The State of Algorand - TOKEN2049 Singapore 2023
TOKEN2049·Mon, Oct 9, 2023, 12:00 AM
Transcript
I'm John Woods I'm the CTO at the algoran foundation algorand is a killer layer one blockchain and I'm here today to tell you a little bit about some of the things we've been up to and where we're thinking of going with algorand it's a short presentation I think it's quite Punchy so thanks for coming out and listening to me so algorand launched on mainenet in 2019 it's four years ago and in that time we have gone from a very initial implementation of algorand to one that is quite sophisticated so over those years we've processed 1.1 billion transactions on the mainnet and we've had over 31 million blocks but critically this all occurred with zero downtime uptime in systems is hard uptime and distributed systems is even harder and so algorand has managed to process over 1 billion transactions all without a single second of downtime and this gives algorand a key characteristic which is absolutely essential reliability so thinking of algorand in terms of enterprise-grade Technologies I think you have to remember that programmable blockchains at their essence are really operating systems they're decentralized operating systems but they're still platforms for executing applications and when you're building a traditional application in an Enterprise context you want to make sure that you're executing on a reliable platform and that's of course why the internet is run by Linux and it's why when Mission critical applications are being executed they're run on embedded or very reliable operating systems algoran gives you that reliability in a decentralized context so algorand is a turing complete layer one which means you can program and express any level of sophistication within a smart contract it has 10 000 TPS in mainnet and it can support that and very importantly where a lot of chains will talk about transactions per second which is not really a metric that I think is that important once it's over a certain threshold of course algorand's TPS values don't dip when you're interacting with a contract the way other chains do so you'll see other chains and they'll say 50 000 TPS but then when you jump into it and you see what is the TPS value when you're executing an amm when you're executing a swap when you're interacting with a sophisticated contract and of course what happens there is that that contract is running inside the virtual machine on the Node on the network and the virtual machine cannot process that as quickly as it can process a standard payment transaction but alkarand has been engineered from the ground up learning from the mistakes that were made with the evm architecture and at the core of algorand it's smart contract execution platform is the AVM I like to think of the AVM like the Apple silicon of the blockchain space it's efficient and fast it focuses on speed and precision and So within the AVM you can actually play with Native 512-bit integers as an example Straight Out of the Box and so one other aspect that I think is critically important is finality and a lot of people talk about finality in the blockchain space there is one unique property of algorithm that sets it apart from its rivals algorand has true instant finality and what I mean by that is that with algorand once a transaction is in a block it is completely final on many other chains including great projects such as ethereum and cardano you will have sometimes short rollbacks block hype battles where a a chain that has been deployed to the to the to the network and has been accepted as valid is rolled back and replace with a long with a longer chain or indeed a different chain and so this happens all the time these blockade battles uh occur these chain roll backs occur all the time and what usually happens in an Enterprise context where you have a mission critical application is that you have to watch for these rollbacks and replay your transactions now that brings a ton of middleware into your into your application within algorand because of the vrf Primitive the consensus algorithm consensus is reached by 100 of the network on every block and so this means as soon as a transaction is in a block which occurs every 3.3 seconds your transaction is completely final so this reduced the costs of execution it reduced the cost of Maintenance of applications on the platform and it gives a killer user experience where folks don't have to worry whether a transaction is really done they don't have to wait for confirmations so what have we been up to over the last few months one of the things that I think is increasingly on the agenda is quantum security and so what I mean by this is all of the blockchains that we're using at the moment use elliptic curve Primitives they use elliptic curve cryptography and so this is a mathematical technique that uses elliptic curves to find over finite fields for the security of the digital signatures it's very well understood and it's been around since the 70s the thing about elliptic curves is that they can be the discrete logarithm problem which is the core of their of their mathematical hardness can be solved by a quantum computer trivially and so we're starting to see serious entities look at the quantum scene and start making changes so recently Google Chrome was updated in version 116 to include kyber which is a key encapsulation technique that's that's post quantum and so Google is looking at their browser technology and they're making code changes to make sure that it's ready for the Advent of a Quantum machine at algorand we're known for our Academia we're known for our prowess within the crypto within the cryptography space of course our founder Sylvia McCallie is one of the the best known cryptographers on Earth so we're taking the step to to include Quantum resistant cryptography in the layer one and so we've done this the right way we've brought Falcon which is a nist approved post-quantum digital signature scheme based on