CVA Social Club: Smart Contracts for Smarter Business

Transcript

Welcome everybody to this uh new um yeah edition of the um CVA social club from the enterprise blockchain working group. Um this so my name is Dennis Flat. chairman of the enterprise blockchain working group of the crypto valley association and in that context we always do a couple of times during the year some sessions with people that are working in the environment of enterprise blockchain solutions and today I'm very pleased to to welcome Bj Bj Hak uh Takman P Takman um from uh defininity the internet computer quite interesting

to hear about that what is the latest stages there and how we can actually use that for enterprise blockchain solutions or enterprise solutions in in general um because it's not all on blockchain but it's a decentralized technology >> uh which is good bern um is now six years with defininity >> correct >> and before that you you worked at the IBM as well in the research area that followed then after your PhD at the EC CH here in in in in Zurich >> and um yeah in that context I think you have a PhD in cryptologology and computer science and then everything comes together and it was kind of logic

that you that you come here and and work at Definitely thank you very much for taking the time and us as well here in the defininity uh foundations headquarter in Zurich and Um and the internet computer I read has over a thousand person years um of development inside. >> Yes. And uh so are you happy with the results or is the computer already finished or will it ever be finished? So it's it's not finished but it's fully operational.

And what I mean by this is that we have a pretty ambitious goal with the internet computer. And let me maybe explain this and in particular how it how it differentiates itself from other decentralized technologies and other blockchain based projects. So if you think about like software and data uh then like you basically have two ways if you want to deploy some software if you want to to store some data you have two two fundamental ways of how you can do about this first. So option A you can use your own hardware and you can run your software and store your data on your own hardware and option B is you use someone else's hardware right that's basically the the the two options if you go with option A uh then I mean you

retain full control you retain full ownership but you bear the entire load of maintaining this so you have to make sure your hardware is operational fix anything that is broken you have to keep your operating systems up and secure and always uh up to the latest level of security patches. You have to debug any upcoming networking issues. So, it's a lot of work uh that you that you incur if you want to keep a service running up 24/7 uh by yourself, even if you're a company, that's a significant load that you're taking on. With option B, it's kind of the opposite. It's like someone else takes care of operating the service, right? That's what you pay them

for usually. Uh Today you would maybe deploy to a cloud provider and they make sure the infrastructure is up well mostly up uh they make sure that whenever something breaks it's it's going to get fixed but on the flip side you give up all of the control right now they are in charge of your code and in charge of your data most likely if you're deploying to cloud the data will be immediately immediate immediately transferred to some pollier database I suppose um someone else can just flick a switch and you lose access to your code and data. And so that's a pretty pretty bad trade-off, right? So either you have to go with option A or

option B and neither of them is really good. And so at option C, huh? >> Yeah. Exactly. So at Definity, what we want to do is we want to think about this in an entirely different way. Like if you think about the internet today, like the internet is fantastic, right? any two entities basically all over the world can immediately communicate uh no matter where they are. It's fast. It's cheap. It's it's fantastic. Um you are just connecting to the internet then you're using shared infrastructure. If you want to make this secure you can add some cryptography and it's going to be secure. So so you're fine on that front as well. So that's great. But the internet is very ephemeral, right? If you send a message, it goes through the

internet, it drops out somewhere, the internet doesn't remember. And so we want to think about code and data as basically ex or about the internet computer as extending the internet for code and data. So so like you can think about it as okay, you now want to have a program that you want to run maybe personal maybe as an organization you want to keep control, you want to keep ownership, but you don't want to run the infrastructure. So what ICP is supposed computer protocol is supposed to achieve is to give you this like you can just deploy to the network. You will have to pay for the resources you're using just as you do with your internet provider uplink, right? You pay for the resources. Uh but you don't have to take

care of having the infrastructure itself operational. And so that's that's the grand vision we have. It's it works. Uh but of course we're not at planet scale yet, right? So we we uh the computer protocol is live since May 2021. So uh by now it's four and a half year four and a half years pretty much. Um and you can use it. It's usable. It's it scales pretty much but not yet planned to scale right. And so in that sense we are extremely proud of what we have achieved. It's I think it's it's really I'm so proud of of all of our teams on in in what they have achieved through their work over the last overall like

