and managed on the device and basically what you did is just happen with your phone it automatically opens up an NFC wallet and then you can spend from that and just enter a pin code to initiate that so it's super easy user experience and our hope is with the relationship with tan gem that we can eventually build more capable wallets at the same price point that we can use for offline off chain transactions the goal there would be to couple that with our Africa work where people can basically have these as super low-cost cash cards they can be loaded have cash register style accounting and they can go and just tap them with each other and eventually they reconcile online when the internet is available to reload and so forth so that that's a long-term play it's something that we'll be building over several years and basically what we're doing is starting with just getting a to support into the tangi of ecosystem and then we'll work with tan gem moving forward to see where we can go with hardware but anyway I'm pretty excited about where that's going I love the price point it's a hundred times cheaper than a ledger you know or Kiki or any of these other things so it'll be pretty easy for people to acquire them and they still offer a reasonable degree of security because the takes a while for the NFC to actually open the app and they still a pinko to spend from the card and it is a trusted Hardware module so it's basically a TPM Charles can you please comment on the Ergo platform ergo now that's a very odd question but ergo is pretty interesting project so that's an idea that came from anion which K research fellow named Alex tripping Oy who actually predates I which K you been in the space for almost as long as I have if not as longer I he was a quarter belapur for NXT and he started something called the score X project that I which case started funding back in 2015 he was one of our earliest people in the company and he did a lot of research with us so Alex took ideas that he learned from ago and papers that he wrote with Leo raisin and his studies of Sigma protocols and he's putting it together into a new cryptocurrency framework called Argo and basically it's kind of the best of all of his research so we like Alex Allah he's a good friend of the company and I'm not sure when and how they're going to launch but what's exciting is that he's able to take some really good research in and put it out the open for kind of a next-generation cryptocurrency so I'm as in much the dark as you guys are about what exactly that's going to do but it'll be quite fun to to see where that goes Sigma protocols are really interesting flight client paper I mentioned this a little earlier but I will say something a particular about fly client reading through the paper it's it's not clear to me that they've proven everything that they claim that they've proven particularly the the dynamic difficulty requirement and the sickness requirement I have to read the proofs again but they don't seem to cover everything that they claim second they were a bit unfair in their comparison with NEPA files because we have a whole family of papers coming soon that will cover many of their criticisms if you look at the work that our scientists tend to do they start with the simple case and that we gradually move our way to more complicated cases this is a very sensible way of constructing proofs and a very sensible way of writing papers because if you try to solve everything in one paper first you're not going to and second you're probably going to have a big mess and it's very complicated so it's good to have composition and it's good dance start as a series of benchmarks and milestones and move the whole ecosystem along but for example with or Boris we started with the synchronous case and then we moved to the semi synchronous case so anyway we have more papers coming and a blog post is coming specifically about fly client and Nipah house and we have in our view proofs that are even smaller than the than the NEPA path then the fly client proofs so I think it's going to be a nice comparison but anyway it's it's good to see a lot of innovation in progress these merkel mountains are interesting constructs good good job Peter taught for having some of your work end up in a good paper and the team that created flight clients quite good and they've been around for a while in fact I think they were presenting their work for fly client back at scaling Bitcoin in 2017 so they've been at this for a bit and it's nice to see that they've gotten some developer adoption I think grynn supports them and I hope that they can get their ideas integrated in other cryptocurrencies after they've gone through peer review and there's been a bit of time for the open source community to get used to them but me papayas are still pretty cool and they're gonna get a whole lot better very quickly what is the scope of your involvement in Indore so Endor is a predictive analytics venture and they're building some really interesting products and what's nice is they actually have an existing marketplace a real product in a real use case for their token and basically people use the token to query against the protocol to ask questions so for example if you want to ask trading questions and get answers back they have a partnership with eToro and people are already using this to do things so this scope of my involvement is to help them develop out some of their product lines then I leverage my network to help them accelerate existing product lines and also see where we can find some synergies between the things I do the things they do I'm an advisor it means it's not a significant time commitment and I every year try to be an advisor for at least one or two different projects a just so I don't become siloed and isolated and be because a lot of cases of these projects have a lot of win wins for our ecosystem in their ecosystem and it's very common for CEOs to sit on other company's boards or to serve as advisors to other companies and regardless of what level you're at the fortune 500 level all the way down to the small to mid-sized enterprises are with VCS this kind of cross-pollination is is quite common and it's nice to be around that they're good guys let's see what else we got here you know this is an example of of what happens when people just don't pay attention or media perverts things so Chuck says you said Cardinal would be the first trillion dollar cryptocurrency I