Pinned post
Alejandro
Verified account Sei@alranpe
683 Following 643 Followers 1.3K Posts
Latest posts 5
This applies to not just math
I think this slide from Terence Tao perfectly captures the future of mathematics with AI assistance.
And I strongly believe the same transformation is coming to cryptography, Ethereum, and protocol research more broadly!View quoted post on X
Secret single-leader election is hard. An adaptive adversary can crash the leader in single-leader consensus if the leader is not hidden.
A very interesting result by @ittaia et al. fix this with Multi-Leader Fixed-View Simplex: k leaders propose, everyone votes on the set of proposals they saw, then a threshold beacon reveals a random priority order after voting and each vote resolves to its highest-priority proposal. This is kind of similar to what we did at what we then called asynchronous Granite back at Protocol Labs (and ended up being psync GossiPBFT).
Two leaders already give you ≥½ chance of finalizing a good view against one mobile crash (expected ~6Δ). Structure stays basically Simplex.
They also check the rational case (leaders delaying for MEV). Second leader drops view-failure rate from ~1% to 0.01% at equilibrium.
arxiv.org/pdf/2611.01323
A very interesting result by @ittaia et al. fix this with Multi-Leader Fixed-View Simplex: k leaders propose, everyone votes on the set of proposals they saw, then a threshold beacon reveals a random priority order after voting and each vote resolves to its highest-priority proposal. This is kind of similar to what we did at what we then called asynchronous Granite back at Protocol Labs (and ended up being psync GossiPBFT).
Two leaders already give you ≥½ chance of finalizing a good view against one mobile crash (expected ~6Δ). Structure stays basically Simplex.
They also check the rational case (leaders delaying for MEV). Second leader drops view-failure rate from ~1% to 0.01% at equilibrium.
arxiv.org/pdf/2611.01323
I love consensus analysis in the context of (k,t)-robust equilibria. In a future where validators must be KYC-ed (to prevent bots), more safety cannot come throwing more virtual entities to a protocol that can handle them, it comes from safer protocols.
Do you know what STAKE stands for in the context of blockchain? Secure and Tolerant Algorithm through (K,t)-robust Equilibrium (STAKE), a new blockchain algorithm that does not suffer double spending attacks despite a coalition of 2n/3 hackers.View quoted post on X
“SoK: Multiple Concurrent Proposer Blockchains”, with @muurshb and @MaxResnick.
We compare 39 MCP designs to understand how their choices affect inclusion, ordering, hiding and performance. Here are some of the main observations. 🧵
We compare 39 MCP designs to understand how their choices affect inclusion, ordering, hiding and performance. Here are some of the main observations. 🧵
