Skip to content
Blockworks

The Ethereum L2 State of the Union | Day 1 | Protocol R&D

Marc-Thomas Arjoon, Harry Kalodner, Alex Gluchowski, Andrew Koller, Ben Jones

0:000:00

Key takeaways

  • Divergent Build Philosophies Converge: Despite very different starting points, all projects now emphasize EVM compatibility, rapid iteration, and a modular stack that can swap in either optimistic or zero-knowledge proofs.
  • Speed of ZK Proofs Hits Home Hardware: Alex Gluchowski unveiled Airbender, a zkSync proving system that can generate a validity proof for a typical Ethereum block in under 35 seconds on a single H100 GPU, slashing prover hardware from 160 machines to one.
  • L1–L2 Value Loop, Not Parasitism: Ben Jones argued that Layer 2 networks burn large amounts of ETH for data availability and import ETH liquidity, creating a mutually reinforcing loop rather than a zero-sum game; slow L1 scaling would be the real systemic risk.
  • Vertical-Specific L2s Are the Next Wave: Andrew Koller sees general-purpose Layer 2 launches as a fading strategy. Future chains will differentiate around native assets or user bases—centralized exchanges can move liquidity on-chain, banks can tokenize real-world assets, etc.
  • Interoperability Will Be UX-Invisible: Both the superchain (Optimism) and the elastic network (zkSync) pursue protocol-level messaging that treats multiple chains as one environment, with asynchronous Solidity patterns and intent-based execution hiding the plumbing from end users.
  • Spam Is a Pricing, Not a Tech, Problem: Harry Kalodner likened 30-50 percent spam transactions to email junk; cheap blockspace guarantees some waste, so the focus should shift to proper fee markets and MEV-aware incentives rather than heavy-handed filters.

On Choosing and Evolving Layer 2 Architectures

Marc-Thomas Arjoon opened by asking each founder to dissect hits and misses. Andrew Koller experimented with multiple stacks but discovered that even non-engineers at Kraken Inc. were comfortable in Solidity, so he abandoned exotic contract languages and adopted the OP Stack for Kraken’s superchain Layer 2. Harry Kalodner traced Offchain Labs’ seven-year arc from a 2014 white paper to a 2021 mainnet, emphasizing that neutrality—never anointing “official” apps—kept Arbitrum attractive while the team ruthlessly cut early technical debt. Ben Jones framed Optimism’s early Optimistic Virtual Machine as the classic “build the whole car first” mistake; deleting and rewriting large sections of code was painful but necessary. Conversely, Alex Gluchowski regretted not “going big” earlier: zkSync first pursued a custom language and VM, delaying full EVM equivalence and limiting adoption until zkSync Era. Each story reinforced that product-market fit for a Layer 2 is a moving target secured only by flexibility and willingness to rewrite.

On Interoperability, Account Abstraction and the End-User Experience

The panel agreed there will be no single “flip-the-switch” moment. Instead, rapid block times, home-hardware proving and native messaging will compound. Ben Jones highlighted Optimism’s work on fungible asset movement and trust-minimised messaging across OP-Stack chains, aiming for “one big instance” semantics. Alex Gluchowski argued that true seamlessness requires designing for it end-to-end: zk rollups speaking to each other can settle in roughly one second thanks to Airbender’s proofs and 100 ms block cadence, whereas bridging between optimistic and ZK systems will remain slower. Both foresee asynchronous Solidity patterns and intent-based architectures becoming standard, echoing web-2 async/await. Harry Kalodner cautioned that proof systems are “the lock on the car door,” not the engine; execution environments must remain modular so stacks can swap optimistic or ZK proofs as tech matures.

On Ethereum L1 Economics and “Parasitic” Narratives

Responding to claims that Layer 2s siphon value, Ben Jones noted that the majority of Layer 2 revenue ultimately buys calldata on Ethereum, burning ETH and reinforcing its monetary premium. He warned that if Ethereum stops expanding data capacity—through larger blobspace or higher gas limits—both layers suffer from rising fees. The real debate, therefore, should focus on shared decentralisation norms, ETH usage in monetary policy, and equitable fee splits, not on zero-sum rhetoric often pushed by rival L1s.

On Spam, Blockspace Pricing and Network Health

A Flashbots study put junk traffic on some Layer 2s at up to 50 percent. Harry Kalodner downplayed the alarm, comparing it to email: when marginal cost approaches zero, marginally useful or even useless data will fill the gap. The answer is dynamic gas pricing, MEV-aware auction design and careful block-gas limits—not draconian censorship. Better fee markets will matter even more as ZK proofs drop the cost of chain execution and invite higher transaction velocity.

On Centralised Exchanges and Real-World Assets

Andrew Koller sees exchanges launching proprietary Layer 2s as a necessary self-disruption. Liquidity, order flow and existing user relationships become the niche differentiator in an increasingly crowded rollup field. He confirmed that Kraken Inc. plans equity tokenisation with partner X-stocks and hinted at mortgage-backed tokens as a personal priority. Alex Gluchowski framed tokenised RWAs as the bridge to convert ETF-only investors into on-chain participants; zkSync already hosts about two billion dollars in such assets and is running pilots with Deutsche Bank and several sovereign entities. All panelists agreed that regulatory clarity in bills like the GENIUS Act could accelerate this migration.

Key actionable insights

  • Track real-time proving benchmarks: consumer-grade ZK proving could collapse hardware costs for enterprise rollups within a year.
  • Monitor L1 data-availability roadmaps; insufficient blobspace threatens both L1 fee stability and L2 business models.
  • Expect specialised, liquidity-anchored Layer 2 launches from major fintechs and exchanges, each paired with creative token incentives.