10ish years. Um but this is not the end, right? We're further we're going to increase the capacity. We're going to extend functionality uh also in the future. So it's not not finished in that sense. >> And now as you said it's four and a half years that your life could be quite interesting. So now really enterprise solutions or solutions that you develop for corporates together with corporates for enabling them to have uh applications on the internet compute and so on. What do you have what kind of examples you have? What kind of showcases you have? Mhm. I would maybe split in two cases here because as with a network uh there is the internet which

is a public network that is shared among everyone and that's useful for applications that everyone is supposed to access and then there's also the you can also have an internet right for applications that you just want to use within a certain context that don't need public access uh you usually would deploy them on an internet and so the same thing kind of exists with ICP. So ICP the main net is kind of the public internet. We also have an intranet version of this which we call utopia. Uh and that is particularly uh well suited for certain types of applications that we see more and more uh for I guess also political reasons reason reasons uh recently. And so I want to maybe focus on one case in the utopia one first uh

which is uh a project that we do with the uh UN development program the monetary authority of Singapore um and the central bank of Cambodia and that's that's uh so that's is a use case in Cambodia that's going to be uh based starting I think Q1 next year going to be based on on a utopia so on an intranet version of the internet computer protocol and so The idea here is that in Cambodia there's a lot like compared to Switzerland there's a lot more like informal business. There's micmemes small businesses uh that serve a larger part of the economy and they're by by how they work they're a lot more informal. Now sometimes they still need

access to financing. They still need access to loans uh for whatever they need to to buy to keep their business operational. But now they face a problem which is since they are so informal there's very little uh documentation and so they cannot just go to a bank and and get a loan because they may not have collateral they could give to the bank. They have have kind of partial documentation and so this project is more than just a technical thing. It's it's like an entire uh like use case solution. >> Mh. >> And the idea of this of this entire project is to bring more uh like credentials more data in a trustworthy way in such in in a platform where banks can access that data and use it to make

their risk estimate about a certain loan. So, so with the data you mean like uh you know balance sheets or even like a a um a depth book that you see okay how many how many incoming funds does the customer have and and and >> could be but it it could also be for example utility bills and so that you see okay this this theme microme has paid their utility bills uh always on time over the last three years probably they're pretty trustworthy and then maybe you have some some maybe they did a bigger contract some year ago uh and and they can have that as kind of a credential. So we'll be very probably every case will be slightly differently what kind of

documentation they will have and use. >> But the idea is to make this kind of a trustworthy verifiable data uh um sharing between the uh providers of that information which could be could be banks, other authorities, could be utility companies, whatever. And on the other hand the consumers which are then banks that make the decision about the loans. >> So actually that the other part is are then verifying what this particular um small small enterprise or the microenterprise actually has uh and then that gets automatically collected in that kind of marketplace or how can I

envision it? Is it a is it then like a platform where I go as a as a smaller medium enterprise apply for a loan or >> more it's more like the bank goes there and says okay here is a set of requirements that we have and that we want to have fulfilled >> and then the platform can see whether there's someone who can provide these uh and they can provide this to the platform and it's going to be forwarded to the bank. So that's basically the idea. >> Mhm. Mhm. Mhm. It's like the the theme will not interact with that platform. It's it's kind of between the entities uh that provide the data or uh consume the data.

>> Okay. Okay. So the need to go to the to the bank actually it's not a cryptocurrency solution. It's really about >> uh document integrity. So >> yeah that that sounds quite interesting. So looking forward to hear how successful that has been launched and then how good the yeah the small enterprises or small small uh medium enterprises there will will >> we of course hope they will flourish >> get to more loans. [laughter] Yeah. And that is on the private environment.

>> That's on the Utopia platform. >> Utopia platform and then on the on the public one what kind of enterprise use case you have there developed in the last four four years. >> Yeah. The one that I would highlight is uh is a collaboration with Federi which is a government agency in Italy that takes care of the made in Italy brand and the origin foundation which is another entity here in Zurich uh that has done actually the development on this and so the idea was of this project is uh it's live uh is to bring uh kind of the made in Italy part on chain so if you uh if if there is a company particip participating in this program. Uh they

can basically print a QR code uh on their products and if you scan the QR code, you get uh to the corresponding application on the computer where you can read kind of what makes that product made in Italy, what's the value provided and so forth. So this is something which which as an application in is inherently public, right? you want every uh every potential consumer of these products to be able to validate uh this information about the product. So this is something that clearly uh has a focus on on ICP mainet. So that's also where it's built.

