Smart Contracts and Beyond

hi everyone this is charles hoskinson broadcasting live from warm sunny colorado today is saturday the 4th and i'm just about to head up to laramie wyoming to catch university wyoming montana state university game and before i go up i just figured i'd make a video and talk a little bit around some of the things that are coming and a little bit around some of the way we thought about how to build a good system now the testnet's out people are actually playing around with it and people are building all kinds of things and as a consequence some of the things that are being built are being built in a very alpha way to say the least and so people are coming up with questions about access control they're coming up questions about concurrency they're coming up with questions about throughput and capabilities of the system the problem is that it's like somebody just takes their first computer science class and they learn about bubble sort and they say god why is my computer so slow well because you're using bubble sort it turns out that there's new better algorithms that can be implemented new better methodologies and techniques that can be implemented and what we've been trying to do as an ecosystem is is balance speed of delivery to market with time for the development of tooling pedagogy and community driven applications so if we had all the time in the world what would happen is you would have a very long test net period about a year and you'd have lots of people come tons of challenge problems and there would be this grand evolution and education of the underlying infrastructure now we're in a competitive environment there's tons of other protocols a lot of people are moving very quickly it doesn't make sense to do a long burnout so you can shorten that if you believe there are no safety concerns and what you can do is create something on the main net where it's clear that there's a certain class of problems that are very easy to solve and you now add lots of value to the system and there's a set of other problems that the community is going to struggle with for a bit probably about three to six months and you let them struggle with those problems until those are resolved so the purpose of this video is to and it's going to be a little bit of a whiteboard video is to kind of talk around where we started and where we're going and why cardano was built the way it was built and hopefully give you guys some more clarity on some things so let's get right to it and then i gotta go okay wow all right so the first things first i wrote a paper way back in the day i think this was 2016 which is why cardano and in this i talked a little bit about our design principles various things that we were thinking about various things we looked about and i also talked about kind of two concepts one was the idea of cardano settlement layer and cardonal control or computation layers okay so the basic idea is that no matter what computing model that we adopt in the main chain there's going to be some use case pattern or cost or performance requirement or concurrency requirement that's not satisfied by that and as a consequence there's going to be somebody who's unhappy with whatever design change design idea that we have so they'd be a lot happier if they lived in a domain that's built bespoke for them okay so the first thing is utxo for accounting versus accounts okay now the problem between these two and this is hotly debated in the space utxo bitcoin follows and there's a whole cohort of people playing around with that and it accounts as what ethereum does it's almost like comparative versus functional a functional programming language like haskell versus imperative language versus javascript there's just going to be radical differences of opinion about software design development accessibility and what's the right tool for the job so when we designed the system we said can we construct something that makes it easy to spawn off side chains that have different computing models okay and different utility for the system so for the purpose of this lecture there's really three yellow the evm and catalyst okay so catalyst is all about consent and a complex collection of voting protocols now if these broading protocols were to be developed with traditional evm or pluto style smart contracts they'd be extremely expensive and heavy so they're basically built into the chain okay and there's a lot of mpc and all kinds of stuff going on here and this is basically what we've been doing with catalyst as a standalone chain now the evm many of you watching who are ethereum developers you're familiar with this model the whole web3 concept all the different libraries you like to use all the different tools you like to use all the layer two solutions this is a vibrant ecosystem and there's a lot of growth and there's a lot of software that's already written there and you guys are pretty comfortable with that one of the big problems though is scalability here principally because you have global concerns for the model to work so basically everybody has to know each other's business the utxo world you have local concerns you really only have to know your part of that graph to be able to do the things you want to do and you preserve inclusive accountability so when you want to do things like shard and scale you want to do off-chain stuff you have a lot of non-determinism here you have absolute determinism and you have a lot of concerns of exactly how to address that deal with that this is one of the reasons why f2 is such a complex piece of software and why they're having such trouble getting these out this is not an easy problem and kudos to them for thinking about it trying to solve it however because you have global concern you have a slightly easier time as long as you're willing to accept these trade-offs with things like concurrency and things like event handling but at the end of the rainbow almost all the solutions that actually in practice deal with this because of cost and scale they have to deal with this as off chain some sort of layer two or centralized solution this is why a lot of people actually when they see defy they say actually we're talking about c5 because in practice to make these things cost effective in scale we actually need to introduce some centralized components okay now yellow is a different thing yellow is actually a register based assembly and it's based on llvm which is a very well understood model that came out i believe in 2003 and it's really built on top of the entire case stack k technology stack and this has always been something that i i've really liked and wanted to enjoy the concept here is really universality ke has always been a kind of a meta programming language so really the idea is something the jvm or the clr something that is sufficiently complex enough that we could actually have a real conversation with a straight face about eventually c support and c plus plus support and