Discussion on Cryptography in Blockchain
A lively discussion is currently taking place on Twitter regarding the merits of lattice versus hash-based cryptography in the context of blockchain technology. Key figures such as Eli Ben Sasson, co-founder of Zcash; Vitalik Buterin, co-founder of Ethereum; and Gabriel Shapiro, general counsel at Delphi Labs, are all actively participating in the dialogue.
Debate Initiation
The debate ignited when a16z crypto published a post advocating for the adoption of lattice-based cryptography. They argued that Ethereum’s planned architecture for signature aggregation, which is hash-only, heavily depends on SNARK recursion—an approach that comes with potential drawbacks such as increased coordination demands, bugs, and unproven security measures.
Responses and Perspectives
In response to a16z’s assertions, Ben Sasson expressed strong dissent, highlighting the vulnerabilities posed by lattices particularly in relation to artificial intelligence attacks, suggesting that these mathematical frameworks might not be as secure as believed.
Conversely, a16z crypto maintained that the use of hash-based cryptography might seem like a safer option, yet it could lead to more fragile systems. Gabriel Shapiro chimed in, backing Ethereum’s approach as a more suitable path forward. He indicated that Ethereum’s plan involves phasing out BLS signatures in favor of hash-based alternatives, specifically variants of Winternitz signatures, while also adopting STARKs for aggregation purposes. He also proposed that if lattice technology proves to withstand AI threats, it could be integrated for wallet and account functionality through a precompile.
Vitalik Buterin added his voice to the conversation, endorsing Shapiro’s perspective. He suggested that once lattice cryptography achieves AI security, it could indeed be incorporated for accounts and wallets via a precompile. Buterin elaborated on his vision for Ethereum, referencing a method he believes could effectively support various cryptographic approaches without future complications: conducting signature generation locally before aggregating them through STARKs to minimize on-chain data.
He noted that this would help circumvent the need for additional precompiles in the future and ultimately strengthen Ethereum’s cryptographic capabilities. Ben Sasson concurred with Buterin’s remarks, affirming that this direction would be beneficial for Ethereum.
Broader Implications for Blockchain Technology
In the broader landscape of blockchain technology, while Bitcoin remains the standout performer, many major platforms have transitioned to proof-of-stake models, where validators vote on blocks by signing them. This transition has been made feasible due to signature aggregation; Ethereum’s use of BLS signatures leverages an algebraic framework that can consolidate thousands of validator signatures into a single entity per block.
However, despite its advantages, BLS signatures are not resistant to quantum attacks, as no existing post-quantum signature scheme can aggregate in the same way. Earlier this year, Buterin outlined a quantum resistance strategy that includes protocol-layer recursive signatures and proof aggregation, suggesting improvements such as EIP-8141 to enable account abstraction, which would allow for the use of any signature algorithm for transaction validation, execution, and gas payment.
