Earlier this month on May 9, 2024, Head of Institutional Research at Uphold, Dr. Martin Heisboeck interviewed Kaspa founder Dr. Yonatan Sompolinsky in the Uphold Institutional x Kaspa webinar. This discussion occurred ahead of the Rusty Kaspa upgrade that went live on May 13, 2024. It is the second time that Dr. Heisboeck has interviewed Dr. Sompolinsky.  

Dr. Heisboeck opened the interview with a story about purchasing a Kaspa branded sticker from a market vendor in a small town in Bohemia. This anecdote served as an illustration of Kaspa's remarkable growth and popularity over the past year. The discussion transitioned to fundamental aspects of the Kaspa network, ranging from general background on blockchain technology to Kaspa’s past and future updates.

The topics discussed include:

Blockchain origins - Dr. Sompolinsky described how issues inherent in online transactions resulted in the creation of blockchain, with the underlying task of consensus protocols being to deal with asynchron. He stated: 

"If you think this through, the entire double-spend problem stems from latency. The entire need to run consensus in the first place stems from asynchrony because two events can happen at different places where the two respective servers or recipients are unaware of the other. So, it's a basic fact that consensus is how we deal with asynchrony. How do we reconcile two conflicting events that happen in different factions of the network? And this is characteristic of consensus protocol[s], and the properties of consensus protocol are good or bad depending on how it reacts or corresponds to network latency."

During the webinar, Dr. Sompolinsky’s internet connection was unstable, forcing him to dial-in on his phone. At this humorous moment he emphasized that the internet is unreliable, which further pushed the significance of internet latency.

Bitcoin ignores the problem of asynchrony simply by adopting a synchronous time model. Synchronous models are not ideal. In his article, The Master of Time, Nicholas Sismil describes "Synchronous models are overly strict and inefficient at modeling the real world without harming safety (e.g., with network outages and attacks of unexpected durations)."

In contrast, Sismil describes asynchronous protocols as follows:

“Asynchronous models attempt to overcome the problems of synchronous models by assuming nothing about network delays — there are no bounds on network latency, even between correctly functioning nodes. The pros are twofold: first, without a time-bound, message delays cannot cause unexpected harm to safety; second, since there are no fixed values for timeouts, nodes adapt to the actual latency of the system.”

Bitcoin Limitations - Kaspa was created due to the limitations of Bitcoin and how it does not adjust to real-world latency, due to its long block interval of about eight minutes. Dr. Sompolinsky’s goal for creating the DAGKNIGHT protocol for Kaspa was to design a proof-of-work consensus protocol that was able to handle transactions with higher block rates and faster confirmation times according to real-world internet latency in a way that is superior to Bitcoin. 

Rusty Kaspa Update - Over the past year and a half, the Kaspa code was rewritten in Rust (known as Rusty Kaspa) to optimize and streamline the architecture, as well as utilize the capacities of the Rust programming language to increase the block rate. Developer Michael Sutton led this initiative and the Rusty Kaspa codebase is currently running on mainnet on most of the nodes as of the webinar. As of May 15, the code has been upgraded and the majority of nodes are now running Rusty Kaspa. 

Why Kaspa is Better than BTC - All blockchain projects have to have a tradeoff with what is known as the trilemma - decentralization, security, and scalability. Kaspa Kaspa not only solves the trilemma, but also solves a challenge that was once deemed impossible — the  ability to function as a partially synchronous protocol within a proof-of-work system. The Kaspa DAGKnight consensus protocol self-adjusts to network latency. For example, when the network's adversarial latency is low, DAGKNIGHT's orderings converge immediately, allowing clients to confirm transactions within a few Round Trip Times. On the other hand, when the network is slow and clogged, the orderings converge more slowly, and thus, allow the transaction confirmations to take longer. 

Kaspa’s MEV Solution - Increasing the block rate enhances the security of the network. In most blockchains, there is a monopolistic control of each block. Entities compete to control the round to capture the Maximum Extractable Value (MEV) that occurs between the transaction=0 (not occurring) and transaction =1 (occurred). This is known as algo (or algorithm) trading in traditional finance. However, in traditional finance it is an assumption that the transaction processor is being honest, but in decentralized finance, it is up to the code to provide this assurance, as there is no gatekeeper. 

High Block Rate - Kaspa is currently working on research to reach an equilibrium where the miner kickback matches the MEV through increased blocks per second. This is achieved due to increased miners' competition for the transaction. While this does not prevent MEV, it ultimately creates a wash as it equals the kickback. Increased block rate also reduces the desire for miners to join mining pools and run their own nodes, increasing decentralization. The goal for Kaspa is to be mining 100 blocks per second (BPS). So if a miner owns 1% of the hash rate, at 100 BPS, then the miner would mine 1 BPS. Dr. Somplinksy expects that most miners will be service providers (like a smart contract wallet) processing transactions for their user base. 

Kaspa and ZK - Due to Kaspa's robust codebase, developers will only need to add 1-3 operation codes (opcodes )to a clone of the codebase to add additional features such as zero-knowledge proofs (ZKPs). This should happen when there is user need and demand, like in a decentralized finance (DeFi) ecosystem

Kaspa’s Sequencing - Blockchain technology is only useful for solutions requiring an immutable ledger of transactions. Finance stands out as the prime example of such a use case, although it would also be beneficial in applications requiring order sensitivity. Hence, Kaspa concentrated its efforts on refining sequencing algorithms to maximize the foundational capabilities of blockchain technology. 

Stablecoins on Kaspa - Dr. Somplinksy stated that there will be a Kaspa stablecoin in the future. The team is currently researching and reviewing the following two questions related to stablecoins: 

Oracles related to Stablecoins - When thinking about stablecoins on Kaspa, it is important to consider which entities report the prices to the network. In decentralized finance (DeFi) applications, it does not matter how well-developed the blockchain is if the pricing oracles are faulty. For instance, what happens if the oracle reports an incorrect price? Also, if the stablecoin needs to be liquidated, which agents get liquidated, and in what order? 

In conclusion, the interview between Dr. Somplinksy and Dr. Heisboeck provided invaluable insights into the evolution of blockchain technology, Kaspa's innovative contributions, and the potential implications for future developments. Dr. Somplinksy quoted Naval Ravikant during the interview, 'Impatience with actions, patience with results,' to emphasize the importance of perseverance and strategic vision in navigating the complexities of technological advancement. With Kaspa's focus on refining sequencing algorithms, addressing challenges such as blockchain latency, and exploring decentralized finance, stablecoins, and other innovations, it is clear that Kaspa is poised to play a significant role in shaping the landscape of tomorrow.

References and Resources: 

[1]. Original link to the webinar from Uphold Institutional. Video reposted with permission to the Kaspa official Youtube channel here. 

[2]. For additional information on synchronous vs. asynchronous consensus protocols, please Nicholas Sismil’s The Master of Time. How DAGKNIGHT Solves an Impossibility Result Unachievable by Bitcoin, Ethereum and Classical BFT Models

[3]. Image from the official Uphold twitter account; post. 

[4]. Additional sources: https://iang.org/papers/BitcoinLatency.pdf