other mainstream languages and have a path and a process to basically get there hang on a second all right so let me just kill that real quick so the point of the settlement layer and the way that plutus and extended utxo were designed is we had some principles with this we said okay well there's going to be these different computing models specialized computing models for governance that are very very heavy and resource intensive and run very deep protocols that cannot be replicated with the standard smart contract model even with utxo and all the optimizations we could potentially do there or with this account based model with ethereum or anything we can do in yellow if we want to copy ethereum we need a special sandbox to do that and allow that to co-evolve with how ethereum evolves if we want to chase new concepts like yellow which is new concept to the industry and all the power that kay gives us and we want to chase universality then we're going to need we're going to need some sort of connection point for that okay so the first goal of plutus was really pushing for csl to ccl interoperability okay so basically the idea is that assets information users have a easy method of flowing between the different systems and calling between the different systems so as side chains turn on you can do your computation off the main chain you can do it inside the computation layer for that you can call it and then results can flow back and forth here this kind of is beating at a methodical pace okay so basically you have a block every so often it's a heartbeat and that's your unit of change and these unit of change they can be batched you can get much more efficient you can compress things you can use compressed representations you can do things like for example be very clever you can use nfts for access control there's a lovely blog post floating around about this okay and then in terms of concurrency you can actually distribute amongst many utxos okay and then you also can do off-chain processing like ethereum does in practice but you can't get around the fact that there's a consistent reliable semi-synchronous heartbeat that the system has this is a fundamental design principle of most blockchain systems bitcoin ethereum other things and there's all kinds of things you can do there's all kinds of things you can do to make things more efficient to be more clever about resource utilization and how you represent information and what you're actually processing on chain about how people have right of access and how access control is going to work and how distribution amongst the utxos is going to work but you still always regardless of blockchain system you're using if you choose to use a block based system for change you're going to have this methodical pace this flow that happens now you can get very clever and you can put some pre-processing into the state pools and they can do all kinds of things to handle event flow and concurrency before some change gets committed or you actually can run things in a dedicated channel and this is kind of the idea of hydra hydra says okay so we're going to have stuff that happens on chain and off chain all right well that's actually starting to get sensible so we have these things that are occurring and somewhere around here we're going to go into this kind of domain it's going to have lots of users interfacing with the domain they're going to have a lot of fun and then at some point those changes get committed as a single batch transaction on chain now you can do such a thing if you've designed your accounting model and your programming model as we did with extended utxo and plutus to work with the layer 2 solution one of the reasons why lightning is having a lot of trouble with this type of process is that the programmability of bitcoin script is extremely limited plutus is actually a general purpose programming language in a certain respect it's a dsl that looks a lot like haskell and it is turing complete now the thing about it is is that when we're doing this you build this a spiral and i'll show you where we're at so when we started this it was just good ol faction push transactions and byron just like bitcoin and then we started adding some more sophistication we added in metadata and then suddenly you can put payloads into the transaction and this enabled things the ethiopia deal and beef chain and a litany of other things like supply chain management why because you can now embed into the transaction metadata that the system doesn't care about but your application does care about so dids are the canonical metadata identity travels on the wire then you you have next step user issued assets so the cardano native asset standard and what are you guys doing you're issuing tens of thousands of nfts because that's right now the level of sophistication our native assets have and now we actually have the next step where we're introducing on chain polluted smart contracts okay and this is huge because first off it allows the interoperability bridges to occur so when side chains start materializing centralized and decentralized you can start moving assets information and users between them including ethereum side chains okay so all that stuff you guys are used to in that world it now can work okay so you have interoperability enabled okay you have things like spending policy so you can issue significantly more for tokens you can issue significantly more complex tokens not nfts which are kind of passive things now you can do things a stablecoin so you guys may have seen the djed paper the paper on an algorithmic stablecoin this is an example you can now use marlowe so that's going to be rolling out as the pab gets integrated and these things get worked and so basically now you can have very complex financial contracts but because they're written in marlow they're simple to write and they enjoy determinism so you understand your resource costs security they enjoy safety an execution and a litany of other things okay now there's open questions about what about dexes well yeah you can do dexes and then actually one of the decks is an ecosystem occamfly wrote two articles one was in july saying hey we're thinking about concurrency and they kind of talk about some of the challenges then they're at another article just came out today august oh well two weeks ago excuse me august 18th they say actually we're ready for concurrency we we figured it out you see and why how did they figure it out well there's a lot of different ways to do it and i'd probably i'm guessing that they are using some sort of off chain batching mechanism and event handler and serializer to accommodate that and then efficiently committing things once they've settled on the off chain infrastructure on the on-chain infrastructure and that's okay that's actually how plutus was designed stuff happens on chain stuff happens off chain you can write both you can think carefully about both some solutions are somewhat