>> So basically a very similar case just that there is not the bank that asks for trustworthy information and data that is when then the consumer ask for trustworthy and validated information because exactly I think that that fed Federally is is is an is an audit company isn't it? They are doing a lot they're doing a lot of audits for these these products when I understood it correctly. >> Ah that's that that's a good question. I don't know what exact part of that kind of value chain in the verification process they play.

>> So for us they are kind of the entity we talk to in order to uh to to understand the use case but like we're more on the technology side. So I'm I'm not sure exactly that >> basically on the sources that I have read preparing this uh this meeting is that they are really taking the production companies or the producers, the farmers, the wine makers and so on really through an intensive auditing process >> of course but this they do for the main >> so they are actually validating then the data that is then in in the system in interesting it's because you have not mentioned any kind of trade cases at the moment as showcases you have really uh uh so where really value gets moved from

one corner to the other the really broad cases where it is about accessing rights accessing information accessing uh uh validated information um that's quite interesting that people are then taking here the internet computer I guess because it is decentralized and it is non-manipulatable And it is uh >> there's actually one part let me try to keep this as non-technical as I can but there's one one part uh that that is from a technical perspective that is very specific about uh ICP as a decentralized protocol compared to basically everything else and that is what information do you need as a client to validate the correctness of

information that you read. Mhm. >> Like if you think about a blockchain, like if you think about Bitcoin or Ethereum or whatever, you basically need a copy of the blockchain in order to really like on your device decide whether a certain part of information is correct or not. There's some simplified cases like Bitcoin SPV where you only need block headers and not the full blocks. But that aside, you need a significant state like at least tens of megabytes, usually more. Um, and so for ICP that information is a very compact public key. So we use a lot of threshold cryptography in ICP. That's kind of our specialtity. That's how we roll. Um, and so in order to verify information from

ICP, you just need a like 100 byte of static information and then any response that comes from ICP, you can immediately validate for authenticity just based on this. Mhm. >> And so in particular you can validate authenticity in every web application like on every embedded device basically uh you can do this which would be completely impossible with with other types of blockchain or blockchain like technologies. >> Yeah that's probably as well I mean on your website theinity website you claim as well that actually um you know smart contracts is the big buzz word.

>> Yeah. and uh and a couple of years everybody was talking about smart contracts and how important they are and they will change the world and so on. Um but at the end I think the reality as usual with hype technology was a little bit yeah [laughter] I mean there are use cases and so on but you say as well that smart contacts are just not they don't just don't have enough data >> and that's why they can only fulfill their job limited why I think you're claiming that with the internet computer this data restriction is actually solved >> I it's it's a lot looser in a sense.

There's a lot more data you can store. So if you if you look at a at an application that is built on Ethereum or built on Solana then what you will usually see is that there is some token component of that application that actually lives on the blockchain as you would expect it. And so if you do a transaction on tokens that will incur a transaction on uh the blockchain and that's perfectly suited for certain like DeFi cases, right? If you if you look at DeFi and you just want to exchange token A for token B or tok take a loan in token C for token D or whatever um then this part of the the application actually lives on chain but all the rest in in those systems is actually

offchain. So if any more heavy computation has to be done that is still hosted on a usually centralized server in the cloud today or if you uh if you use a front end in order to interact with a smart contract that front end will be stored uh in in the cloud. And now this may seem like a theoretical concern and why would you care? I think one great reason why you care is is security. Um, so you may remember that half a year ago or a year ago there was one of the biggest hacks ever. There was the bybit hack. And the way this was executed is um that someone got access to a machine of one of the lead developers that could deploy the front

end. Mhm. >> And so what they then did is they introduced code into malicious code into uh the front end while it was on the developer machine and the developer was pushing a malicious version of the code out to to AWS. And now the code was pretty specific to exactly attacking certain wallets wallets at BB. So it was pretty smart. Uh it was a well-run attack. I mean we have to give them that. Um and it was successful, right? And the problem here was that there is the full control over the front end was with one developer and later it was stored just on a single cloud server.

