@Happy Being - I thought it might help to incorporate some (mock up) components & functions/players into your diagram which could be key to i. help flesh out an ““Open Ended Evolutionary Algorithm” (OEEA) design path ii. establishing our Open AI equivalent discussion group equivalent to “centralise some of the thinking” as described by David (which I will respond to separately) and iii. create a common vision/language & safe data commons project office. It’s rough and looking at how to illustrate the components/players which would appear to need to come together for mutual benefit has been a big learning/integration exercise …. so I hope it helps and that I’m not daft as well.
In this first diagram I have connected your diagram and the r3.0 diagram under the heading “a radically democratising algorithmic driven economic model/platform” as they appear to be different views of the same meta algorithm outcome and thus complimentary.
png)
I’m assuming that the Meta algorithm facilitates the “ intermediation of mathematical models and algorithmic metrics as well as the design of data flow architecture?"
In this 2nd version of your architecture diagram I have inserted the aligned UBER AI preso “Multidimensional Neural networks” slide and a number of smaller diagrams numbered 1-10 which flesh out existing or add key architecture components/requirements which as you say need to be evolved e.g. indexes to underpin a decentralised search app/ultimate path
Summary of diagrams 1-10 highlighted in the previous slide
1. User - UI - This is a “ very rough” solidonsafe/”sense” decentralised search page mock up around a google help diagram … to table some concepts/ideas and picture of a possible (decentralised search, public data commons utility and user established governance system) outcome to start a discussion as to what we collectively want to create, what it looks like, the benefits, what sits behind it and the easiest (project commonssense) path to it.
2. Sharing model – This diagram illustrates a forager connecting to a sharing (public data) model described in the Open Data Institute “The role of data in AI business models” paper as the optimum AI model i.e. Mutual AGI prosumer vs proprietary AI dataplume. Nick Shadbolt makes the comment “ there ought to be some way of managing large scale data sharing efforts that are in everybody’s interest” …… which in theory the safe: data commons mutual AGI dataplume communitylink model path can enable.
3. Multidimensional Neural networks optimisation – If I understand this correctly this UBER AI Deep neural network diagram illustrates what I call as smart product matrix to define the phenotype elites/interlinks to “mutate locate replace” and which can be overlay over any community, industry product need i.e. codemap. A key initially use for a safe:commons phenotype matrix method is for safe network Data Council neural network ecosystem definition and subsequent reuse for members ecosystem i.e. Davids actors/hyperneat example.
4.Meta Algorithm – How do we earning rewards in a sharing model/OEEA? This is a link to an interesting methodology i.e.Real Time Offerings with meta utility token for multi-currency engagement and transition from NNR to SRR value exchange … we can add the time based standards approach (using the CCDM resource & need optimatisation method)
5. Deep Learning - This “Prediction Function of One Hidden layer diagram/formula” is from the Neil Lawrence Deep Learning presentation “Dimensionality & activation of the neural network … to define one of the Multi-dimensions ?) activation functions & parameters for the sharing model/OEEA formula (using the Unison language definition?s] i.e. a radically democratising algorithmic driven economic model/platform. His Guardian Data Trusts could allay our privacy fears article addresses directly the need/problem we are looking to address.
6. Safe/UIA webID sign up – This diagram illustrates a mock-up of a Holistic Code UI/UX user interface to create y/our personal via peer (in the centre) , product, project engagement definition. In the collective intelligence field they call it defining y/our learning profile within a sharing model or collective intelligence.
7. Data Council “Sensor” (solidonsafe) NODE – I am assuming each proposed Data Council Node thru the safenetwork will be an enterprise network of members in their own right to create their own codemap/neural network. The NZ Govt Health Dept has implemented a fully templated Health Data Council CCDM CCDM/data governance methodology from ward thru to hospital level data councils which provides an organisational template to build upon.
8. Rough Engagement xls & statement diagram to capture how contribution & rewards earned via the meta algorithm. To practice what we preach this meta algorithms starts with a Foundation” co sig co dev group where the project delivers to what each is promoting in a via a Collective Real Time offer to capture our generic “go explore”/OEEA method as we go.
9. Learning Academy – This diagram is a mock up overlay of the Ui Path Academy which deploys the related training programs for “actors/roles” within the Ui path Industrial AI platform/software. Our OEEA model requires learning courses in a continuous learning environment (what I call a “Living university”/ULB ) which could evolve as David described via… “the unisonweb.org lib for actors to create simulated environments in which the agents evolve and directly affect the environment in which they are tested for groups or network of actors to work together, for physical world application” using a Multidimensional Smart product matrix archive for collective training course needs definition & OEE algorithm optimisation.
10. Smart product matrix… is the overlay to deliver to a pre-defined smart AI product outcome to define the new interaction/connection gamification pts i.e. the multi-dimensional archive of phenotypic elites (each with a 0 Reserve & index meta token to distribute pre agreed margins to contributors) and so we can gradually automate some of the process.
IMHO to bring the key co-dev stakeholders together around a Open AI equivalent Foundation discussion group and safe commons/OEEA path will require a specific converging path/framework & complimentary opportunity/outcome to assess. This will need a prior discussion to align our Epic stories and address the questions … Which EPIC stories to we share? Where do our EPIC’s & stories intersect & how do they interact? What new shared EPIC’s stories do we need to establish? What adjustments do we have to make to our existing EPIC’s stories/project plan …. to frame the co-development opportunity to easily include other’s “activation functions & parameters” in key contribution areas.
This is all a bit rough so I’m hoping it helps and makes some “sense”.