Explore a novel approach to consensus inspired by CRDTs and category theory in this conference talk from Strange Loop 2022. Delve into the Convergent Proof of Stake consensus algorithm and its implementation for efficient real-time decentralized computation using Convex Lisp. Learn how immutable persistent data structures and digital signatures enhance security in blockchain technology. Discover how this innovative method addresses the shortcomings of traditional blockchain, offering improved scalability and efficiency. Follow along as the speaker, Mike Anderson, demonstrates the practical applications of this technology, including voting systems and smart contracts. Gain insights into the structure of global state, data types, and the concept of treating every account as a Lisp machine. Understand the economic considerations necessary for implementing this system and how it compares to existing blockchain solutions.
Overview
Syllabus
Intro
Some Requirements
Convergence example
Three requirements for a CRDT merge function
Some examples of CRDTS
Ideally we want universal state transitions
Voting Example
One solution is Blockchain
Not a good solution
What happens if somebody adds a new block?
Fault Tolerance
What if blocks conflict?
Simultaneous confirmation
We can now define a merge function!
One big problem?
Data structures to the rescue
Optimising Beliefs
Convergent Proof of Stake is close to optimal
Convex Virtual Machine
Structure of Global State
Data Types
Every Account is Lisp Machine!
Smart Contracts - Vote Counting Actor
Smart Contracts - Usage from User Account
A little necessary economics
Taught by
Strange Loop Conference