>> Mhm. >> And so in in uh in in uh the inner computer ecosystem we think that's not a good way of storing such applications. We think if these aspects of the applications are crucial for the security and the correctness of the application, then they need to be handled with the same care that we do that we use for the token part because the token part was fine, right? It was really just a front-end exploit. And so in the inner computer protocol, you can host the entire application uh fully on chain. You can host the back end on on chain like one we call them canisters.

So basically one smart contract on on their computer can have up to 500 GB of memory um which is approxim approximately as large as complete other blockchains right um and um you can host the front end on chain. So by hosting it onchain you make it verifiable that there is no back doors. So the entire thing becomes a verifiable application uh that you before using it you could validate everything that runs on chain.

>> Um and so you can really be sure that nothing has been tempered with when you use the these applications. And so we think that is that is kind of a missing piece in the current web through story that we have to solve and you can solve it on ICP >> and and I mean that that's probably what we then the the link as well to the enterprise world because what's an enterprise what what are enterprise solutions actually they are always working on business critical >> processes within within enterprises and the last thing you want to happen is that they are uh corrupted.

>> Yes, >> exactly. >> That's that's what we're saying. That's why I think this solution is quite good. I want quickly to open as well to the audience for some questions if they have please. >> I can justify the first one. >> Um so I'm not a technical person. Um and I I still don't really understand. So can I imagine it the way so let's say comparing it to other let's say the Ethereum blockchain is my understanding correct that while on the Ethereum network there is like the there is there's a smart contracts which are running on chain and then typically the apps are then deployed offchain as you explained whereas on the internet computer the whole stack application is

running on chain distributed which means that they are then each and every node operator is running the full code and verifies every time the the application is executing something. Is this >> Yes, >> that is correct. The internet computer uses a different model of decentralization. Um I can go into that later if you want to. Um but using this we call it deterministic decentralization. Using this we can uh use a lot lower replication factors. So usual applications run with like 13 times replication but we have many of the subnets. So the the the uh the the network scales horizontally. When we add more compute

more machines to the network the network capacity grows. So that's very different from other blockchain based networks. And so due to this construction we can actually offer a lot more resources to the individual programs. So the processing capability of the network is just so much bigger that we can enable hosting full applications on chain which is difficult otherwise >> and it's true that so in the public domain that you're running so I could set up a node and then I could basically deploy my own code which would then be running on the entire network. Is this kind of is it a similar logic or you

have some sort of a validation mechanism that >> it's it's kind of decoupled. You can use the you don't need to run your own node. You can just deploy to the network. You can validate everything you get from the network cryptographically. There is no need for you to run a node. However, you can run a node because you want to contribute capacity to the network. Now the process of running a node is a bit different from other uh from other blockchains. For example, we uh like we use very symmetric configurations of the nodes to ensure that they have approximately the same speed and processing and so forth. So there's a couple differences but you can become a node provider on ICP and contribute to

the network. But that's com you can completely decouple this from using the network. probably that's that's a good bridge to my next question is so what does it actually what do I actually as a corporate as an enterprise maybe a smaller medium enterprise or a larger enterprise what does it actually take me to release and program let's say >> a website or or something um on the internet computer what do I need to take >> yeah so um we like since We we see ourselves as kind of a a decentralized cloud. We try to be in line with web standards, right? So what what the

programs that run on the intercomputer protocol use a standard called web assembly which you probably have used multiple times today without noticing because every uh current browser supports it and there's lots of plugins that are used uh in just in the internet. So it's a by by by by now it's a almost invisible but important part of the web infrastructure and so back in 2019 when we designed that those parts of the structure web assembly was still pretty nent we were among the first ones that were using web assembly in the back end but today there's actually like fastly cloudfare a lot more companies that offer also web assembly hosting uh in that case in a centralized back end

we offer it in a decentralized back end. So it's a it's a pretty widely used standard today and there's lots of compilers that compile to water web app assembly. However, it's still a bit of a nent standard in terms of programming back end. So there's some part of the API that we had to fill in because the standard didn't have them uh and that are ICP specific. So um if you you can compile to web assembly you need a specific software development kit though to make use of all the functionalities that that uh ICP provides. Um we at Definity uh provide two uh software development kits. One is Rust which is a widely used uh system ccentric memory save pretty modern language. Uh the

other one is Mokco which is one that we built in-house um which is more high level. Um if you're learning new it should be easier to learn than Rust. Um it's very well integrated of course with all the inner computer features. Um so it's kind of it's a different approach. Those are the ones we we use or we provide as defininity. There's a couple communitydriven ones for JavaScript for Python for C++. So you need this kind of software development and the skills to do that >> and then you can just deploy basically.

