Currents 094: Matthew Pirkowski on Blockchain Consensus Mechanisms

Subscribe: Spotify | RSS | More

Jim talks with Matthew Pirkowski about the kinds of consensus mechanisms that can be used to secure blockchains. They discuss active inference, proof of work vs proof of stake & the relationship between them, auto-catalytic networks, proof of work in emergent nature, what consensus means & why it needs to be protected, integrity of the ledger, an analogy with clocks, accelerating entropy, photosynthesis, exploring vs exploiting tensions in emergent systems, coordinating central points of reference, energetic openness, the relationship between energy & information, resistance to manipulation, postmodernity & symbols untethered to reality, the evolution of evolvability, adaptive drift, a stable foundation for building infrastructure, the tight relationship between information theory & thermodynamics, whether existing cryptocurrencies exist in a Goldilocks zone vs an arbitrary spot in design space, bugs of global reserve currencies, whether investing in Bitcoin is an anti-social act, currency vs wealth, personal stores of abstract potential energy, and much more.

Matthew Pirkowski works at the intersection of software, psychology, and complex systems. These interests first took root while studying Evolutionary Psychology and assisting with Behavioral Economic research at Yale’s Comparative Cognition Laboratory. From there Matthew began a career in software engineering, where he applied these interests to the development of software interfaces used by millions around the world, most notably as a member of Netflix’s Television UI team, where he worked on experimental initiatives conceptualizing and prototyping the future of entertainment software.

Presently, Matthew is building the underlying modeling architecture at Bioform Labs, a company focused on using the Active Inference toolkit to model organizations as emergent cybernetic organisms. He believes these models can help organizations manage their deployment of and interaction with AI-based agents, as well as more adaptively manage their own emergent complexity.