centralized some solutions are not centralized but guys your imagination is limiting you if you think that well we can only do one thing every block pays no you can either distribute the utxo or you can be very clever about using off chain infrastructure for batching or we can introduce more complicated concepts over time that's the point of the spiral you see here we're safe we've achieved something that's amazing smart contracts on a utxo model had bitcoin had this capability ethereum would not exist that's what we basically did we solved a 12 year old problem that satoshi couldn't solve and we brought that into the utxo model this is a safe programmability of ut excel but that doesn't mean that that particular solution is the end-all-be-all because there's going to be all kinds of patterns like oracles indexes and other things that need to be thought carefully about and there's always going to be this balance of what do you run off chain and what do you run on chain and that always comes down to concurrency operating cost and it comes down to security and trust okay these are your factors whether you're ethereum or cardano you have to think about this as an application developer oracles for example you could do completely on chain oracles and make them completely deterministic that means all that data has to be committed to the chain so for example let's say the price of eight at a usd well what if you're doing that every block you're adding an enormous amount of bloat to the system well you could do a completely off-chain system all right well that's a different trust model maybe you federate it to improve the model maybe you introduce bonds or other mechanisms or punishments to to improve the veracity of a model you see but then it's all about access to that off chain can a smart contract query and talk to it we have to build some mechanism for that so you need some sort of bridge for it okay well these are design considerations that you have when you actually put an oracle together the same for a dex where is the batching happening the batching and matching you can always choose the bubble sort and use the slowest possible method and say oh yeah the way i wrote it only one transaction can happen per block sure that's absolutely possible you also can come up with very clever off-chain ways of doing it now as we get a little further along and we add the computation layers what gets really exciting is you now have emergent interactions okay so for example you could have an evm style smart contract and that evm smile smart contract can now call and talk to a plutus contract and vice versa so maybe it's easier in your mind to do the batching and matching and these other types of things on an account based model because you just understand that and you want to do that great you could leave the value here and use this as the control contract that actually handles it with all the user interactions what's our advantage over ethereum well one of the really cool things about having a root of trust is this is the single most expensive thing to establish and a cryptocurrency proof of work does is for bitcoin and it costs billions of dollars every year and massive amounts of electricity okay but once you have that route of trust you can then use it as a guarantee and then build efficient light infrastructure above it okay so i'll give you an example right now ethereum uses proof of work to establish the root of trust they're trying to move to a proof of stake system because they know once they have that they can shard and have different computation domains starting to look a cardano style architecture now isn't it wait a minute here we have oroborus and right now we're running prous this year we'll likely upgrade to genesis and then we'll do a little bit of optimization to that but once you have that as our root of trust why can't we run obft a significantly faster protocol that presumes the trusted quorum and use the spos as the quorum well wait a minute here now we're running a bft protocol these things have been shown to operate in the thousand tps or better range and there's papers like rapid chain that talk about how to shard obft plus you can run them concurrently right because you can use the same spo set to run many different chains sorry my pen is acting up today so you could run copy one copy two and load balance and so forth well that's actually quite easy to do and we have something called mamba that we've been working on specifically for ethereum support and ela support and we're going to talk about that at the conference on the 25th and you betcha we're definitely looking at using obft led by the spos okay so if you're running much faster and you don't have to worry about the selection of the people you have a stable quorum you can highly optimize and your trust model is actually reduced to the trust model of oroborus so you're not actually changing the underlying trust model that we already have for cardona so you by having this strong root you can do this and you can replicate many chains so we can use obft for yellow we can use obft for evm but here's where it gets really wonky you could use obft for pollutus why well cardano can be a side chain to itself so you can run a high efficiency model and you can even have different models like for example people like dags they like these block trees they like graphs they like all kinds of crazy stuff about how to handle blocks in parallel in fact we wrote a paper called parallel chains okay and there's a lot of follow-up work there there's a paper called prism which actually talks a lot about this by promote it's not to be confused with our prison paper the smallmouth and it talks about actually how to create parallelism so you actually do all of these things at the same time all right so having this heterogeneous model with stake pools as the route of trust that are always increasing means you can use them to replicate new chains new resource models it's a win-win for them because guess what they get they get more revenue streams they make more money because they're actually running multiple pieces of infrastructure and lots of parallel transactions and so forth okay so that's the point that's later in the spiral but the point is that the spiral each step of the way is secure and each step of the way doesn't cause a mass disruption and we just kind of grow on the spiral and what happens is you get faster and faster and faster so for example mamba the plan about side chains that we've been pushing that is running concurrently with hydra which is the other scalability solution the layer two scalability solution which is running concurrently with an optimization plan that we have to basically compress some more things and speed up some things and increase block sizes and so forth to improve the performance on the main chain and that's running concurrently with blue ocean research like for example concurrent state machines and what's really cool is what's going