aquabu
commented
5 years ago +1 @harlantwood
+ Reasoning Engine For Attestations
It would also be great to prototype a reasoning engine for attestations - along the lines of Prolog / Datalog. Ideally this would provide some common rules of inference to make sense of the implications of the attestations for specific use cases.
Here are some reasoning_engine(trust_graph, inference) use cases that are especially interesting to me.
Use Cases
Journalism Attestations
This will be extra important for upcoming election seasons where there are strong incentives for nearly all campaigns to distort and amplify claims to benefit their election goals. The same mechanisms for validating claims are also useful inputs to DAO communities as they scale and need to communicate truthfully with DAO members (human or machine). The data model that we came up with during Web of Trust was like this:
- Assertion = Proposition + Evaluation + Evidence. With each being signed JSON-LD claims.
- Attestation Schema & Code
- Portable Reputation Toolkit Use Cases
For Journalism cryptographically signed attestations indicating where video & photos originate from will be especially important as we enter into the era of deep fakes. Adding GPS proofs would be an additional step. Even if narratives vary, having a grounding in what did or did not occur can temper demagog-ery.
Portable Reputation For Work Platforms
My view is that portable reputation for work should be more like an asset (proof of awesome) and less like a liability (social credit score). Thus the portable reputation should be related to what you have achieved in a positive sense. We modeled something like this here for a project called WorkNation
Computational Law
Smart contract oracles are dependent on verifying certain worldly facts through various means. Legal contracts are dependent on determining the proper functioning of smart contracts and the truth value of certain factual statements. Attestations plus a reasoning engine implemented as a non Turing complete cellular automoton would be quite useful for bridging the worlds of law and computational contracts. I am not only referring to the smart contract code part. I also mean the computational representation of human readable legal documents that are required for a contract to be recognized and interpreted by a legal jurisdiction.
For example, see the good work of Mark D. Flood and Oliver R. Goodenough on [Legal] Contract as Automaton: The Computational Representation of Financial Agreements. This work is directly related to DAO decision making. Co-author Oliver Goodenough was instrumental in creating legislation for the Vermont Blockchain Based LLC which is designed to create a limited liability structure for decentralized orgs of which Dorg.tech is the first.
An example would be the reasoning for when to distribute payment for an insurance contract based on attestations about worldly events.
Contact
I'm @noahthorp on decentralized web slack, telegram, twitter. Or noah@comakery.com
Idea
Decentralized Self Sovereign Isomorphic Trust Graphs
What if all our decentralized apps and platforms could share trust/ratings/attestation information, in the form of portable cryptographically signed claims?
What if this self-sovereign, user owned and controlled, privately sharable trust network could become the backbone to unite our many disparate decentralized web apps and platforms?
Finally, what if by uniting our networks in this way we could snowball into an effectively larger, deeper, more powerful network effect than today's legacy social networks and chat apps?
Critical Features
Here are the features that seem to me most needed to make this work well:
source,target, semanticcontent, normalized scalarvalue)Trust Graph project
A small team, including myself and fellow DWebCamper Noah Thorp have been working off and on for several years on the open source decentralized ratings platform Trust Graph, which has all of these goals, but still has a long way to go to achieve them. We've learned a lot along the way, and want to offer that here; but we're also unattached to any one solution; we are suggesting here the building blocks for a generic solution, and together the community can choose to use components of Trust Graph or other projects or build from scratch.
Skillsets
What skillsets would be useful to develop this idea?
Contact
I'm new to Matrix; my username is showing up as
@harlantwood:matrix.orgon#dweb-camp-2019-projects:matrix.orgI'm also
@harlanon the Decentralized Web Slack group.I'm also @harlantwood on Twitter.