“Arbitrum was not created as an ‘Ethereum service,’ but because Ethereum provides high settlement security,” he emphasized.
Goldfeder shoots down claims that a beefed-up mainnet will eclipse L2 throughput. Think back to those wild days when Arbitrum and Base hammered over 1,000 transactions per second, while L1 idled far below. Insider tip: monitor Dune dashboards during NFT drops—L1 TPS rarely breaks 50 sustainably, but L2s stack calldata efficiently post-Dencun.
He warns Ethereum risks pushing big institutions to spin up rival L1s if it turns unfriendly to rollups. Nuance here: institutions crave Ethereum’s liquidity but bolt at 10x fees; one bad block explorer spike, and they’re eyeing Solana forks. Mistake to avoid: ignoring sequencer failover tests—downtime kills trust faster than gas wars.
Optimism’s Karl Floersch: Decentralization Hurdles Are Real
Karl Floersch from the Optimism Foundation backs the decentralization drive but flags the gritty barriers.
Withdrawals drag for weeks due to challenge windows, Stage 2 proofs aren’t plug-and-play yet, and cross-chain tools feel half-baked. Floersch likes Buterin’s nod to native precompiles for easier verification—saves cycles on SNARK checks. From experience, precompiles cut prover latency by 30% in my stacks; without them, ZK syncs timeout under load.
Why the delay? Provers guzzle GPUs, and fault proofs need adversarial testing that sims rarely catch. Tip: run your own Stage 1.5 fork—exposes custody risks sequencers hide in prod.
Base’s Jesse Pollak: Beyond Cheap Ethereum
Jesse Pollak, leading Base, echoes that L2s can’t just mimic low-fee Ethereum.
Base prioritizes onboarding devs with account abstraction and privacy hooks—features L1 lags on. Practitioners know: AA slashes UX friction; one-click wallets convert 2x better in A/B tests. Privacy layers let apps handle regulated data without full zk overhead.
Assumption to challenge: L2s as Ethereum clones. Reality? They’re app-specific chains, inheriting security while customizing gas for games or AI inference.
StarkWare’s Eli Ben-Sasson: ZK Already Fits the Bill
Eli Ben-Sasson, StarkWare CEO, kept it short, implying ZK rollups like Starknet match Buterin’s ideal.
“Say Starknet without saying Starknet,” he wrote on X.
ZK proofs bake in decentralization from day one—no sequencer bottlenecks. Hands-on: compiling Cairo feels clunky at first, but recursive proofs scale vertically where optimistic waits for disputes.
Broader Ecosystem Shifts
These responses show consensus on progress but divergence on priorities. L2s evolve by niching—gaming on high TPS, privacy for finance. Buterin’s Feb 1 idea for creator incentives
From the field, the real edge lies in hybrid stacks: L2s with L1 precompiles for proofs, plus offchain DA for governance. Mistake devs make: chasing Stage 2 perfection before Stage 1 stability. Why? Users pay for speed now, security later. Ethereum’s scaling mosaic thrives on this tension—L1 base, L2 fireworks.
This debate lights a path: L2s won’t fade; they’ll specialize, securing Ethereum’s multi-chain future.