never said that you can go back for the tape you can listen to me verbatim and I never said it will be a trillion dollar cryptocurrency I said card ah no I'd like it to be a trillion dollar cryptocurrency and then I went on a five-minute discussion about what I actually mean by that you know my goal is for Cardno to be a self-sustaining ecosystem like the u.s. dollar is or the euro is if you're paid in dollars you don't say boy I want to buy a sandwich so let me go look at the exchange rate and and you know change my dollars over to a euro so I can buy a pizza or buy a taco no one ever says that they just go and buy it it's a self-sustained ecosystem where you can exchange your units within it for products and services and you have a circular economy and yet somehow this got turned into Charles said it's gonna be a trillion dollar cryptocurrency and somehow it's going to be a trillion dollar cryptocurrency within a year grow the [ __ ] up people and learn to listen a tiff with Chieko crypto you know let's talk about let's talk about Chico crypto just for a moment he's basically a shock jock TMZ style journalist in the space what he does is he takes rumors speculation and just falsehoods and tailors them together into video to get a lot of views creates controversy and that tries to provoke people that he criticizes into coming onto his show so that he can get more views I come on to a show he gets a lot of views he makes money he gets more brand he gets more reputation so so let me get this straight why would I do that why would I reward someone who lies and who's scum you know fuck'em I'm not gonna go on a show you know as far as I'm concerned you know he's gonna be in the ash heap in the waist Bend of history so let go don't validate people who lie mother is quite good mother had a knee replacement surgery two days ago and she's recovering unfortunately I'm in Hong Kong and they are in Wyoming so I just have to call and hope for the best it's the reality of being a CEO on the road you miss a lot of stuff my grandmother also died in February and I couldn't attend anything because I was traveling but I guess it's all a vacation reddit hi Charles I saw some fun about card on oh and bliss not being quite a resistant we're not using bliss it's an old part of the road map get with the get with the promotion guys we're probably going to end up using hash face crypto for that how is a simple transaction different from running a smart contract so the the goal of Plutus is to actually turn all transactions into a smart contract and some some way so basically if you look at the difference between what aetherium does and a UT exosystem like hard on o does their different data structures so one is kind of a mutable ledger and you send messages to it to wake it up and then you're gonna have all the state management that you have to contend with and it's really difficult especially as the system grows and especially as you want to shard the system to keep track of everything so you have to add a lot of complexity into your model when you look at aut Excel system it's what's called a data flow graph and basically it's just saying alright you have all these threads and you know you have outputs and all you have to do to weight one of them up is just claim that you you have the right to spend it and then if you have the proof that it validates its bends so what we've done is we've taken that model and we've extended it to include some state information and include the value and include some data and we call this the extended UT EXO model so you keep the same semantics of Redeemer validator that Bitcoin introduced over ten years ago but now you've added a capability of having enough information in it that you actually can have code running this is what Plutus is basically all about now the advantage of this type of a system is that now all transactions can be either as simple as push value from Alice to Bob or they can be as complicated as a smart contract that you would see like creating a currency or something like that so there's a whole spectrum of complexity there and you can chain these things together and you know you can also more easily interact with that on chain transaction with off chain code because when you actually look at these smart contracts generally what they are they're wrapped in complexity from things living outside of the system so you'll have some JavaScript code and some web 3d action going on and that's running on a server or a client and then that's talking to something living on chain that's the etherium model with Carano it'll just be glued as template Haskell Haskell and then that runs and it runs as a single unit and you can write all the code together and it's very easy to validate that the off chain and on chain code is actually working correctly together and we're actually working with GS CGS and with webassembly and you'll be able to port that code off chain code to run on anywhere javascript can run so can run as a node package or you can run it the browser via web assembly so we have a nice framework there and you can do a lot of property based testing and you know a lot of cool validations and say things are working correctly because it's running this data flow model that's immutable what's really nice about is easy much easier to shard these types of transactions so basically the answer your question succinctly is that all transactions are technically smart contracts in this Redeemer validator model it's just their complexity is up to the user and how they run these types of things there are a lot of really interesting use cases that do require further extensions especially if we're thinking about things like puffs up and Redeemer VAT and contingent settlement so recently it was just announced in Bitcoin magazine that Texas is considering passing a law that if someone is to receive a cryptocurrency they have to know the real world identity of the person sending it to them first which is a pretty crazy notion if you're taking a look at something like Bitcoin the whole point of Bitcoin is people can just push value to you and you don't know who they are and you can't stop them these are push transactions but somehow Texas is under the opinion that we cryptocurrency users have some sort of control over people who push value to us we don't but if you live in a world of contingent settlement you would have the ability to control that because what you would do is you would say hey what