Charles Hoskinson, Vitalik Buterin Duke It Out Over ZK-Proofs, Ethereum Design

A technical discussion on X (formerly Twitter) between Ethereum co-founder Vitalik Buterin and Cardano founder Charles Hoskinson quickly blew up into a wider critique of blockchain design, governance, and the practical, real-world viability of zero-knowledge (zk) proof systems.

It all began with a mathematical musing from Ethereum co-founder Vitalik Buterin, who explored the theoretical efficiency of distributing zk proof workloads across multiple provers.

A fun math aside, on the idea of splitting a large zk proving workload between multiple provers.

Suppose you have N provers, and you have a proving workload that you split into N parts (so, one part per prover). You require provers to pre-register, but registration is…

— vitalik.eth (@VitalikButerin) May 21, 2025

Buterin proposed that by using open-access registration and fault-tolerant retries, the system could achieve verification in logarithmic star time–a function that grows incredibly slowly.

Buterin’s ZK-Proof Math Sparks Hoskinson’s Real-World Rebuttal

While the math was elegant, Charles Hoskinson quickly pounced on its real-world impracticalities, calling the model naive in the face of adversarial attacks, cloud failure risks, and hardware bottlenecks.

Hoskinson didn’t just pick apart the math; he used the opening to paint a bigger picture of what he views as Ethereum’s systemic flaws. He argued that Buterin’s proposed retry model for zk-proofs would crumble when faced with Sybil attacks, where bad actors could spin up thousands of fake provers just to jam the network.

@VitalikButerin I’ll bite. Your log*(N) retry model stumbles on real-world constraints. Open registration lets an attacker spin up thousands of provers at near zero cost, turning your 20 % failure rate into an adversarial control lever that can knock out any shard on demand.…

— Charles Hoskinson (@IOHK_Charles) May 22, 2025

This, he explained, is because the hefty hardware demands of actual zkEVM circuits make a super-quick three-second retry loop unworkable. He also red-flagged the risk of network self-sabotage, where escalating retries could flood the system with unneeded bandwidth and computational burdens.

Cardano vs. Ethereum: Governance and Layering Philosophies Clash Again

This isn’t the first time Hoskinson has positioned Cardano as a more grounded alternative to Ethereum.

Back in September 2024, he criticized Ethereum’s governance, comparing it to a “dictatorship” and emphasizing how Cardano’s Voltaire-era overhaul avoids the extremes of both Bitcoin’s minimalism and Ethereum’s founder-centric approach.

Buterin, on the other hand, has not remained idle. Facing growing scrutiny over Ethereum’s reliance on Layer-2 networks like Arbitrum and Optimism, he recently proposed a bold Layer-Zero upgrade.

This next-generation architecture would replace the Ethereum Virtual Machine with a zkVM based on RISC-V standards, dramatically improving efficiency.

According to preliminary estimates, the upgrade could slash proof sizes from 346MB to just 1.5MB, reduce computation cycles by over 95%, and boost transaction throughput up to 30-fold using GPU acceleration.

Yet, Hoskinson remains skeptical of Ethereum’s direction. He argues that an overreliance on Layer-2s siphons value away from the base layer, risks fragmentation, and leaves Ethereum vulnerable to being overtaken by more robust alternatives or even Bitcoin.

He has gone so far as to claim Ethereum may not survive the next 15 years if it continues along this path.

Source