Decentralise: Decentralise and diversify, ‣.

Distribute (spatially): distributed neural representations, distribution in infrastructure, etc. makes systems reliable.

Disintermediate: where possible, provide direct access. The purpose of mediation is to manage the flow (reduce the volume), not the information itself.

Disaggregate: units of information should be as small as possible.

Disintegrate: entities in a network are not ‘components’ of one another.

Democratise: autonomy + Diversity.

Dynamize: it's only through the process of change that a network learns. (Or moving through the reference frame: see Humans learn only within reference frames.)

De-segregate: learning is not a separate activity. The same tools that are used in practice are used to learn. This is an important distinction between brain intelligence and AI with separate train and inference stages.


References I define Openness as one of four elements of what I call 'the semantic condition'. The other elements are autonomy, Diversity, and interactivity. The semantic condition, in turn, is what informs the design of networks to enable them to learn. Networks - and therefore, people and technology and systems and the rest - that embody the semantic condition will be responsive to and adapt to the environment, and hence be resistant to cascade phenomena leading to stasis and network death. That's a lot to pack into a single paragraph, but I've written about all that elsewhere.