to happen is all of these solutions so there's right now eight development companies we got together and i think seven of them are going to be presenting at the conference and we asked them go build a dex okay and go build an oracle and go build a lot of different stuff they're actually going to be showing their code show code and doing tutorials and talking about their experiences the good the bad and the ugly okay so that creates a feedback loop and that feedback loop basically gives us more ideas about new features so for example you'll see in ergo they have this concept of read only utxo that's an idea and there's dozens of potential new features you could do pre-processing to actually take some of that processing that would be done with off chain infrastructure for ordering and batching and you actually put something like that into the spos okay so there's a feedback loop that now exists with the thousands of plutus pioneers and the companies that are actually building on cardano where we can add them and these two things this one and this one it's about six to 12 months so about two to four hfc events because we average an hfc event every quarter i think there's going to be three this year but we'd like to get and to have it about four so there's six to 12 months of taking it for a spin and what'll happen is you have all these people are just so dishonest and just toxic who say well what charles is saying is cardano won't be useful for six to twelve months now i didn't say that at all at this juncture right here you guys can do pretty much anything a turing complete smart contract can do we're gonna have stable coins materializing we're gonna have interoperability materializing you can do all kinds of locking mechanisms for multi-sig you can start connecting centralized off-chain infrastructure in fact dc spark is planning on doing that you can run complex financial contracts and we're actually a pretty efficient system and all of our benchmarking shows about two orders of magnitude cheaper than ethereum right now in this month those things will be solidified and absolutely there's going to be some creative decisions that have to be made with oracle's indexes and partners have already started making some of those creative decisions and those things will work their way into production this year and next year and are they super efficient of course not why because it's a new model no one has ever done extended utxo before and plutus before outside of ergo and at the scale we're doing it no one's ever done that before the point is that you want to establish a stable route of trust that's what we've done here and have enough programmability at the settlement layer where you can get people connected to the domains that they want to be and also move large amounts of value efficiently you can represent and move value efficiently okay that's why we created marlow you see cardano is going to be in our view a dominant platform for lending for insurance for bond issuance for all kinds of financial products now when you talk about rich interactive experiences involving millions of concurrent users blockchains are not built for that ethereum isn't built for that cardano's not built for that's not the point of these types of systems you need bespoke domains for those types of systems and those bespoke domains all have their trade-offs and their assumptions about how they run if you're really clever about your design and your model then what you can do is you can use the fact that you've created a very strong root of trust to instantiate those bespoke domains with different economics and different ideas and then suddenly you can utilize them but what's really nice about that is this concept of encapsulation so not every one of these bespoke domains will be successful and let's say one of them collapses or has a problem or a flaw that doesn't cascade back to the main chain so the billions of dollars of value that's on the main chain in the settlement layer is never really disruptive you say but charles what what about the fuel source and things like well guess what you can do you can say okay well the same mechanism that allows token locking the same mechanism that allows staking the same mechanism that allows voting you can use this to generate an alternative fuel with different economics on the side chain okay so the sky is the limit you can put in demurrage you can put in is that where the tokens expire you can put an infinite inflation you can do all kinds of crazy that you want on that new chain well ada will never leave the main chain but yet it generates new fuel on the side chain so let's say that that side chain has a tragic economic flaw and that thing collapses is your ada harmed no your aid is a-okay you see and that's the beauty and that's the magic of what we've constructed here you when you want to do mass acceleration of smart contracts you're going to need off chain processing like hydra you're going to need things like optimistic roll-ups and state channels these are just facts of life which is why ethereum is investing so many millions of dollars into them and thinking about these things this is not something that makes sense on the main chain you just can't build these systems in a way where they scale at billions of transactions per second on one system you have to divide and conquer you have to batch you have to pre-process you have to move things into different domains some things aren't going to live there forever and also if you just use common sense what does it mean if your average transaction size is half a kilobyte let's do some simple math here okay so every two transactions you have a kilobyte okay so let's scale it up let's go to a thousand two thousand transactions well now we're at one megabyte oh you see what's going on here let's go to two million transactions now we're at one gigabyte so if you're processing millions of tps you're actually saying that your blocks are in the gigabyte size every second where does that go in one month where does that go in one year where did that go in 10 years what is the weight of the blockchain if all of these things are immutable and never go away and they stay there forever you say oh well no that's not completely accurate because we're going to batch them and we're going to put compressed representations like hashes or use zero knowledge proofs or we're going to do off-chain processing well then guess what you're in my model you're in my model you're here you're here with the computation layers you're here with the settlement layer right you're here with hydra and what we did is we thought about this very carefully it took six years of careful thought of saying how do we pull in 35 years of knowledge from the functional programming space because we want to do things like high assurance development we want to do things like resource determinism we want to do things like local processing as opposed to having to know the global state of the system