we can do we can basically say I have to consent for this transaction to sell so Alice would push the Bob it would sit in pending and then bob has to sign something for the transaction to sell now you could then structure these transactions in a way that Alice has to transmit to Bob probably through a subscription over a private channel some identity information and if that matches Bob's require 4ky CML court in the state of texas if they passed this law then Bob would sign you could even automate the system potentially if there's attestation of the identity document it's an example of how you can stay in compliance with the law by a gradual escalation of the protocol so what you do is you'd say okay we have basic transactions and then we have contingent settlement and then within the address structure itself you might even be able to embed that requirement so if you're receiving payment in Texas if they passed this law you'd be able to stay in compliance if you're seeing payment in the other 49 states you'd be able to use a regular address structure and it's up the user beside that and you can sort these things out with Politis contracts and so forth so so it's it's really all about the capabilities of the system and we designed Cardno to try to be as layered as possible so that we can keep everybody happy and also allow you to keep your principles for the most part and no it's not really an enforceable law hey Charles were all of those release dates you pulled out of your ass what do you think Brian what do you think do you think we just made stuff up you think we just got a hat and we said hey hey you're here July sounds great June sounds great that's not high engineering work so we got our people together we have project managers and we had charters and a risk register and we had a lot of Engineers and we said what can we do and what can we do it and let's estimate this yeah processes behind it they worked real hard and then we get into a bunch of bottlenecks we had to really realize that the code that we were working with had too much technical debt which meant that a lot of it had to be rewritten okay that takes time second we realized that we had a lot of people in the ecosystem that the foundation was supposed to take care of but they didn't under Parsons leadership that we had to take care of so we had to build a help desk that they were supposed to build but we built it anyway because well someone had to do that we had to take care of all the exchanges and their maintenance and support we didn't have any personnel on staff for that because that was the Foundation's job but they didn't do that also had to build out a community management core delayed us slowed us down a little bit furthermore we at science delays delegation and state pools and or boards it took a little longer than we anticipated to go from the lab to something that we could write a specification for so April of last year turned into summer turned into fall turned into q1 and that's how software engineering works when you're doing science now all along the way we inform people all along the way we had a mas all along the way we tried to make best effort to show people where we're at we're open source project we were never one for github commits when it became clear that a lot of people were having issues with Dedalus we didn't just rest on our laurels we went ahead and created icarus which turned it to your eye and now 15 percent of the network runs on your ROI we ship five major updates to cardenal with 1.5 it's just about out five major updates 1.1 1.2 1.3 1.4 and we dealt with a lot of ups and downs we went reach Ange doll the api's from v-0 to v1 much better API is easier to integrate against and much easier to use for enterprise clients so huge amount of work was done you know I'm not happy that things took a little bit longer than we thought but all things considered to go from nothing to launch to something and in just a year have a very clear crystallized vision and massive progress behind that vision forty papers more than 20 peer-reviewed that's an amazing progress but what's your standard well honestly what is your standard let's say we did ship everything and we had smart contracts and decentralisation if your opinion is that the price would somehow go up maybe a little but we're in a recession I could cure cancer I could sail on a air balloon around the whole world in 40 days go to Mars won't change anything price will still be in the toilet because that's where the markets are at right now so so we're doing the best we can with what we have and the lessons we've learned in the science that we're dealing with and I really am proud of the people who have put in hard work and I'm proud of the community that's been patient and I'm proud of the things that we've learned along the way our estimates continue to get more accurate and they continue to get easier to follow we said that or horse BFT shipping this quarter and shelli test net shipping this quarters is an incredibly strong possibility that that's gonna happen we couldn't do that a year ago we obviously weren't so good at estimates so that's a huge improvement and I'm proud of that and things are going from maybe we can get it done in six months - we have a clear path and understanding how to get that done in a reasonable timeframe there's still a lot of process as the processes that need to be changed there's still a lot of work that needs to be done but we're always moving in the right direction week by week month by month and there's a lot of really good people at i/o HK who wake up every single day in some cases on the weekends - and they put in more than 40 hours and they write some of the best code the space has seen and we keep our principles with that code hey Charles enigma what do you think of the enigma project and their secret smart contracts so enigma is an MPC project out of MIT the kids come out of Media Lab they're actually I have a transitive connection to them because I think Alex Pentland is an advisor for them but Alex Pentland is also an adviser for Endor and the head of Endor oh I believe was a graduate student under or postdoc under Alex so they have an academic connection there so basically they're running private MPC based smart contracts and I think they're accelerating them through SGX fundamentally sound and it's a pretty good idea to do these things this way we have a whole bunch of MPC research we do mostly out of Tokyo Tech at