AMA Highlights — Kadena
By Daniel Dal Bello, Director.
September 4, 2020–12 min read.
On Saturday 29 August, we welcomed Will Martino and Doug Beardsley from Kadena into the Hillrise Group Telegram chat for an AMA. Will is the current CEO and founder, and Doug is Director of Engineering.
Kadena is a “hybrid” blockchain platform with several distinct products, including both public and private blockchain architecture and a new smart contract language. Will was previously Lead Engineer of JP Morgan’s Juno project before leaving to pursue Kadena and Doug is an experienced software engineer with previous tech startup experience. Kadena’s mainnet token KDA has recently begun trading across several crypto-markets.
In this post, we have compiled key questions and answers from the event.
Daniel Dal Bello
Welcome to you both, thank you for being here today. Kadena to me has been flying under the radar for some time because this is really quite a fascinating company and product suite. Could you introduce yourselves and tell us about your journeys into the blockchain world?
Thanks for having us! I’m Will Martino, Cofounder and CEO of Kadena. Before Kadena, I was Lead Engineer for Juno, JPM’s first blockchain and what became JPMCoin v0. Before that, I was at the SEC as the Lead Engineer for a quant group that was building forensic tools that are now deployed nationwide. When I was at the SEC I also was the Founding Tech Lead of the SEC’s Crypto Working Group. My co-founder Stuart Popejoy and I met at JPM and founded Kadena to address the shortcomings we found when we tried to get the EVM/ETH to do more than simple tokens/debt at an industrial scale.
Hi, I’m Doug Beardsley. Thanks for having us, Daniel. I’ve been programming since I was ten years old. After undergrad, I worked in defence contracting for 6 or 7 years. After that, I worked in finance, and recently I’ve done a fair amount of frontend development. I’ve been writing Haskell professionally for 10 years and am the co-organizer of the Compose Conference and the New York Haskell Users Group. I originally met Will and Stuart through the Haskell community and was really attracted to Kadena because engineering excellence is in the company’s DNA from day 1.
Hi Will — great panel with Oracle, Microsoft and Baseline protocol today. What are thoughts involving Baseline?
My first introduction to Baseline was today. As it’s an open spec that we can build against, and our previously proprietary Kuro private/enterprise blockchain is now open source, there’s a lot of potential. My main issue is that we, and everyone else, tried to get the EVM to scale 4–5 years ago (scale meaning do more than simple tokens) and generally everyone failed. So I’m not a huge fan of private EVM state machines timestamping to ETH and instead feel that a scalable base layer, like Kadena’s mainnet, and bridging enterprise/industrial logic between contracts found on the public and private network(s) is the way forward.
That said, Baseline could evolve into that eventually, but the EVM will never make it there.
What is the most ambitious thing that Kadena wants to achieve in the next year?
Wow, we’ve got so much going on it’s hard to pick just one thing. If we zoom out past just one year I think Kadena wants to replace both Bitcoin AND Ethereum. A week ago we proved that our proof of work blockchain can scale when we transitioned from 10 chains to 20 chains live and in production. Bitcoin will always have a place as digital gold but we can actually deliver on the promise of digital cash. Because we can scale, we can also become the smart contract platform of the future that won’t be mired in crazy high gas fees and have a smart contract language that is humanly readable on the blockchain, easy to use, and purpose-built for the needs of smart contracts.
Daniel Dal Bello
On this Doug, when you say the smart contract platform is human-readable and easy to use, do you expect someone with no computer science or programming background to be able to pick up and learn this language to a functional level and if so in what time frame?
Pact is definitely a huge step forward in making smart contracts easier to use. We have some non-engineers in Kadena who were writing some basic Pact code fairly quickly. There are a number of things that you have to do by hand in Solidity that Pact gives you out of the box. Multi-sig is one example that comes to mind immediately. It is supported natively by what we call ‘keysets’ in Pact. These things definitely reduce the barrier to entry.
I know there are Fortune 100/500 companies already using Kadena. When will Kadena release the names of the big clients and have their logo on your website as an official client? We think this is important for more awareness in this market.
You and me both, but it’s not something we can directly control. We were lucky in that Blend, a newly minted west coast unicorn was willing to publish their efforts on using Kadena as well as describe how and why they switched from Ethereum to Kadena (hint: Solidity is really that bad, ETH 1.0 really can’t scale).
Daniel Dal Bello
Will, I’m really curious to know more about your background that you mentioned at JP Morgan before you left with Stuart to start Kadena. There are a lot of similarities to the Juno project that you previously worked on in comparison to what you are doing now. Juno was presented to support a new smart contract language in a private blockchain which was to build on a proof-of-work system that was scalable.
