Aparna Krishnan is head of education at Blockchain at Berkeley and co-founder of Mechanism Labs, an open source blockchain research lab. Earlier this year, Aparna was awarded a scholarship by the DFINITY Foundation for Mechanism Labs’ research into consensus mechanisms. This episode is essentially a primer for advanced discussion of consensus in decentralized networks.

https://mechanismlabs.io/https://github.com/aparnakrhttps://twitter.com/aparnalockedhttps://medium.com/@aparnalocked

Aparna KrishnanCo-founder mechanism labs open source research labAll work and research is on GithubTelegram: @mechanism_labsCo-founder of the education team at Blockchain BerkeleyConsensus researcherTeaches executive education courses

ConsensusProof of WorkProof of StakeOld Field

HistoryCynthia Dwork developed stronger adversarial modelsDid not have many applicationsBlockchain has brought cryptography and consensus into the mainstream eye

ProtocolsDFINITYTendermintBitcoinFocus has been on proof of stake protocolsMining may not be sustainableLong term sustainabilities and lack of externalities is importantProof of stake offers efficientlyAs does proof of elapsed time and proof of space and time

ConsensusSybil controlComing to agreement relies on traditional consensusPoS, PoW refer to the sybil control mechanismPoS - Putting down capitalFinancial penalty for misbehaviorToken holders are participantsPoW - burning energyOne cpu, one voteNo connection between token custody and rewardsP o Elapsed Time Proof of Authority placing reputationA cost of playing ballTraditional Consensus (PBFT)No concept of probabilistic finalityAll honest nodes come to final agreementClosed, permissionedBlockchainAll nodes may with a high degree of probability agreeA probabilistic guarantee

Longest Chain RuleLongest sequence of blocks is the EthereumGhostCan be attacked by a stealthy entityOpen, permissionless, decentralizedInefficientNode churnBetter liveness properties

Early PoS 2013PeerCoinNXTBitsharesPrimitiveState grinding attack vulnerableRandomness derived from blockhashes

New GenerationDFINITYUses threshold relay decentralized randomnessTendermintRound robinHas social layer fallbackEthereum CasperThunderellaAlgorandHides the block creator until the block has been created

In Tendermint a minority below the assumption cannot break finalityIn Bitcoin >10% can change the finality through threatening to censor

Hard to get both economic and cryptographic security. Not clear that bribing and collusion attacks have been solved. No protocols have shown sophisticated economic models.

Mechanism LabsFocussing on incentive schemesWhat does it mean to have a stable, scalable protocolScalability

Proof of ReplicationRecent work with Verifiable delay functionsIn commit/reveal schemes there is an opportunity to manipulate by giving only one person the ability to reveal the randomnessVerifiable delay functions allow anyone to reveal the randomnessASIC resistance because not parallelizable