and introducing non-determinism these types of things we wanted to account for a rich metadata standard we wanted to account for native assets and the point of the 12th is now to turn on the single most important feature which is the on-chain smart contract component and then you have things the plutus application back-end which is off-chain infrastructure and you have things like off-chain processing you have canonical design patterns that come in those are built over time with the community to accommodate that finite scarce resource that heartbeat that you have just it was with ethereum and just it shall be with us the difference is we have built in the keys to the holy grail which is we have a strong root of trust which means we can instantiate many solutions and encapsulate them and if those new solutions fail they don't blow back and destroy the main chain it protects the users in a certain respect furthermore there already are people building occam's building dozens of others are building well-financed projects and over the next three to six months those developers will bring those 1.0 turn them on into cardano now we could have done all of that on the testnet had a rich long-term test net time period but that means then we're delaying stable coins we're delaying interoperability we're delaying connecting side chains on we're delaying spending policies for your tokens we're delaying your ability to do bonds and security tokens and all kinds of things by waiting for those applications to build some of this canonical infrastructure there's no safety concern for doing a hard fork combinator event pollutus is a safe thing we have a very good understanding of it and we've gotten to a point where we feel pretty good about it so we actually can move quickly here but what that does mean is it takes a little bit more time for canonical approaches to solidify and then you end up having what always happens you have abstraction where tough implementation details you no longer have to worry about you have templates okay where other people solve the problem for you and you borrow those particular solutions you have services okay there's tons of off-chain services like infera and fire blocks and all these other things that exist in the ethereum ecosystem why because they need to do this off-chain stuff they need to have a management layer for these types of things especially if you're in a light client environment okay you have standards and apis okay you have all kinds of these types of things develop design patterns and that will be built by the plutus pioneers the plutus core team and built by the application developers in the cardano ecosystem now each of these start handling problems like concurrency and access control and all kinds of things like that already there's several people using nfts for access control to create unique identifiers for the contract and unique identifiers for the the different parts of the the different users of the system and the nfts are ephemeral they can be destroyed easily created easily because of the native asset standard it's a very novel way of using that for extended utxo and of course every step of the way optimization occurs your operating cost gets reduced and the other thing is this model is super amenable to security and predictability so at the conference i think we have four vendors that are going to be talking about formal verification of smart contracts okay so for the first time in a while people are starting to take this seriously and this model is amenable to that and you have an upgrade cycle you see there'll be probably one more hfc event before the end of the year and then that's going to be adding a lot of capabilities and functionality and emergent things that we've learned and at some point next year the ccls are going to turn on and you have decentralized side chains that add in programmability from the account based model and from the yellow based model and also catalyst for bespoke computation so all of a sudden you can take really complex governance and put it there by the way you also can do a super complicated dex as a side chain and you can put in all kinds of bespoke features for that just we did for catalyst it's just a sensible model and if there's enough commercial demand enough commercial activity it makes sense to do that and you just keep living on the spiral what makes cardano fun and unique is this is an interactive process it's a participatory process by you the community and it's a friction free process we know come september 12th it's going to be a pretty easy upgrade all things considered certainly could be an issue but it's probably going to be an easy upgrade and we know the use of this system is not going to break or regress cardano in ways where you wake up and the system is suddenly unusable the existence of it creates new challenges the shipping of shelley created the demand for proxy keys and counts and partial delegation and there's a lot of work at the legend rule level that has to be done to accommodate that and so that's being rolled out but once that gets done we'll be best in class in that respect and similarly there's a lot of little things that have to be resolved with the launch of plutus that will enable more and more things the problem our industry has is they basically say well if you don't support my thing or i can't see a way to make my thing work what you have constructed is absolutely useless forever useless and it's just a toy and then they use vanity metrics not even understanding what they're saying you're only at this performance level you only do well increase the block size oh well okay they don't really understand how all these components are interconnected to each other and or operate with each other what they really don't understand is the single hardest thing to do and this is what bitcoin solved and it's why it's worth a trillion dollars is the establishment of a root of trust what cardano has done is be the largest proof-of-stake system in the world with a an efficient route of trust with a beautiful model that will make it easy to roll out computation layers this was the goal that we wrote down i actually wrote the paper but this right here why cardano back in 2016. i said we want to have this idea of a computation layer we want to design in layers this is the notion we need we need to somehow figure this out and it was a humongous technical challenge and there were a lot of moving pieces to it and what we've done is we've basically figured that out and now it's just a matter of deciding the most efficient connection protocols implementing things like mithril and we'll have an update on mithril for example that's another one of those concurrent projects we have here that will assist in this entire process and now because we have stable foundation we can work in parallel different teams can do things competing teams can do things and by the way competing teams was one of the design principles that we