And if you want to if you want to just try it out, we have a couple things as well. So one is this year it's caffeine. It's kind of our newest venture. Uh it's an AI based tool uh that you can use to build applications in particular whatever prototypes smaller applications is is fantastic to do this. Um caffeine.ai if you want to try uh you you'll get free credits to start. Um, and so with this, you can just chat with an AI on the other side. You can tell it, oh, I want to have a website for my party where people can register and uh later after the party

can put all their photos so they get shared and people can like photos. And then it's going to build an application that does exactly this and deploy it directly to their own computer. or as an enterprise, you could use that, for example, for prototyping something that you want to build and see how your partners or customers or whatever like the ideas. Um, and depending on the complexity, it may be enough or you may still have to do some other development later. But I think you know when when I look at the corporate world and I mean we do as well in my company a lot of um product developments with with corporates and so on then of course I always see that they are working of course as well with a legacy

>> and they are working with a standard procedure and people that are you know working with standard codes and so on and so on. So is it as simple as you say to develop or do I need then somebody from definitity that helps me on the project as a as a corporate when I'm really interested in in in in doing a solution like uh uh like the UN uh thing that you that you did for example. So I think first of all it depends a lot on whether you want to target mainet which has a pretty easy entry or a utopia where you have to where we where you have to set up the

network which is kind of a complicated process by itself >> by nature because you're setting up a decentralized network. So that by nature is is harder than just uh kind of installing a single single computer, right? Um if but if you're if you're just thinking about the software development um I mean sure there would be some some level of learning curve for everyone who who adapts but like if you have general purpose programmers in your company uh they will be able to adapt. Um or we also have a program called the ICP alliance in in which we have a couple uh companies that focus on developing on ICP. So if you want to outsource the

development then to a different company that's that's also an option. So I think it depends on the particular case but it's like this is not rocket science right it's it's like >> building software which is what software developers do. Yeah, but in a corporate environment it's always a bit difficult especially when you go to the large corporates. I mean when you have a small and medium enterprise they probably can or have a business idea. I do that these [clears throat] things but in the corporate I think you know you know in the crypto valley enterprise blockchain working group we once had a really deep sessions on okay what are the success factors or enterprise blockchain solutions and we actually concluded them on on five ones. So first of all it needs to add value in the means of it

needs to replace something or make something better that hasn't existed before because why corporates are like that why changing why changing a a winning horse >> you don't do that I mean we had enough use cases about oh we need to replace uh uh payments yeah but you don't need to replace domestic payment because they work already very efficient you know when you when you want to use something then do crossboarder payments, you know, and these things. It really needs to be value ad. And then another thing is uh that it needs to be easy to bring real assets, real world assets into digital assets so that they are exchangeable or replicatable or

presentable. uh like in this Italy uh project you know I mean a bottle of wine how you bring the quality of a bottle of wine into into a onchain solution uh that's the thing then of course the reach of the people uh how many participants you have how big is the network and the fourth one um apart from a governance model which is the fifth one but the fourth one was the integration with existing workflows what we concluded So because usually you don't go into an you know into a blue ocean or green field environment there is usually

already something there are already databases that you need to connect these databases are centralized there are workflows you need to correct that are enterprise resource systems you need to correct and so on. H how h how I think this is quite complicated with the having a decentralized solution connecting with components um of an enterprise in which are the more in in a centralized environment or >> not necessarily it depends a lot on the particular case. So as as I mentioned before like ICP uh tries to speak web as much as we can and so in particular also

uh the the smart contracts the kenesis on ICP uh they speak web as as much as we can make them speak web and so in particular um that means they can offer an API to the outside that is HTTPS right so you can you can just make an HTTPS uh call to a canister and the canister will be able to uh respond to it with an HPS response. So there's no no particular uh like uh module you need to put in between. It's just from an exist like if your existing uh system can make an HTTPS out call to do an API which is today the most common