the moment but we're extending the team and bringing it into our European core Vasily zika's for example co-op writes a lot of MPC protocols in using MPC for smart contracts that are off chain does make a lot of sense so at first it depends on the nature of the contract it depends on if it's interactive or not interactive it depends on whether you want to store and save that data or throw it away after the game ends so there's a laundry list of things to think about and there's a lot of competition down in smart contracts that are private for example there was a Z cash related paper that came out from Matt green called de XE it's about 64 pages it's a big piece of meat but long and short is it accomplishes roughly the same thing it's a slightly different model so so it's an interesting space and it's it's nice to see some MIT kids working real hard on getting something out hello could you comment about other Iowa Cape projects so score X is Alex project and he's he's kind of frozen that to work on her Co RS coin is retired ours comment was something that we constructed to learn more about cryptocurrencies but there's not a lot of commercial use in in ours coin so we we're not doing much there with respect to key adidas we are actually actively working on that Polina who's right now the one of the professor's teaching the Haskell course down in Ethiopia will likely be rolled over to be a technical lead to investigate q adidas and see what we can do within what I would like to do is take some of the fundamental concepts that Bill White wrote in Q and etus and combine that with homotopy type theory and see if we can pull them together with a completely new foundation so there's a quite a bit of work to do there to explore that space but basically the long is short is that Q and E just seeks to be a marketplace for deduction so the concept is that people have something that needs to be proven they offer a bounty for it and then basically there's mechanisms within Q adidas for it to be proven now my hope is that we be able to liberate the entire field of mathematics at some point from academia and put it into more of a an independent of university setting where independent mathematicians can make their living proving things on a blockchain rather than just writing academic papers and being subsidized by university salaries so it's a very different model but it's it's an interesting model and it does require a lot of very heavy proof machinery and it does require a heck of a lot of really complicated concepts so that's a stretched long arc project and it's something that we're just in the prototyping and exploration phase and there's a lot of heavy decisions that have to be made about even how a person would go about that but it's an inherited project and we've been sitting on it for about two years now because we were waiting to get the right team and we're almost where we need to be to have a team for that by the way about jargon probably 99.997% of the world is unfamiliar with what moto p-type theory so don't worry about it hey Charles what's the strategy for teaching 15,000 Plutus developers this year I thought it was 10,000 but anyway it's a large number the goal of treat and teaching a lot of Plutus developers was basically saying we're going to start with teaching Plutus to a small cohort we're going to learn a huge amount from that process and then we're going to take those lessons and convert them into a MOOC and then what we're going to try to do is get a high enrollment with that MOOC and then people who pass it probably get some secondary training to them so that we can get close to that number and we're taking kind of a multi prong approach so we're going to continue the in-person training we're going to start working with channel partners to distribute content the Clio guys will probably get some money to or grant to build up some new content we may also use udemy as a platform I did that with the Bitcoin class that I taught with about 80,000 students and then mergo has its commercial training that it's doing down in India and I think the class size is going to be around 3,500 that they anticipate so our plan is to do this over the next 18 months so between you know this entire year and next year I and the hope is exiting 2020 we should be over 10,000 that's the hope but that depends on basically how fast the MOOC content I can get out into the public when Pluto's playground and Plutus is really ready for mainstream adoption and mock chain is able to fully emulate everything that we want and then also there's going to be a lot of kinks to get out with like IDE support and tooling and so forth working with university of edinburgh tokyo institute of technology and university of athens we also have commercial relationships in some capacity with Cambridge Oxford Lancaster Kent University we had one with Rensselaer Polytechnic Institute we also subsidized research out of University of Illinois urbana-champaign and the University of Connecticut what's your take when it comes to scientific research on our chains rolling when they read a set of formal semantics and they have a working system will comment on it there are a lot of smart contract programming languages that that are on market and you know in my view it's less about the language and it's more about the ecosystem around the language and also it's more about how you intend on capturing business intent and delivering a reasonable and an experience in modeling that intent and delivering that intent to your customer so so basically you have a situation where you say okay I want to do X and I want to do X for person Y and in situation Z so you have a domain and you have a customer and you have users so actors and events and value that you're having a discussion around so before you write any code at all you have to have some sort of coherent way of putting that down and capturing those requirements so that you actually understand what you're trying to implement because the problem is that when you write the code you're writing code for a computer you're writing code for a protocol for a system and that's that's not human so there's no guarantee that what you've actually written will actually match what you intend you know you for example multi-sig so when parody wrote the multi-sig contract they had an intent of how multi-sig should work we all kind of knew that but then the code didn't actually reflect that there was a little bit of a drift