have a heavy use of interdisciplinary team and infosec fast iteration competing teams these were some of the design goals that we had inside inside this the the setup for cardano and we're achieving that now because of the architecture but reddit won't agree telegram won't agree twitter won't agree guys it's noise and it's fud it is one of the hardest things in the world to build a truly scalable distributed system that is decentralized and admits byzantine actors and to make that system programmable it's so hard that the bitcoin community for over 12 years has been trying to figure out how to do that in a way that they consider to be safe we as a project took an enormous amount of inspiration from the successes of bitcoin and the successes of ethereum and a lot of the ideas that have come out of the third generation and what we tried to do is put together a representative sample of the best things that we think for example most of the flaws that we see in d5 and historically ethereum smart contracts come in trying to deal with complicated financial arrangements that's why we created a dsl to make those simple and safe you need to admit that there's a world outside of your system so you need the ability to batch and match and process off chain so we built that right into the system and into the language and we're showing people how to do that with the plutus pioneers program and all these applications are coming up with that we wanted to connect to the formal methods and high assurance world so we chose a functional language as the entry point so that there'd be a significantly easier and lower cost path to do that level of verification because there's so many hacks and flaws and other such things concurrency comes naturally it comes either with off chain interaction state channels it comes with concurrent state machines and there's ways to do that they can be brought in it comes with clever load balancing of different utxos linking them together with nfts there's all kinds of design patterns that you can do and we spent the money to bring in different independent development firms to build things because those firms are going to solve things a different way and they make that code available to all of you not in six years but this month and next month and the month after and there's a stack exchange cardano stack exchange people can ask questions and if you still don't the model well then just take the other model that you guys like there you go and just live in that model and all that tooling is reusable but the difference is they run on proof of work right now and they're super expensive and we're going to run on obft right now that's kind of like what binance smart chain is doing it's a lot cheaper okay everybody gets a cow everybody gets a chicken chicken in every pot that's the way we built the damn thing okay and what we're shipping on the 12th is a massive massive upgrade because you can now have full programmability you now have spending policy you now have interoperability we can do stable coins you can do marlow these things all will come naturally and yes dexes will be here yes oracles will be here and they'll certainly have all kinds of cool features and functionality there now it'll take time for those things to mature become more cost effective and efficient and that's just a natural evolution of the system and our users have confidence that'll come because they see the spiral and they understand each part of the spiral is a new update a new upgrade these types of things and they've seen it happen before and what's nice is parallelism comes as you get deeper in the spiral you can now do more things at the same time and we already see in practice the application of intensive high computation high expense stuff being done in bespoke domains and those domains will get connected on the cardano none of this was possible before plutus existed and now that it does all of these things are possible okay you don't throw out all the innovations of the last 12 years you don't throw out the concept of a blockchain and what bitcoin has accomplished in that methodical pace you don't throw out the all the security models that have been painstakingly figured out you don't throw out all kinds of stuff just because some junior developer can't figure out how to do something what you have to do is you have to teach them and you have to show them and you have to build the infrastructure you have to build the tools you have to build the abstractions you have to build the services okay and these are economic opportunities especially on the services side uspos you keep saying we need more money we want to make more money well guess who's going to be the blockchain as a service provider for a lot of these things you guys it'll be done in a federated network it's just another revenue stream at some point that's the power of cardano and because we thought about it carefully we get things like babel fees we get more predictable fees we get resource determinism we get local processing a significantly easier time sharding god it's going to be fun to run sharding experiments with pluto's and parallel chains we can do that and we have a model for how to roll that out it first has a disconnected side chain and eventually as a connected chain and it's the very same thing that mamba is going to give us with evm will give us that okay all of that comes and it's built into the design of the system it's built into the evolution of the system and you regress nothing you don't lose the security model where it costs tens of billions of dollars to attack the system you don't lose the upgrade ability you don't lose the native asset standard you don't lose the metadata standard you don't lose all the tooling that's being built for plutus and all the formal verification capabilities for plutus you don't lose all the new design models and if it takes an additional six months to add a new computation model that just means those developers will do something else for that six months and when you have it you have it this is a marathon it's not a sprint and if i am being told the only way for a system to compete is it has to have everything today then then first off i'll ask what real life massive change has occurred because of blockchain in our industry are we running the internet on blockchain are we running our national id system your voting system your supply chain system your stock markets your central bank's currencies on blockchain quite yet the intellectually honest answer is no and the reality is these changes will take years to decades to percolate through society there are no standards in the industry yet the so-called market standard ethereum is literally upgrading itself to a radically different design and interoperability well the the things that they're going to keep the same we will have so it's all going to be commoditized and it's going to be about cost ease of development it's going to be about maintenance cost the life cycle