ZK Summit12 - Lisbon Oct-24

I attended the Zero Knowledge Summit 12 in Lisbon. But, should I have gone, as someone who doesn't have a technical background? This post will summarise some of the talks and hopefully convince you to go to the next one!

One of the first talks was on circuits and showed equations with xA, vkA, argsA etc. I hadn't done Maths since my A-Levels. I didn't understand much. But, no, I hadn't made a mistake in attending. How do you learn if you don't know what you don't know?

The header shows my NFC card, which allows event attendees to share personal information (social media etc.) with each other.

The card was created by @cursive_team & @zkv_xyz. @andrewlu_ & @viv_boop from the Cursive team presented their thoughts on social connections:

This extra connection is needed because people are becoming lonelier.

Allowing people to share personal information privately, is the next step in IRL event and digital connection.

The concept of digital pheromones is fascinating. Like a digital smell, which is similar to pheromones in animals. These refer to the signals we can share with attendees. The signals can refer to personal information, e.g. if you think zkVMs are underrated or overrated.

@schaliasosvons presented his research - analyzing and benchmarking ZK-Rollups. His presentation was focussed on zkSync Era and Polygon zkEVM. Full research paper is here - stefanoschaliasos.github.io

But, why don't we just use Optimisitc Rollups? Optimistic Rollups assume all transactions are valid. A fraud proof needs to be submitted onchain, to prove the transaction is invalid. To give time for this, the time to finality is 1 week. All data is published too.

ZK Rollups avoid this delay by using validity proofs, resulting in time-to-finality of a few minutes (not 1 week). The cryptographic proof is verified onchain (without re-execution). The data is compressed by using state-diffs or succinct proofs (not all data). 0. Pending transaction (tx) - ZK Rollups are used when Layer (L2) users interact with an application and submit a (tx) to a sequencer, which creates blocks and batches them into a single batch. 2. Pre-confirmed tx - The tx is executed and pre-confirmed in the L2.

3. Committed tx - The data is then committed in the L1 (Ethereum) and the L1 publishes data of the batch in the L1 (Data Availability Layer). The L2 sends a batch to a prover, which then creates a succinct proof of the batch.

The proof is submitted to the L1 smart contract which will run the verifier and verify the correct execution of the tx's (Settlement Layer). 4. Tx's are now verified and finalised.

@0xPolygon zkEVM & @zksync Era both use two proof systems (one for the initial proving and another, smaller one for submitting to the L1). Polygon zkEVM submits Tx data, whereas zkSync Era publishes State diffs.

zkSync Era uses huge batches, resulting in a higher cost per batch ($18.37), but much lower cost per tx ($0.0048). Polygon uses smaller batches, then aggregates multiple batches into a single proof. The cost per batch is lower ($1.34), but higher per tx ($0.0511).

The penultimate talk was by @Zac_Aztec - presenting on @aztecnetwork, a privacy preserving programmable L2 for Ethereum. This project is 7 years in the making and mainet is set to go live! Aztec's aim is to have user privacy, data privacy and code privacy (of smart contracts).

The final talk was with @AnnaRRose & friends. @aisconnolly left a great thought on the end game for ZK. Is it when ZK is trusted by Granny, or when the government trusts ZK passports? Will Granny trust the government or ZK?

Finally, special thanks to @CelestiaOrg & @payy_link for the after party. Night time views over Lisbon were stunning!

Many thanks,