there so the first question you ask when you write these things how formal and how rigorous do you want to be with the modeling do you just want to have an aural understanding do you want to write something down do you want to write something down in a language that machines can understand so that when you implement it you can verify the implementation matches your understanding and as you work your way up that stack it requires more and more effort and the tools that you use are a bit more exotic so for example you could just write a business requirements document that's like level one then level two you can model that document maybe with a little bit of a technical specification and and some specific domain requirements in a slightly more formal language so one's probably just prose it's English and then the next one actually may have some formulas and some math and some you know those legalese and things like that then you can take it up a notch and write it in a modeling language like TL a plus for example which you can do model checking with and computers can understand and now all of a sudden this is something that's a lot more mathy now to go from English which anybody can write to kind of a domain-specific document it's probably written by an engineer or lawyers or other people to TLA which is written by formal methods expert each it's like a pyramid you kind of lose the amount of available people and also you get diminishing returns so there's a big discussion right now in the smart contract world about well what is good enough for a modeling intent and understanding intent of what you actually want to do and this is independent of your smart contract system because once you've written that all down then you have a question of well where do I want to run that and then what is the best way to implement this and then these are questions of privacy security cost portability future proofing that's a completely different set of questions so there's a whole bunch of smart contract programming languages some are domain-specific like Gamal or Marlo some are general purpose like Plutus and solidity rolling and others are are you know kind of live in between and they do things for you and some are easy to port to other platforms like for example you might be able to write your smart contract and rust and then you can port it between different webassembly style platforms where others are very tightly locked to the platform for example solidity is right now very tightly locked to aetherium and it's difficult to port it to a different model so so these are the two sides of smart contract language development one is the modeling of intent and then the other is the implementation of intent and then there's this meta No did we prove that your code matches your intent so you have to solve all three of these problems okay let's see what else we got stable coins are definitely going to have a prominent role in the future all right give me a good question here guys I think we're going to complete the entire to answer your question Center Studios about the 2020 funding I feel we're probably going to complete the entire scope with the contract within the original bounds because we're actually moving faster so we were a little slow out the gate but we'll finish strong Charles would you be willing to join a POS versus POS discussion with Kyle cemani and will Martino I'm happy to talk to will and I let the cadena guys very much I like their work a lot and I think the great engineers Kyle said I was working with the accuser and I'm a criminal and I defraud a Japanese people he's a piece of [ __ ] I don't want to be in the same room with him you know anybody who publicly accuses me of being a criminal who's a real person like Twitter trolls who cares but he's a real person they can just go [ __ ] themselves so no desire to be in the same room with him and I'll tell him that to his face let's see here cardano's already in Hawaii we actually have a lot of Hawaii fans in fact I was in Maui in Lahaina and the waitress said was a Anita holder so it's really amazing to see how far spread things are I was in Melbourne in Australia going through the airport and somebody ran into me and recognized me and we took a picture together so it's a truly global movement and it's really exciting to see how fast things have grown that's cedar come on guys give me a good question I studied severe II an additive number theory and I was interested in more efficient algorithms to decompose integers into their prime factors and then I was really interested about how prime factors changed under addition so if you knew how an integer X decomposed into a particular factorization when you added some arbitrary amount to a K what statements could you make about the new integers decomposition in relation to the old one is it just truly random or is there some order to that chaos and could you use that order to solve a whole class of interesting problems like Collatz and Goldbach and so forth so that was the mathematical research I did and it wasn't very successful so I will be under the category of failed mathematician and that's okay Louis Philippe I really like the Lightning Network and will probably integrate something with Cardno there it's an interesting question how do you work with formal methods and that's a whole big rabbit hole so you can go everywhere from model checking to refinements to property based testing to by simulation so it's a game of diminishing returns so the very first thing you do when you're talking about formal methods is you start and this is kind of like the the minimum viable formal method I'd say is just writing a latex specification in a formal language like mathematics to discuss your design and that is probably a lot simpler than what your implementation is going to end up looking like now the advantage of writing a latex specification is it's incredibly easy to translate mathematics into Haskell it's almost a a one-for-one matching so you can kind of eyeball the Haskell code and eyeball the latex back and you can look at them and say yeah they're probably the same thing now the disadvantage of that approach is that you you can't get a formal proof that they're exactly the same because one is just late ekend the other is actual code but you know there's a visual proof where you can look at it and say well that looks pretty good now the output behaviors of that Haskell code is basically a reference specification and