way of doing APIs then the cancer will be able to answer to that API call. On the flip side, the canister can also make HTTP calls, which is also something that that other blockchains usually cannot do. Um, and so the canister can both provide an API to the outside and consume an API that is provided by by an existing system. And so depending on on what parts of an entire system you want to uh potentially want to move onchain on ICP um you can keep other parts um offchain and and that's still going to work and communicate as long as uh

everyone speaks HTTP which again today is kind of the standard way of building APIs. >> Mhm. And so in that sense it's hard to to generally answer all potential use cases. >> Uh but it it gives a very good tool set to integrate with existing use cases also in in a way where you maybe move things partially. >> Okay. So, so what would be then the kind of use cases or sub use cases or partial use cases you would say if a corporate has these and these issues or or topics or faces these and these challenges really thinking about the um internet uh computer would be

would be a good option. >> Mh. >> What would be what would be your advice to these to the corporates? >> So there's again the two different things. one if I if I answer the the utopia the in infinite one first I mean we've been talking to lots of different entities over the last months which relates to the kind of degradation of political culture and trust internationally and so these are usually like government like a or or NGO type type organizations that now realize oh they are actually putting a lot of trust in the United States owned clouds, right? And they realize that they don't

feel comfortable with this whether it's an existing or a new use case. And so they want to they now realize uh that they actually care about their own sovereignty, about their own data. So a lot of the contexts we have these days and and upcoming projects in the utopia space are really about specific parts of their data which are super sensitive. So it could be related to weapons, it could be related to money like these types of things where you really really want to make sure >> kind of paradox. I mean it's a super sensitive data and you are instead of putting it into a safe scene and protecting it, you're putting it

>> what you're using then is the utopia vision which is the one where you operate uh the infrastructure. So you only gain if there's some independence between the different parts that operate that infrastructure. So you will need for example multiple different data centers or the nodes should be uh controlled by different entities for example in the same country or in the same government. So you actually benefit uh from from the from the guarantees otherwise >> if one controls all nodes you don't gain right. So um and so then but what Utopia helps you in that case is to protect the integrity against insider attacks.

>> Mhm. >> So now once you have this even if there's one party that for whatever reason is malicious and tries to uh kind of change the data or read some data they are supposed to. Um they cannot because the network is not going to allow them. So if you are willing to to take this additional it's it's additional complexity right it's additional complexity but if you're willing to take this you get a significantly increased level of security and sovereignty >> and so it's use cases that really benefit from this where we see now a lot of uh requests for for utopia >> and for mainet I think it's really if

you want to build something um where you may want to integrate with newer things like stable coin payments or that kind of thing but your entire information flow you want it not to be hosted like you don't want the the trust of the cloud you want to have this thing secure I think that's a case where where I would say okay then if you want this or if you want public verifiability of what you do then it's a case for IP ICP mainet so as with federally like you can go and validate not only the certificate that federally gives you you could validate the actual software that generates that certificate.

>> Mhm. >> Right. So it's like you get a very sophisticated level of verifiability that you cannot get otherwise. >> Okay. Okay. Very promising future for a lot of yeah confidential or sensitive data that we have out there. There's some more questions from the audience. You see how internet blockchain solutions blocks and their platforms maybe create their own blockchain ecosystem and how you see your position because

we know trying stuff. How do you see the future in blockchain institutions blockchain? >> So besides ICP solving a different problem than other blockchain. So it's not really necessarily competition for us. Um one one strength of ICP that I have not mentioned at all uh so far is that ICP is inte is is very good at integrating with other chains. So it an ICP smart contract can uh make out calls or it can hold Bitcoin, it can use Bitcoin, it can make out calls to Ethereum in a way that is secured by the network itself.

Um so you can so ICP can enable smart contracts to interact with uh with tokens with contracts on other chains. Right? So for us, if there's more other chains um that we need to interact with in order to to help the project the the projects building on ICP, that's fine. We can do this. So for us, we're like that's that's not competition for us. That's collaboration. And so that's the way we see it. I don't think we really have uh competition in that space.

