Nov 8, 2021 · 46m · a16z
Scaling Up Blockchains with Zero-Knowledge Proofs
Alex Gluchowski (Alex G) · 21m spoken Ali Yahya · 14m spoken Dan Boneh · 4m spoken Alex Vlasov (Alex V) · 2m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this a16z panel discussion, host Ali Yahya and Stanford Professor Dan Boneh join Matter Labs co-founders Alex Gluchowski and Alexsandr Vlasov to explore how zero-knowledge proofs, zkSync, and zkEVM solve Ethereum's scalability bottleneck while preserving mainnet security guarantees.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 4.2 Guest teaching 3.1 Guest disagreement 0.9 The host pushing back 0.3
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 25:33 Probing asset security trade-offs
Dan Boneh directly puts guest Alex G on the spot by asking where he would personally store an expensive asset like a CryptoPunk, forcing an explicit admission of security trade-offs between ZK-rollups and ZK-Porter.
Hardest push from the host ▶ 36:59 Pivoting away from host framingDan Boneh explicitly redirects host Ali Yahya's multi-part outline on industry trends to demand a technical explanation of zkEVM circuit mechanics instead.
Biggest teaching moment ▶ 11:00 Correcting the ZK privacy misconceptionAlex G reframes a foundational premise about zero-knowledge technology, clarifying that scalability relies strictly on succinctness rather than privacy properties.
The host holds their own ▶ 43:49 Host analysis on language network effectsHost Ali Yahya demonstrates sharp domain authority by explaining how developer toolchains and language network effects make Solidity the indisputable lingua franca of Web3.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Title Card and Important Legal Disclosures | 6 | 0 | 0 | 0 | Host Ali Yahya opens with an extensive monologue setting up the thesis for blockchain computing and technical context around Ethereum throughput bottlenecks. He demonstrates solid domain knowledge regarding transaction limits and computational paradigms without guest interaction. | |
| Alex Gluchowski on Matter Labs' Genesis and Mission | 1 | 2 | 1 | 0 | Alex Gluchowski details his personal background in post-Soviet Ukraine and the founding vision of Matter Labs. The host passively listens while the guest frames economic sovereignty and scaling priorities. | |
| Demystifying SNARKs: Succinctness vs. Zero-Knowledge Privacy | 3 | 5 | 2 | 0 | When asked how he discovered zero-knowledge proofs, Alex G gently corrects a major industry misconception, explaining that scalability relies on succinctness rather than zero-knowledge privacy. The host acknowledges and accepts the distinction smoothly. | |
| Dan Boneh and Ali Yahya on the Math and Mechanics of SNARKs | 6 | 2 | 0 | 0 | Dan Boneh explains SNARK soundness and proof succinctness. Host Ali Yahya actively extends the technical framing by detailing validator computation redundancy and how single provers solve the 20,000-node overhead problem. | |
| zkSync Roadmap and the Achievement of zkEVM Compatibility | 2 | 4 | 1 | 0 | Alex G outlines the zkSync roadmap, highlighting how engineering breakthroughs surpassed initial skepticism from industry figures like Vitalik Buterin regarding generic ZK-rollups. The host acts purely as a prompt for the guest's progress update. | |
| Architecture of a ZK-Rollup and Inheriting Mainnet Security | 3 | 3 | 0 | 0 | The guest explains ZK-rollup architecture using the Yap island Rai stones analogy for state commitments, while Dan Boneh emphasizes cryptographic soundness guarantees. The exchange is highly collaborative and educational. | |
| Comparing ZK-Rollup (zkSync) vs. ZK-Porter: Cost vs. Security Trade-offs | 6 | 4 | 2 | 2 | Dan Boneh presses Alex G on where he would personally store a valuable CryptoPunk asset to probe security boundaries. Alex G delineates between ZK-rollup security and ZK-Porter's proof-of-stake data availability trade-offs, which the host then synthesizes effectively. | |
| The Engineering Breakthrough Behind the zkEVM Architecture | 4 | 6 | 2 | 1 | Dan Boneh bypasses the host's high-level topic overview to ask directly about the internal engineering of zkEVM. Alex G delivers a deep technical breakdown on TinyRAM circuits, recursion, and overhead separation. | |
| The Strategic Necessity of EVM and Solidity Compatibility | 7 | 2 | 0 | 0 | Alex G explains why Matter Labs pivoted from Zinc to Solidity compatibility. Host Ali Yahya strongly reinforces this point with expert analysis on language network effects, toolchain maturity, and developer ecosystem lock-in. |
Statements from this episode (16)
Yahya: Ethereum can only process about 14 transactions per second
“It currently can only really process around 14 quote unquote transactions or 14 kind of, you can think of them as computations per, per second.”
