When Ethereum launches if you'd like to contribute to the project you can create a smart contract which can act as part of the proof of concept. Ultimately it's going to depend a lot on how robust Ethereum is for developers but ultimately contracts are like scripts which run on top of Ethereum.
DAVS has to be platform agnostic but it's first home will be Ethereum. It should be able to communicate with different blockchains and protocols though. This means it has to be designed to be modular, extensible, and able to interpolate with the Internet of Blockchains, and other decentralized apps.
If you've got programming skills we are looking for developers to help write smart contracts. The overall design is currently being discussed in Facebook with my blog acting as a place to introduce new concepts which are design options or possible features. The main thing is that we want it to be a decentralized bottom up emergent design so it's not so important to try to predict every possible interaction or evolution the software could take but more about making it as adaptable to change as possible.
DAVS has to be designed for adaptability because it has to survive any of the technological changes which come about.
Some points below can bring it all together:
DAVS is to be (open) because it must contain these properties
- All source code must be open so that it can be audited and reviewed.
- We should choose the languages which are simple for people to understand or which are widely used, comment our code to make it easier read, and try to encourage code reuse.
- All participants should be able to help direct the evolution of the architecture and should be in control of their data.
DAVS is to be (decentralized) because it must contain these properties
- Fault tolerant design
- Disruption tolerant network design
- Modular design
- Compartmentalized for security where necessary, extensible design.
DAVS must be (autonomous) because it must contrain these properties
Formally Wikipedia defines these concepts as autonomic computing:
Or it can be described as organic computing:
DAVS will not be fully autonomous in the sense of an AI running everything without participation from humans. In fact it will be up to the humans to determine how much AI or how much human control they want. Humans will be able to express their preferences, vote, in various ways to provide feedback to the system.
The mechanism that DAVS can use to achieve a semi-autonomous state is currently only theoretical. I think it can work and it does seem based on examples such as Bitshares that you can have a self funding mechanism, you can have consensus mechanisms. The concept I put forward is for "virtual evolutionary organizations" http://darkai.org/?page_id=171 .
A virtual evolutionary organization is a self emergent design which uses AI to "evolve" from one architecture to the next or to optimize itself according to participant feedback. So the key for making software semi-autonomous is the feedback mechanism whether it be voting, collabrative filtering, learning conditional preference networks (for AI recommender systems), The point is you can have traditional democracy, but the architecture itself can learn from how you use it as well.
Most of this is theoratical and doing it in practice might not be so simple. On the other hand if you're talking about a decentralized autonomous virtual state then you have to be very ambitious because of the nature of what it is. The design of it in my opinion is one of the most important aspects to get right so that even if we find we cannot build it as intended in the design at least if we can build a prototype or some proof of concept then maybe in the future someone can pick up the design and finish building it.
The other reason is that the design has to be future proof. If you look at Bitcoin that isn't future proof at all. It's a good design for currency but it's not designed to be very good for a governance protocol because it doesn't have democractic processes at all, you can't vote, you can't give the system feedback, it just runs.
At the same time it's designed in a very top down traditional architecture with miners at the top generating coins, defending the network from a 51% attack. On the other hand what I'm asking we do is go with an emergent engineering approach as outlined in this paper: http://doursat.free.fr/docs/Ulieru_Doursat_2011_nets_IJAACS.pdf
Mainly because you cannot really plan the perfect design for something like this in a top down manner. As a distributed design it's more resilient if it has no "top" to corrupt.
There is too much to put all my ideas in one post and I'm not as good at explaining as some others. So I encourage you to peer review anything I write about in my blog and any design ideas which you think are stupid please shoot it down.