Did that experience motivate another four-letter private blockchain product you’re working on now? What motivated you to leave a position where you were working on an exciting new enterprise solution like Juno?
The lineage of our permissioned blockchain, which was for enterprise use but now that it’s FOSS [free and open-source software] it’s also for layer-2 scaling, is as follows:
- Tanagroa (BFT RAFT by Stanford Grad Student R&D project)
- Juno (Tangaroa + EVM contracts by JPM Blockchain group)
- ScalableBFT (Juno + Pact + huge performance improvements by Kadena)
- Kuro (ScalableBFT + operational/administrative features + confidentiality upgrades by Kadena)
Kuro is our permissioned ledger, and at first, was a substrate for Pact to run on. Kadena’s mainnet came about because we were advancing and solving so many problems in smart contracts with Pact’s design that we wanted to democratize the language to the public blockchain community. It was only when researching public blockchains and how to put pact on one that we discovered how to directly solve the decade-old scaling via state-sharding with proof of work, thus allowing Kadena to sidestep proof of stake’s security issues and beat every state-sharding protocol to market.
Daniel Dal Bello
Thanks, Will, interesting to hear about that transition from building the permissioned chain and researching public chains into you developing the public side too.
I recently read through a paper talking about some of the use cases you have deployed for clients that are live now. I’m interested in getting a better understanding of the interaction between both of your blockchains, the advantages of Pact, and how your consensus aligns with client needs in these scenarios. Maybe you could give us an example?
As for consensus, we see it more about being control of data privacy and administrative oversight. After Bitcoin came out, people tried to get enterprises timestamping onto it and while that works it isn’t useful enough to really drive adoption. But it had the advantage of keeping all data private.
After Ethereum came out, people (including me) tried to get EVM contracts to work for enterprise use-cases. When things deployed directly to mainnet, there were problems because bugs are all too common/impossible to avoid in EVM languages + there is so little native support for normal patterns (upgrades/governance is a contract add-on, multisig is a contact add-on) that there was little space to do actual work on the Ethereum mainnet. The boilerplate ate the available gas. By the way, this is why as DeFi spikes, we see gas fees go through the roof even though the cost of trading ETH itself is still very low. It’s that excessive, needless gas that’s wasted every transaction on functionality that should be natively supported.
So, people tried to get Ethereum to work on a private network (Quorum most notably) and sadly the EVM just isn’t up to the job. It’s general-purpose sure, but it’s not designed with industrial use in mind and ends up getting in its own way. So that effort failed, however, the 2016–2020 Enterprise Blockchain efforts educated the C-suites which is very important.
Our platform fits in by observing the successes and failures of the EVM/Solidity approach, learning from how they fail when you try getting EVM/Solidity to work in an enterprise context, and also solving layer-1 scaling. With Kadena, we’re talking about a battle-tested 4-year effort to build the network that enterprises need but is fully open source and democratized!
I did a presentation about the advantages of Pact a couple of years ago. Here’s a shortlist of some advantages of Pact:
- Human readable code on the blockchain (as opposed to inscrutable EVM bytecode).
- NOT Turing-complete. This is important for making it much easier to do formal verification.
- Important crypto primitives such as multi-sig built-in so users won’t get them wrong.
- Pact has the ability to limit how much can be withdrawn from an account in any single transaction.
Daniel Dal Bello
Could you explain in basic terms what formal verification within Pact is and what impact that has on security overall?
Certainly. Formal verification is a way of proving things about your code for all possible inputs. This last bit is the thing that can be tricky to wrap your mind around. The usual way of writing tests is to try a few values and make sure they look right. If you go to https://pact.kadena.io/example/Verification, you can see an example of our formal verification. It has a very simple absolute value function that returns the correct result for every possible input except one. But when you pass it -4839125, it returns -1 which is something that an absolute value function should never do.
Now, imagine you were trying to test this. There’s no way you’d ever by yourself test it with the value -4839125. So this bug would fly under the radar. But Pact’s formal verification analyzes the way the program is constructed and figures out very quickly that -4839125 falsifies the “result >= 0” property defined on line 19. This is super powerful and gives developers tools to make their smart contracts much more robust than anything else I’m aware of today.
Daniel Dal Bello
What is your revenue model as a business?
We’re a base-layer protocol, so the model is to change the world. We firmly believe that answering the single biggest question in crypto — can a layer-1 decentralized blockchain scale? — is ample for us to continue to drive adoption for the next decade. That said, as the platform matures, we see Kadena offering services on the platform itself (e.g. insurance) which we can leverage the project’s treasury to collateralise. Decentralisation matters, however industrial adoption will need interim centralized services to ease the onboarding path.
