# MACI: 2024-04-22 ## Resources - [MACI roadmap](https://maci.pse.dev/roadmap) ## Agenda - MACI plans - Short term vs. long term - Questions for V - Thoughts on technical direction? Research ideas? - Opinion on strategic approach? ## Notes ### MACI - What's hurting adoption? - It's hard to run - Why? - Should just be running a Python program - Technical hurdles - Getting non-crypto people using it - Currently: every vote is an on-chain tx - V idea: Plasma technique - Operator. Relay message to operator. Operator publishes batches. SEnd proof that you're included in the list. As a user 1st resort: send to coordinator, 2nd resort: post directly on-chain - Adds back coordinator complexity. Requires online component (running a daemon vs. just scripts) - El gamal - https://github.com/privacy-scaling-explorations/maci/issues/796 - Risk of creating more UX friction at the cost of more privacy - User should only need to do 2 things (1 register, 2 vote, maybe 3 to check if vote was included) - Coordinator - MACI properties (current) - Correctness (snarks) - Censorship resistance (on chain) - privacy (depends on coordaintor) - collusion resistance (depends on coordinator) - Next gen - Correctness (snarks) - Censorship resistance (on chain) - privacy (now able to guarantee) - collusion resistance (depends on coordinator) - Can we ever remove this? FHE or MPC - W/ FHE, anyone could run the tallying computation. Overhead w/ FHE & SNARKS. Threshold for key generation. Threshold is final decryption. Set n of m for whatever you want. If it fails for not enough decryption, - Don't need prover in FHE. Just need verifier. From a dev standpoint, not super hard. Just an overhead question. - Talk to Auryn about his project - Can we use TEE for coordinator? - https://github.com/privacy-scaling-explorations/maci/discussions/1385 - V: def good. Are you able to run a zkSNARK prover inside TEE? If yes, that's the best you can get until FHE - Single MACI instance (for keys across instances) - challenge is more pressure on the gatekeeping / proof of personhood - Upcoming technical roadmap - See where that's going - V talking to Barry - Wants to start doing things outside PSE - Barry: Early stage work, algorithms, - V POV: excited about this for a long term. cryptographic identity, voting, reputation. See this actually happen. Starting to feel the impatience. Time to shift gears & create things that make impact - within & outside of our space. e.g. zkemail, anon aahaar, one way that'd be useful: my coins in my 4/6 multisig. Some gov't ID is one of my guardians. For that to happen, that service needs to be simple, anyone can download in 1 click. Plug in ID & get ETH address. take address & plug into Safe. Or fork safe UI & specify email address / gov ID / whatever - Barry feedback: current PSE. Setup for early stage protyping. Not as well suited for actually brings things to "next level". - Strategy needed? - Picking an expertis pool? e.g. zk for privacy vs. zk for scalability. zkevm is nerd-friendly. - Summa vs. proof of passport. Different things - e.g. Summa - idea has been out since blog post (really 10 years). what is a good outcome? we want CEX to actually use this. Current work doesn't do that. Why not? current team is lacking that skill. Or lacking - 3 buckets of use cases - Identity - TLSN - Voting - Overlap/integration w/ 0xPARC - Separate convo? - They're doing Zupass. It's amazing. Need to make sure it's an open standard in all the right ways. Making sure it feels like a natural part of the ETH ecosystem. Worry that things trend towards that. The fact that it doesn't feel signifcantly easier to use Zupass & EThereum together vs. seperately - e..g tix. ETH thing becasue you can buy on chain. But also Zupass b/c if you have an off-chain ticket you can proof it. Ideally these things work together. Can have NFT ticket & prove immediately without a server - V: would love to see attention towards real stuff out there in practice. Getting things out in use. DAO adoption. ## Action items - [ ] Sam to touch base with Auryn. He's working on FHE stuff - [ ] Check out his paper: https://docs.google.com/document/d/1N_G5BH7klvFV3vdwe4ZRuoVoFNYsVq7ZU_1BOBHxN4A/edit?usp=drivesdk - [ ] V to post about Plasma approach (off-chain voting) on ethresearch - [ ] Sam to detail thoughts on PSE roadmap/strategy. Share with V & keep in the loop