I don't know if you're you're finished but yeah okay >> can you share a little bit about I don't know if is there a consensus mechanism on the network and and what are incentives to basically I I think you said to contribute >> resources >> by running a node >> yes so ICP is is what other blockchains would call a sharded network we call it a subnet a network consisting of subnets so there is groups of nodes again these subnets uh that uh together run a consensus algorithm. So the consensus mechanism we use is um as a standard threshold consensus. It's uh it's kind

of a so in I would say if you think about the traditional consensus is like PBFT but moved to a blockchain context. So, it's a bit simpler because we use the change structures, but it's like this is it's a pretty cool thing, but it's not the most groundbreaking feature that we have. I think it's a very good consensus mechanism. It's very fast. Um, but it's it's not our main distinguishing factor. What is our main distinguishing factor is that we build this threshold cryptography around it. So every subnet has cryptographic keys that are whole distributed among the nodes. So not

every node has a copy but only all nodes together have a copy of the key. Um and this we use in order to do this verifiability of information. We also use it to do the interaction between uh the subnets. So every subnet does its own consensus. Every subnet does its own cryptographic key operations and then they communicate with each other using what they do with this cryptography and consensus. That makes sense >> necessary to actually achieve consensus.

>> We also no we we use it in in one particular part of the consensus but it's more like um that's what we do. We build it on top of the consensus. It's part of the core protocol. um but we need it on the on on the next layer to do the interactions. So that's kind of the the I just meant like this is the part that is really unique to ICP that doesn't exist elsewhere. Uh whereas the consensus is good but you know you could use a different consensus algorithm. >> The incentives >> the incentives is um if you're a node provider you have to like there's ICP has a governance system where everyone can participate. You can get ICP tokens, you can stake them in the governance system and then you can participate and

so the governance system will decide about applications to become a node provider basically. So if you want to become a node provider, you say, "Hey, governance system, I want to run a node uh in the following data center." And then the governance system will say, "Look, this node is not something that we need because there's too much capacity there already." Or it will say, "Ah, that's fantastic. We always wanted to have a node in [sighs] Cambodia. We don't have one. So great that you're offering one and we gladly accept you to the governance system."

And then if you run your node and if your node uh contributes to the network, you will receive token payments from the network that will remunerate basically your expenses for running the node plus of course an upside because you have work uh keeping your node uh up and running and the network wants to help you in providing good service. Good. Please mentioned that smart contract call HTTP or do HTTP requests.

>> Yes. >> Does that make it easier to hack this smart contract? >> Um, doesn't make it easier [sighs] for this. completely designed data that could destroy your smart contract. >> I mean, you can always have bugs in software, right? And and you can write code that has bugs. Uh but I don't think it fundamentally changes the picture here because your your smart contract

will anyway process data that comes from the outside world. And the main difference here is usually smart contracts process data that is actively sent by users. In the HTTP case, uh the smart contract is processing data that it requested from the outside world, but it's still kind of the same fundamental thing. So I don't think it makes hacking easier or harder. Um you of course if you have a more complex software system you may just due to the complexity you may introduce additional things but I don't think this particular feature as a feature is something that adds vulnerabilities. Mhm.

>> So it really comes down to find the right use case for the using this decentralized internet computer. >> Mh. >> And um and that is probably what a lot of corporates will will will get engaged with. And I think what I what I took away is when when you have an data that you that is sensitive and where it makes sense that it is decentralized stored >> and decentralized so validated and verified >> and processed >> and processed. Um so then that is the where you your where the sol where the internet computer brings the biggest

advantage for an enterprise solution. >> I think yes that's what I would say. >> Yeah that is that is what I take away and I think that is as well a good uh finding at the end for this session. Thank you very much B for taking the time and uh and giving us these insights um as a head of research at Definity in where the internet computer can be used for enterprise solutions. I think yeah as you said you are still already four years on the market. The first solutions have been already launched. First showcases are there another interesting ones are in the pipeline and the coming.

You have a lot of interesting talks with people that are looking how to protect sensitive data from American um yeah providers actually [laughter] >> and uh and um and and and I think it will be quite interesting next couple of years with the internet computer and I am really looking forward to see how it will succeed. Thank you very much. So thank you very much for the time. Okay. And thanks to everybody as well online for joining us here for another session of the uh crypto valley um social club of the enterprise blockchain vault and I

hope next year we will have again other interesting talks and I'm very pleased to welcome you there as well and again thank you very much.