Daniel Dal Bello
I’d like to know a little more about your Kuro from both of your perspectives.
Will, where do you think Kuro fits into the world of enterprise blockchain solutions? We know that while both small and large firms are interested in the benefits that the tech offers that privacy is a big concern. We spoke briefly earlier about corporate giants like Microsoft, EY, IBM and of course JP Morgan all exploring and building their own tech.
With your expertise Doug I would like to know more about the technical architecture of the on-chain privacy and how your solution in this arena is positioned in comparison to other privacy solutions.
The single biggest feature is that it is a real blockchain. That means it behaves like a blockchain, i.e. it starts when you turn it on, and it is easy to administer. Hyperledger Fabric was famous for often failing to boot and being a broken database, Quorum was famous for being a mess to architect with, and R3’s Corda had a development team that rebelled and quit.
One of Kadena’s biggest problems is that, despite shipping quickly and often, despite having an incredibly reliable platform, no one writes an article about “their stuff just works”. For example, on August 20th we upgraded our sharded mainnet by 2x and it went flawlessly. This was the first time EVER that someone upgraded a state-sharded platform, but while we got press that it was happening, we didn’t get press about how boring it is to have drama-free releases. This is unlike ETH 2.0 that gets miles of press every time it gurgles. We undoubtedly miss out of the free press that failure brings, but we vastly prefer our approach to diligent release testing and accurate roadmaps.
“This was the first time EVER that someone upgraded a state-sharded platform, but while we got press that it was happening, we didn’t get press about how boring it is to have drama-free releases.”
Roughly speaking, what you can do is encrypt things client-side before you put them into a Kuro transaction, then you can decrypt them coming out the other end. This technique can be used in either a private or public blockchain environment. The Finprint work used this idea to keep financial data private and put the end consumer in control of it.
How does KDA plan to market to developers?
Glad you asked. We’ve been working on exactly that. I don’t want to spoil the details, but stay tuned in the next few weeks for a new initiative on that front. More broadly speaking, with my Director of Engineering hat on, I’d say that we plan to market to developers through engineering excellence. By providing better tools that allow developers to not get bogged down in details and enable them to focus on the important logic of what they’re trying to build.
Also, Pact is just a better language that’s easier to use, easier to learn, safer to use, and runs on the only decentralized+scaled layer-1 in existence. I think the scaled layer-1 platform will be a big selling point versus the hub-and-spoke model where developers can build an app but to deploy it they need to deploy an entire blockchain instead of just deploying a contract.
Is the team doing anything DeFi related? If so can you elaborate on some of them?
Yes, now that we’ve successfully demonstrated that we can scale, DeFi is becoming a bigger and bigger priority for us. It’s still early, so I don’t know exactly what form this will take in terms of what our engineering team works on, but we want to put in place the raw materials for a vibrant DeFi ecosystem. In my mind this starts with things like stablecoins and DEXs. The ideal thing would be to get some partnerships so we don’t have to build them ourselves. We’re already in talks with several DeFi companies that are getting crushed by Ethereum gas fees that are very interested in exploring alternatives like Kadena. One thing that will probably make it onto our internal engineering roadmap is building the crypto primitives necessary to implement decentralized bridges to various other blockchains.
“We’re already in talks with several DeFi companies that are getting crushed by Ethereum gas fees that are very interested in exploring alternatives like Kadena.”
I’m personally excited about the DeFi stuff. We are in conversations with a number of different companies. These conversations are governed by NDA though, so we can’t talk specifics. I know this isn’t a really satisfying answer, but it is nonetheless the truth.
Daniel Dal Bello
We’ve just touched on a few things like DeFi, so here is my final question. Let’s fast-forward 12 months, where do you see Kadena? What goals do you want to achieve in that time?
I feel that we’ve seen enough demand to upgrade mainnet to 50 chains, if not 100, and we’re listed on most/all major exchanges. But the single biggest change in crypto that Kadena will bring to the sector is the direct challenge it poses to the last 5 year’s narrative of “proof of stake is the only way forward.”
We have decentralized sharding, in production, and have demonstrated that is can securely scale. We’re at least 1 year ahead of our nearest state-sharding competitor, and that assumess they don’t run into all the issues that proof of stake has when you state-shard (i.e. security goes down the tube and the network degrades into a supernode architecture a la Hedera Hashgraph). This will be a massive shift for the community.
Kadena is very much in a “and yet it moves” moment in crypto.
Hillrise Group supports ambitious Web3 startups with early-stage venture capital and fundamental research.
Kadena’s public blockchain utilizes a ‘braided’ high-throughput proof-of-work architecture that improves on throughput and scalability.