then what you can do is you can take that code and go through a series of refinements until you end up with the production code and then what you can do is kind of like a poor-man's by simulation on both sides so you can say oh okay this is what it ought to be and this is what I'm getting and they match okay they look right and then you can also use it as the basis for writing using quick check tests and these types of things so it's that's that's the kind of formal methods we're using at the moment what we're probably going to do is at some point upgrade that to you know kind of level to where we'd write these things in a formal specification language like PLA but at the moment were you just using latex deriving some Haskell code as a reference base from that and then doing serious refinements from this and then having some sort of way of matching them along the way and conducting property based testing and to be fair this is probably going to to capture I'd say 80 to 90% of all the bugs that formal methods would is going to capture where you look for formal proofs or when you're dealing with systems like flight control software or things where when they fail people die like pacemakers or things like that and you know you have a long time horizon and the devices don't tend to change when you ship a plane the planes in service for a long time unless it's a 737 max and they fall out of the sky but I'm sure they'll resolve that but anyway levity aside planes are planes are in the sky for a long time and medical hardware you know if you put a pacemaker in or some biomedical device probably going to be in the patient for decades hopefully so given that you have a device that once design it doesn't change much it is worth the investment of spending one or two years using formal methods to at the highest possible assurance level to verify that the code is correct the problem the Cardinal protocol is were still in an active state of research so there's a lot of flux and there's there's a lot of little changes and subtle things that need to be done which under normal development circumstances can be done rather quickly but if you're engaging with like maximum level formal methods this would really really slow down the the development process so the approach that we've taken where we dual a tech which can be very rapidly updated and changed some cases same day and then have derived Haskell code from that that's really easy to write because it's almost a one-to-one correspondence and then being able to write property based tests against that to the to the new code is quite fast in fact in some cases it only takes a few weeks to go through this entire pipeline and it doesn't require a lot of heavy machinery like we don't have to use dependent types or use a theorem prover like Agda or coq or interest and as a consequence it doesn't slow us down very much but yet we still get a big chunk of the value proposition that formal methods provides as we grow as a company as our tooling gets more sophisticated and as code moves from the changing often to probably not going to change for a long time if ever then what we'll probably do is take some of that and the code that's not going to change very often and start moving that code up the assurance stack will also have some longer-term discussions about proving properties of our specifications for example the formal wallet spec polina before she went to ethiopia did spend a quite a bit of time writing some [ __ ] proofs about these things so so that's how we use it Duncan Koontz has done several videos and also Phil Conte and Jared are gonna go on the card ah no effect and talk about formal methods in detail so look for that probably gonna be like an hour long conversation what do you recommend watching on Netflix you know I'm a big fan of animated show on Netflix called Bojack horseman I don't know if you guys have ever seen that but it's a really interesting show incredibly well-written deals with depression and a lot of other things and burnt out celebrities and it's like kind of a parody of Hollywood as if everybody was an animal character strange show but really a very entertaining and also an emotional roller coaster highly recommend so in the lab the TPS of EDA has been our between 100 to 250 somewhere in that range that was when energy 32 did a quick and dirty model for or for Sprouse so I would expect the performance to be no greater than 250 TPS for the launch but we'll know more after the test that comes out we'll stress test it but you know that's great because that's single thread performance and then what you can do is you can run ethics in parallel once we have Hydra and so that we can get a lot more performance and then we can put a layer 2 protocol on top to get even more performance so I think our TPS will be quite competitive with a lot of these third generation protocols you just have to get all the foundations right and we're almost done with the Genesis research line or Bors Kronos and also dealing with spikes of dishonest majority are really the last two milestones there alongside some updates on the proofs and that at that point the foundation is is granite its rock-solid we have all the properties we want and then it's just moving on to sharding and that's where you get your TPS ray you the papers I found on the official website don't seem to be reviewed by reputable scholars but by small crypto related firms which dollars are currently reviewing work do you understand how peer review works do you understand Christ man go back to school look up crypto look up euro crypt look up nd SS look up CCS come on you jack off I'm having fun today I'm very jet-lagged and I have no filter right now or our patience for this idiocy everybody's a [ __ ] armchair expert on how academia works that's an interesting question Cheryl so you see smart contracts making deep inroads into legally binding contracts the answer is yes because there are several jurisdictions that have passed laws recognizing the legal rights of smart contracts and that's going to continue happening and then I think within the next five or ten years we're probably going to have a patchwork of jurisdictions that when wired together will give you pretty much global coverage so if you if you have a smart contract that contract will actually have rights and what is it having rights mean it means that there's some way of sorting out the owners of the digital signatures and the owners of the assets