lattices to algorand so the entire history of the algoron blockchain is secured by a lot of space PQ signature scheme and so this means that even if a Quantum machine was to be attacking the chain at this moment it wouldn't be able to disrupt the history of the chain another wonderful uh technology that we've developed over the last year that I'm particularly proud of is algo kit one of the things that I think was a drawback of building on algorand was that it's not as simple as evm you can't just fire up your JavaScript devs and get them through all that applications on algorand you needed to deal with tooling that wasn't as mature as ethereums and so we recognize this and over the last year we've released Advocate which is a complete end-to-end solution for realizing your applications on algorand and this takes you on the entire Journey from build to test and to deploy and so we're on version 1.4 now and we have done a ton of things algorand helps you build unit tests it quickly helps you compile test with cicd Incorporated and indeed Deploy on either test net or mainnet within a few seconds we've also in version 1.4 added the ability for algo kit to help you actually build the web UI front-end for your dapps because of course adapt is not just a smart contract on the chain your users also need to interact with it so Advocate even helps you there it helps you build the front end as well as the back end so I'm particularly proud of what we've done with agricare but we're not stopping there so let's talk about where we're thinking of going so of course on the protocol side everyone's focused on TPS and numbers so we're going to be implementing Dynamic roundtimes and increasing throughput but I think even as it stands algo runs throughput is really high maybe more importantly to me is decentralization so I'll go around at the moment is a hub and spoke model that means that the blocks are produced in a permissionless way by participation nodes or consensus nodes but that data is propagated across the internet uh with a hub of relays a bit like a Ring Road and so we're moving away from this relay Style Network to a gossip or peer-to-peer Style Network using lib peer-to-peer and so what will happen is we will be increasing the number of consensus nodes that are on the network we'll be deprecating the relays and all of the data transmission or data propagation will occur right through a peer-to-peer gossip-based network not requiring any relays and we're going to make sure that we do that in a way that doesn't affect our finality of course at the heart of the consensus nor the current block time which is about 3.
3 seconds also another major tectonic shift in algoron's vision is that we're going to start incentivizing consensus a lot of folks are maybe unaware I know I was when I first started looking at algorand that block production on algoron is not rewarded the way it is on ethereum cardano bitcoin Etc and so to go hand in hand with our peer-to-peer gossip-based Network we're going to be incentivizing consensus and so this has a number of impacts number one it increases the number of channels for data propagation from peer-to-peer and most critically number two it will see a massive surge in the amount of algo that's being staked when individuals who have maybe smaller positions are together forming a large mass that's securing the network so I'm particularly looking forward to moving forward with real decentralization so finally I'd just like to talk a little bit about algo Kit 2.0 and so this is the kind of Magnum Opus of developer user experience on algorand as I mentioned a little bit earlier platforms like algorander are effectively operating systems okay and if you think of the developer experience on Android with develop with Android Studio or you look at vs code it's great right people enjoy using it it's the same on the Mac with xcode xcode is a great user experience and so algo Kit 2.0 is going to address one of the sub-optimal parts of building an algorithm right now so build test and deploy is now complete it's good we have great CI CD support we've got great test support there's only one part of the developer experience that I'm not happy with and that's the language currently when you're building an algorand you've got to use teal and teal is like an assembly like language it's not very fun to develop in it's certainly not fun to debug with and it's hard to learn and I think maybe most importantly for end users when you want to build an algorithm you want to take advantage of some of the cool things that we have like the 3.3 second block time the instant finality and all that cool stuff you got to go and you've got to hire specialist teal developer it's expensive and it's expensive to maintain it's expensive to develop rather and expensive to maintain so without Kit 2.0 we're bringing pure python development to algorithm and what this is going to allow you to do allow you to do is to build applications for algorithms layer 1 in real python without concession so this is not going to be a weird syntactic variant on python actual Python and you'll even be able to do line-by-line debugging and so this brings a huge change in terms of the inclusivity for algorand so everyone from kids and University who are learning prototyping in Python for the first time all the way to of course the Hot Topic at the moment the ml AI guys who all build in Frameworks around python everyone will have the knowledge to deploy contracts on algorithm and so this is a significant change in how easy it's going to be to build and how inclusive algorand is as well as as well as uh that as I mentioned earlier on this changes the cost profile so it's no longer going to cost a lot of money to hire developers to build your apps and of course it makes maintenance trivial so that's a look at some of the things we've been up to thanks so much for listening to me I do really appreciate it thanks a lot
Automatic transcript — names and jargon may be misspelled.