Skip to main contentdfsdf

Michel Bauwens

Playnet

"playnet
a plan for meeting each other's needs."

Shared by Michel Bauwens, 1 save total

Michel Bauwens

Talk by Evgeny Morozov - Socialism after AI - INDEP

"the question is no longer only who owns the machines, but who controls the infrastructures through which societies learn, create, remember, govern and make worlds."

Shared by Michel Bauwens, 1 save total

Michel Bauwens

INDEP online Talk with Marek Solis - Playnet - Viable Socialist Computation in the 21st Century - INDEP

"Playnet.earth is a community aiming to replace the market as the mechanism that coordinates society – with a continuously updated, computer-assisted system that plans production according to people’s desires, the labour actually available, the technical constraints of the tools and skills at hand, and the ecological limits we can’t cross?

The socialist calculation debate of the 1920s–40s asked whether a fully detailed planned economy was computationally possible. Kantorovich, Koopmans, and others showed the mathematics worked; what was missing was a machine large enough to run it. Those machines now exists while Amazon, Walmart, and Uber already operate modified versions of those solvers at civilizational scale, walled off and profit-captured, solving over a limited subset of humanities supply and demand. Playnet proposes generalizing the same computational methods for satisfying our basic needs in common while changing the optimization objective. The talk will share the latest work of the playnet & valueflows communities as an invitation to collaboration with the INDEP community."

Shared by Michel Bauwens, 1 save total

Michel Bauwens

Where Meaning Lives - Sacha Pignot

"ValueFlows separates observation from valuation. It is a vocabulary for distributed economic networks built on resource, event and agent accounting. Its core records flows: an economic event is an observed movement with a provider, a receiver, a resource and a measured quantity. What matters is what the core leaves out. It gives you a rigorous account of what happened and does not fix how a network should assign worth to it. Worth is a reading, and it belongs to whoever is doing the reading. That is why incompatible communities can adopt it without being forced to converge, which is not common in accounting ontologies. The same discipline shows up inside the core: a resource carries both an accounting quantity, what an agent has rights over, and an on hand quantity, what is physically present. Two numbers for one physical situation, because rights and presence are two different frames. The ontology keeps both rather than resolving them into one true quantity.

ADAM separates addressing from interpretation. Coasys’ agent centric meta ontology is built on agents, languages and perspectives, using Holochain underneath. Every piece of data is an expression: a statement signed by its author, carrying author, timestamp, content and proof. Their own framing is that this produces a web of verifiable claims rather than a body of objective data, which is the same distinction built into the storage primitive. What is checkable is that this agent said this at this time, and that is checkable absolutely. What is not thereby settled is whether it is true or what it means. Then a language is a pluggable adapter defining how an expression is stored and shared, saying nothing about meaning, while a perspective is a personal semantic graph that gives meaning to expressions through links. Publishing one creates a shared space others can join. The passage from a private reading to a shared interpretive space is an ordinary runtime operation rather than something fixed before anyone has read anything.

Functional programming separates composition from emergence. This one looks like the wrong ally, since referential transparency says the meaning of an expression does not depend on its context of evaluation, which is the flat contradiction of the hermeneutic position. It serves the pattern anyway, for a reason usually stated badly. FP does not remove complexity, it relocates all of it into composition. When every rule is locally deterministic and nothing hides in mutable state, a surprising system level behaviour is attributable to interaction rather than to something concealed. That is what makes emergence observable instead of indistinguishable from a bug. And the tension with situatedness resolves in the type signature: rather than letting context be ambient, make it an argument, observe : Observer -> EventStream -> View. A reader over a fold. Meaning still depends on position, but the dependency is declared rather than suffered.

The honest tradeoff: compositionality assumes the whole is the composition of its parts, and strong emergence is exactly what is not recoverable from composition. This gives you a faithful description of an ecology, not a prediction of its behaviour. Compositional semantics at the level of rules, non compositional behaviour at the level of the system. Only simulation covers the second."

Shared by Michel Bauwens, 1 save total

Show more items

Diigo is about better ways to research, share and collaborate on information. Learn more »

Join Diigo