the security model and also product market fit the marketplaces that you're in cardo has always been the project where we've tried very hard to balance all of these things and not be everything to everybody but understand how people can self-serve and they can add capabilities and functionalities to it so if you want specific tutorials on concurrency they're coming and they'll come as early as the conference because you're actually going to see running live code of the different development companies that have come in if you want a detailed information on how to be an app developer well we've already done two cohort cohorts of the flutist pioneers program the third one will probably be the best because it's the first that'll have the pab integrated with it that pab integration will be in beta state in september 10th and we'll keep adding to it and then over time interoperability comes and side chains come in partners are building like dc spark and we're building them too and that adds the legacy programming models that a lot of you developers are comfortable with and it's there and then what's really cool is you'll have interoperability between these systems so they can start talking and calling each other and yeah sure a lot of clever infrastructure has to be built and you have to keep upgrading the pab and all types of things sure that'll come it's a process and every day we get a little bit better every day we get a little bit stronger and there's more utility inside the system but no regressions in the system there is no notion of a radical change where we have to throw out everything and build a cardano too really no it's more of a spiral that's always been the point and you just keep adding and adding and adding and adding and adding and adding and you have the intellectual honesty to admit the wins of the past bitcoin was a huge win utxo makes a lot of sense for a settlement layer there are different account models there are equivalencies between accounting models in some cases non-equivalencies so it makes sense to permit diversity and that's what we do with csl and ccl and we wrote that down in 2016 explaining that to everybody okay this was not exactly a new concept it's been around for quite some time and i think it was published in english in 2017. so you've had four years to kind of know about that and everything is built out in the open it's all transparent code of the pioneers and all these other people so it's my saturday i now have to go up to laramie i will see you guys later i have to go quickly but because i don't want to miss the kickoff but i hope this helps people a little bit keep their their eye on the prize and understand where we're at what's going on the 12th is a major day but it is just a day just like mary was a major day shelly was a major day and there's many more to come there is a nice pulse of innovation and we're getting that into a quarterly cadence where major changes can be made quarterly minor changes are made intermediate and it's a two-way relationship there's a lot of people building things as those things are built those buildings will change in underlying infrastructure and a lot of experimentation is going to happen but every time that happens it's done with safety in mind predictability in mind and stability in mind so that you don't have mass regressions certainly a lot of things that i'd like to be sped up especially on the wallet side i'd certainly a better light client experience these are high priorities and you just invest you hire great teams and you work at it for a long time the hard part was figuring out the flow how do you write great papers how do you write high assurance code how do you get predictable releases how do you get upgradeability how do you build a great community all of those wins aren't going away they're here and that's the single hardest component everything else is implementation details off-chain infrastructure on-chain infrastructure augmentations to plutus these types of things but what's nice is there's some degree of stability and predictability about things one of the reasons why i love wolfram as a partner if you look at the wolfram language there's been a remarkable commitment to backward compatibility there some of the stuff in the i think is the 80s still runs on their infrastructure today even though they've added so many new capabilities and features and magic in wolfram's plan and that was not by accident was by dogged design and commitment and it's the same situation here your byron transactions will always run shelley transactions will always run an hour in gogan era and there's many more things to do many more things to add but those capabilities are there that's a magical thing for us as an ecosystem and that's what we're doing so i hope this helps a little bit and as i said there's a time and a place for these technical conversations and i feel workshops are the best place to do that i feel training is the best place to do that that's why we have the plutus pioneer program that's why 3 000 people have run through it that's why hackathons exist that's why cardano stack exchanging says everybody over the internet there's a lot of people out there that are cynical and no matter what we announce what we do they're just going to pooh-pooh it either because they don't understand it they don't it they don't value it or they have an economic incentive to lie it's one of those categories so i hope this helps a bit one of the gifts i got recently was let me go ahead and kill the screen share one of the gifts i got recently was happy happy octopus and you can flip it around and you get statipus and some days we're happy puss and some days we're sadopolis but you can't take things too seriously most people do and unfortunately that's why a lot of problems materialize in the world we're having fun it's a beautiful ecosystem and there's a lot of great people and a lot of great things being constructed and the point of cardano's be around in 50 years or 100 years we got the model of innovation right and we got the model of collaboration right that is more important than anything else all the other stuff it's inevitable it'll happen might take months might take weeks might take years but it'll happen and if you don't the model well eventually you can add your own and what we've done is i think the most balanced hybridization between the innovations of bitcoin and the innovations of ethereum with an eye on how do we achieve the third generation how do we achieve scale governance and interoperability and i'm pretty happy about that model and there's 500 people in my company well over i think 200 at this point are working full-time on nothing but cardano there's 12 other development companies that are working on infrastructure in cardano everything from ledger support to hardcore stuff on the network stack and there's a whole constellation of people now building dabs and so forth and there's catalyst with a billion dollars available