Gluchowski: Blockchains will disintermediate global value exchange like the internet did information
“The same thing that internet has done to information, I think blockchains will do to the world of value, to the network of pages. We can exchange value directly without intermediaries overcoming the trust issues between people without involving third parties, …”
Gluchowski: Hyper-Bitcoinization based solely on Bitcoin is utopian and impractical
“The idea of hyper Bitcoinization just based on Bitcoin is kind of utopia because of a lot of usability issues and a lot of issues with just general lack of functionality, which you need for a real financial system.”
Gluchowski: Blockchain Scaling Relies on Succinctness, Not Zero-Knowledge Privacy
“It was about zero knowledge aspect of zero knowledge probes, which we actually paradoxically do not use in scalability.”
Gluchowski: zkSync 1.0 Has Processed Over Four Million Transactions
“We have processed over four million transactions.”
Gluchowski: 99% of Gitcoin Grant Transactions Processed via zkSync
“Almost all the Gitcoin grant runs have been run on the, on, on ZK Sync with 99% of transactions going through ZK Sync. Rather than on mainnet.”
Gluchowski: Matter Labs Deployed Uniswap V2 Port to zkSync Testnet
“We just launched our first fully ported EVM that last week on it. It's the port of Uniswap version two”
Gluchowski: ZK-rollup security is roughly equivalent to Ethereum mainnet
“So in the roll up your security is roughly equivalent to the security of Ethereum.”
Gluchowski: Optimistic rollups cannot support Validium extensions due to game theory reliance
“There's another Type of roll-ups called optimistic roll-ups. They unfortunately cannot do this. They do not rely on math and cryptography. They rely on game theory and validators, and they cannot have this extension.”
Gluchowski: ZK-rollups will maintain a 50x cost reduction ratio over Ethereum mainnet
“So right now the rollups we have for ZK rollups we have for payments are about 50 times cheaper than native payments on layer one. And I expect that this ratio to remain roughly the same for also for smart contracts.”
Boneh: Matter Labs achieved zkEVM capability decades ahead of expectations
“Compiling Uniswap V-II solidity into you know, a snark capable circuit. We thought this was going to be decades away and you got it to work.”
Gluchowski: zkEVM separates program logic from heavy operations via recursion
“So the trick that we used in ZKVM to make the EVM compatible ZKROLA possible Was to separate the tiny RAM circuit, which is responsible only for the program logic from the circuits, which are optimized for doing heavy operations, which are not generic. They're…”
Gluchowski: Recursive ZK proof verification became affordable on Ethereum in 2020
“And recursion is something that became available and became affordable on Ethereum just last year. Before that, it was very expensive. We had different proof systems that supported recursion, but they were way too expensive to be used on Ethereum.”
Gluchowski: Matter Labs Created Custom Language Zinc Before Pivoting to EVM
“We actually started originally with building a custom language, which was optimized for zero knowledge pros. It was called Zinc. But we quickly realized that it will be really hard to onboard the existing ecosystem of Ethereum and let everyone rewrite their pr…”
Gluchowski: Alternative L1 Blockchains Failed to Gain Traction Without EVM Compatibility
“We see it with alternative L ones. We have a couple of Ethereum killers. Having launched in in the previous year with custom systems, with different virtual machines and different languages and they didn't really get traction until they implemented Ethereum co…”