We need a good way to store what is true. For example, for verifying the identity of people. But also other things: “Capital city of Greece is Athens”. It can be an App that handles the facts and the opinions about the facts.
Let me give a simple explanation by a simple example:
We have “statements”:
X has name Norimi.
X is from Lithuania.
X likes horror movies.
X hates rainy weather.
X has friend Y.
We have “opinions about statements”:
Y agrees 1 is true.
Z agrees 2 is true.
Y disagrees 2 is true.
Y agrees 5 is true.
X agrees 5 is true.
Now we know Y and X are friends because both Y and X say 5 is true. So we think X is not from Lithuania. Because we think 8 is more important than 7 because we know Y is friend of X but we don’t know about X and Z relationship. We think X has name “Norimi” because friend of X says that.
Easy to post statements. Easy to post opinions. Just need a format.
One statement or opinion is one small block of data. Statement can be anonymous. Opinion must be signed. Alternative: Only statements. Signed. If I say the same thing that another person says, I have to repeat it.
More complicated: Some things are more important than other things. For example: “X is friend of Y”. This is a “meta statement” because it changes what other statements mean. If X and Y says this statement is true, X’s opinion about Y and Y’s opinion about X has bigger weight. Existing work:
Big questions about safe network search: a) How to find the statements about the things? b) How to find the opinions about the statements? c) How to collect these efficiently in real time?
Related to the links you posted I think Ruben’s article is relevant (re searching across data sources) - link below.
Also, there’s a Solid project called twee-fi
where you can review tweets etc.
I think there’s a lot of people thinking along the lines of your OP and that SAFE will play a part in this (I remember @Seneca talking about it ages ago wrt Project Decorum, and I was thinking about this too back then).
There are two separate things: a) Know that content belongs to account. Easy with safe tech. b) Account belongs to an actual person. Not a made up persona. Hard because not tech problem.
But tech can help make it easy to express structured knowledge. If somebody says “I know that is true” and I know that person personally and trust him, I can trust that that thing is really true.
If we can search this knowledge graph, we can access much information about things and about how much to trust that information. So this store is multipurpose. A framework that can be a service helping other services.
Thank you for the information. By the way, I have a bridge in San Francisco, you wanna buy it? I’m the owner, post-truth, on the internet.
I hope I am on the right page here with what you are talking about, but isn’t this substantiation of facts the concept behind Factom? It’s a blockchain technology where information is verified by several parties as I understand it. The more people verifying the “fact”, the more reliable the information.
Example why not the same: If I know you, I can trust you (or maybe I know I can’t trust you ) If another guy says something, it matters less. Because I don’t know him.
In my idea, it matters who says something. Our relationship decides how much I care. His relationship with the subject decides how reliable is his opinion. Not just a simple voting system.
Look at the references from a previous post: RDF, semantic model, triple store. They express real “structured knowledge”. A graph of statements. Not just a simple voting system like Factom (if it works how you said).
Another difference (if I understand Factom) is that I can change my opinion. It is not set in stone as with blockchain. Voting is right to be permanent. But this one has a different purpose.
Agreeing on a given topic does not mean X and Y are friends or agree on ALL topics. It doesn’t even mean they know each other. I could agree on 80% of what one could say and the 20% we differ on could be the difference between being best friends and mortal enemies. So you need to take into account the relative value of a topic to an individual before you can calculate the likelihood that it will affect a relationship status. It’s not enough that a topic is agreed on, it’s which topics specifically are agreed on and which are not.
In the example: Both X and Y says “X has friend Y” is true. It means they are friends.
If Y is friend of X, when Y says something about X, it’s more reliable.
That’s all. There was no suggestion they agree on everything.
Yes. That is why I referenced prior work. RDF, semantic models, and so on. The safe network can build on those things. But I agree, it is a complicated subject.
Than from your point of view, are the tokens and futures? Unsubstantiated fiction. Yes, you will show white lsty, a pile of debt. In the end, zilch. So why bother with the truth. Can a lie take for granted. And the fact that a physical person knows another physical person, this is nonsense. You your wife or girlfriend KNOW? Are you sure about that. So much so that individuals are not suitable, but to sew the chip under the skin ))))
This is not global trust. My account need to trust your account. Until it does, you can have a million of them: My choices are not affected by them at all.
If I accidentally trust your bad account: When I notice it’s a bad account, I say: “DON’T TRUST THIS ACCOUNT.” Then: Everybody you trust (your other 99 accounts) will be trusted less by me. So: Thank you for your hard work to mark all the bad accounts I shouldn’t trust.
If your bad account trusts a good account because you want to discredit: If it’s a good account, other good accounts will trust them, so your discrediting attempt will be seen, recognized, and ignored.
) If another guy says something, it matters less. Because I don’t know him.