and all kinds of people voting that's not going away that's only going to add and add and add the model that we've developed has the ability to absorb those business and technical requirements and upgrade and scale up the system now now and then there are checkpoints the cardone 360 is a checkpoint the conference on the 25th is a checkpoint and as that model matures and materializes and gets really strong next year the foundation can do tons of hackathons to encourage more developers and that'll happen in addition we're seeing no indication of slow down on catalyst and projects and programs and people there's a lot of push there one thing i'll never succumb to is cynicism you when you have a hundred billion dollar ecosystem you have to be very careful and delicate about how quickly innovation is pushed and how quickly you add features and functionality or else you run into a lot of problems what we've been able to balance is that reality with the absolute need to keep moving forward aggressively which is why consistently we've been number one in github commits and i know read it again and again and again oh but those mean nothing those mean nothing those can be gamed guys we're not we're not pushing trivial commits where we're like oh well we changed the variable name for the string these this is real commits this is real code anytime someone says that you'd ask them to go through the commits and identify which ones are superfluous it's an indication of a massive collective engineering effort at a scale this space hasn't seen before we're moving frankly faster than most people ethereum is still trying to figure out how to build that proof-of-stake route of trust not only are we there ours is secured in the tens of billions and it has a beautiful upgrade path and we clearly understand where auroboris needs to go and we also understand how to build secondary chains that can connect on they're still working on that the programming model that we have at times can be a bit inflexible and restrictive but it was built specifically for interoperability and for the movement of vast amounts of value and for safety verification resource determinism inclusive accountability and the requirement of only local state not global okay so given that that was the design space for version one it's unreasonable to go ahead and say well it doesn't do x or well it's hard to do x right now that'll come and it's a natural organic progression of these things and you enjoy all the things that we've already achieved and accomplished as an ecosystem if you can't get that maybe this isn't the ecosystem for you go somewhere else if you think everybody else has already solved these problems you're being lied to these are some of the hardest engineering problems in the world they're not going to be solved for decades because there's only partial solutions and every iteration is another example of a slightly wider solution that we have okay that's where we live as an ecosystem in a space that's why bitcoin doesn't have smart contracts like ethereum does or cardano does at the base layer not for lack of trying look at taproot look at simplicity and their jets and all these other things that they're doing there are a lot of really smart people in that ecosystem who hold great credentials and have been doing this for over a decade we're really really trying to add these things in but they understand there's a trillion dollars at risk and they only get one shot and if it's done incorrectly all goes to hell it's a hard situation to be in so you can only move at a glacial pace what we found out how to do as an ecosystem is be able to upgrade value like that but move at a much faster pace so we can out evolve them over time and do it in a very inclusive way where you get a say through catalyst and do it in a way where the unit where the universities are involved and the application developers involved there's a beautiful feedback loop of community commentary that comes through and very well qualified people very strong opinions about things who will inevitably have a huge amount of influence over the evolution of the system that is the heart and soul of cardano and what really matters and that's why it has so much value it's not about some arbitrary feature or functionality it's about the collective wisdom of a crowd of millions of like-minded people that want to evolve an organic system to become something that can work for the entire world and solve the problems of the world and are committed to that emotionally philosophically spiritually financially in their own ways that is why the value is there as long as that crowd's there the value stays and that crowd's not going away in fact it keeps growing and getting larger and we have the right process to ensure that these innovations are done responsibly peer review formal methods and the hard four combinator these types of things they work together like hand and glove and they allow a systematic methodical pace of innovation for the system and we thought really carefully and wrote down years ago about how the model works you have to build in layers you have to allow ephemeral domains permanent domains off-chain processing you have to introduce heterogeneity in a very responsible way but yet preserve inclusive accountability these are not buzzwords these are things encapsulated in papers in protocol design and now code and the single most exciting event of our ecosystem the launch of gogan is imminent it's happening in eight days there thousands of developers are going to start building breaking and thinking of things and that will force faster evolution of the system more tooling of the system and over an arc of time we'll be in a very good position we never have regressions because well people stay so ask yourself where are we going to be next year this same place no it'll be a significantly more decentralized and vibrant ecosystem where were we last year where were we the year before look at the growth look at the progress that model that mindset is the key and you can absorb everything else so i hope that makes sense to people if it doesn't come to the conference and maybe it'll make sense there participate in the hackathons maybe it'll make sense there if not go camping for six months and come back and see where we're at and notice how it's more and look at the evolution look at the progress that's being made guys this is an adult project it's professional involved a lot of people a lot of careers are connected to this there's an enormous amount of capital invested in this infrastructure and other such things people are really working hard and they're really thinking carefully because they understand what's at stake this has nothing to do with octopus swap or whatever or issuing nfts representing cute stuffed lobsters this has everything to do with saying we need a financial operating system for the countries that don't have it that's where we started back in a ted talk in bermuda in 2014.