and when the contract executes that execution will be respected by a court of law so if value gets sent to Alice to Bob you know that transaction will actually be recognized by court and its admissible evidence and so forth now your mileage may vary and it's completely contextual to what the contract is actually doing and jurisdictions are probably not going to give you a complete coverage you're always under the local laws and so forth but I think legal rights are coming and if you're really interested in these things one group I'd highly recommend you fall actually two groups I recommend you follow Aaron right out of consensus and the open law project and I'd also recommend you follow coin center that's Jared grid OHS group and both of them produce a lot of good content and they're pretty smart guys actually guys I met token 2049 speaking of Tron and Justin Sun is apparently going to be at this conference I've never met him in person invited to several events that he's probably going to be at it would be really fun to get a picture with Justin Sun so I might be able to give you guys your Tron picture you've all been waiting for will have a very lovely conversation DC or marvel marvel definitely Marvel big Doctor Strange fan also big Iron Man fan in my suitcase I flew out Qantas airlines from Australia here to Hong Kong and they just decided not to put my bag on the plane so I have no bag I have the clothing on my back we'll see how long that lasts my mood is subject to deteriorate a bit if I don't get my bag back probably also gonna get pretty smelly what does the main income that I which Kate has we're doing good we make money we're a revenue driven company I love how everybody also knows how to run my company and everybody also knows how much money I make you guys are just so smart great I should just retire and have you guys run everything you actually tone Vaes wills question if you see tone fades you have to speak to him Charles he isn't a fan of Carano and love to see you both talk I've talked to tone many times I saw him in Ukranian I've been on a show and we're never gonna agree but I like tone vase I get along with them it's it's fun to sit down with them and talk with them I you know he says everything's a scam but I should be in jail in a theory it's a scam and it should be a security even when the SEC says it's not but we still have a lovely conversation and he's just one of those guys in the cryptocurrency space and he's gonna be here forever just like me and we're gonna be having conversations every six months to a year forever so I'll let you know if we run into him all right what else we got here come on guys give me some good questions when Joe Rogan podcast would be a lot of fun guys I don't think I could top the Alex Jones Jill Rogan podcast though that was that was perhaps the best podcast of all time you do dimensional aliens that you can only see on DMT truly amazing it was it was a work of art you know about laguage multipliers and scaling absolutely nothing sorry you have me beat name one other developer apart from Charles does he know any does Tom vanes know how to use github like sent him to our github link you can see like 70 70 developers and they have PhDs 24 of them have PhDs and math and computer science and physics man ah damn tone and I'm not concerned about Justin son forking the code he didn't even want a fork mantis I was actually kind of sad about that all right guys come on give me a good question to end on something for a nice sound bite alright smen roll love how you avoid certain questions what question would you like to like me to answer I'd love to I'd love to see it come on give me your best shot but you got for me come on you're so silent talking big man Oh screw you these people and Nostradamus know I'm not familiar with that person maybe I am but it doesn't ring a bell at the moment next power are you quick right dr. Jekyll and mr. Hyde situation I assure you Jesus you guys crack me up I love it well this interesting question here is I heard you say an interview wrote a spec after you finish coding not formal method if true you know they again it depends on your definition of these things it also depends on the ordering of things so it's entirely reasonable to write version one without using formal methods as a way of establishing actually how it ought to work and getting a better understanding the business domain then to revisit the design write a formal specification then refactor the code to meet that and prove that you've successfully matched that specification in fact I'd argue that's probably the better way of doing things just because you will have a more accurate specification if you end up writing the spec first and then writing the code you end up having a development process that looks a lot like waterfall and it's just really slow and the problem is the spec is probably not going to fully model or cover all of your business needs specification is only as good as what it's modeling and if you don't fully understand your business domain or you need to make a lot of changes that's that's a pretty heavy deal so it's usually a good idea to do a version one using good design principles in an agile process iterate a few times and then when you actually fully understand what you want to do then you write a spec and then you prove that you're meeting that specification and seldom will you actually meet the specification there's gonna be some Delta between the two but I think we're just having a semantics issue here that's a good question to you that the Haskell algorithm has an extensive learning curve so first does not algorithm is the programming language Minh is about well is the language too complicated and you're necessarily excluding all your developers here's the thing at the end of the day this is a very boutique platform as is aetherium as is Bitcoin and there's not going to be a lot of developers in the beginning and these platforms involve money privacy identity they involve your property and as a consequence the consequence of failure is very high when you read a web page and you screw something up with the formatting the the end result of that is your user has a broken page or the button is too high or too low and you say crap that's bad it's embarrassing it's like a spelling mistake nobody dies when privacy code fails people can die when Money code fails people lose everything so you have to take a step back and ask yourself for your first cohort of developers the people who are going to deliver the 1.0 user experiences is it a better idea to say let's have them be as broad as possible as many people as possible with the full understanding that those people probably aren't going to know what they're doing and they're gonna deliver a lot of chunk to market which means that the user experience will suffer and people will lose their money or is it a better idea to try to have your first cohort of developers be somewhat informed somewhat skilled and deliver great experiences which are secure and actually work and showcase what the platform can do in other words if you're building the Xbox would you rather launch it with 1000 crappy games or would you rather launch it with five to ten really good games now we already have the 1000 crappy games we have aetherium we have lots and lots of people writing a lot of stuff and they're having a heck of a lot of fun pushing it out and most of it doesn't work and most of it really is insecure and if you're actually trying to put it into a production system it's gonna burn like hell and so given that that market segments already taken I I don't really want to compete with that market segment at the moment what I much rather do is have 50 to 100 really really good dabs and 50 to 100 really good applications written by very experienced nice developers and by going nice you can do that because right now there's nothing for the functional programming community in the cryptocurrency space it's very distasteful to go to a hassle Oh camel close your Scala a shark developer and say I know you love that ecosystem you love tight seal of monads you love these things but ya throw all that away and go write javascript code like the whole reason that they that they've moved over to Haskell is because they don't want to do that and so what we're doing is offering a niche market for that group of people to come in and deliver great experiences now the way we've designed things after that initial cohort runs everything out it's easy for us to open it up to imperative languages and it's easy for us to open up for different development models now there's already market precedents for this if you look at the development the iPhone the initial court of iPhone applications was tightly curated you know wasn't just like everybody was just writing stuff Apple was very careful about user experience and so much so that Steve Jobs went to John doar and they set up a special fund specifically to cap a financed iPhone app so I think it was around a hundred million dollars and they went to planner Perkins for good reason because they were the best VC in the valley and still are one of the best faeces in the valley so I agree with Steve Jobs here I think that he had the right strategy for how to introduce a new platform in new capabilities to the consumer you have to hold their hand a little bit and that necessarily means that you're gonna put a higher burden on the developer for version 1 but that translates to a better user experience more security and overall more functionality then over time you can open up the ecosystem and make the ecosystem far more accessible and then you get a next wave of developers in and you actually have the best of all world's because you still have these old guard cohort that are very experienced they know what they're doing and they've shipped things and they become mentors advisors Stack Exchange question answers for the new people who are coming in because otherwise it's just going to be a free-for-all and it's gonna be a mess so that's why we chose Pascal and Plutus and this approach and we invested in some technologies that we think over a longer time horizon can bring in the rest of the groups of people but let's get 50 to 100 real apps that are doing great work and actually used and people care about and let's make sure they're written really well and they have a great user experience and then once we have those things together then we're in a position to open it up and get the thousand and then the 10,000 and a hundred thousand and so forth and if you don't like that well go go play with aetherium there is a new Explorer coming out what do I think of the Voynich manuscript it's a crazy crazy book I actually once saw the original copy and it'd be really fun to see if it's translated I think there is no translation for it I think it's one of those esoteric tomes that came from the Middle Ages that people wrote or it was a shorthand for some sort of shorthand a dialect of Latin and there is never going to be a proper translation because we don't have a rosetta stone for it if you want to see a really cool medieval document though look at the codex gigas it's a gigantic Bible supposedly written by a monk inspired by the devil okay suto is the dog from Disney after the planet all right one last question come on now give me something good my favorite video game that's a good way to end it now this is a very old old game it came out I think 1992 or 1993 by synthetic studios or something like that it's called legends of hour and actually it's the game that inspired the Bethesda guys to do Elder Scrolls Arena and that whole open world model it was the very first 3d open-world game and it takes place in this little city called middle dwarf and you can play a character II go around and you have to join all these different guilds and the goal is to basically save the exiled King and find your cousin Sven now what's really interesting is I played the game off and on as a kid for 12 years and I was always trying to win it and I could never find out a way to win the game you know it's just like well how do I win this game and eventually through the magic of eBay I came into possession of a clue book for the game and and it had a whole description of how to win every quest and everything and and it turns out that they never actually wrote an ending for the game which was which was a really horrible thing to to discover but it's a great analogy for life I suppose you know and maybe it's a good invitation to write our own ending so that was my favorite game still is always will be highly recommended you can probably play it still if you find a copy of it floating around somewhere and running on DOSBox and if you ever if you ever do let me know your experiences so thank you guys all for a wonderful day and I'm really tired and I'm gonna get some great sleep and I'll see you all u Hong Kong folks at token 2049