1
Judgement System / Re: Rules and karma
« on: July 06, 2014, 03:54:39 am »Maybe if trust percentages were given rather than a black and white all or none system where the guideline was "how much of your NXT would you trust this person with."
This is kind of a half formed BS idea but what if we actually enforced that by, every now and then, granting someone else the ability to spend the person that trusted them's NXT. That would keep people from fooling around with their trust network. But if we were to do that, we would need a way to keep people from gaming trust then cashing out on it big time.
Yup we'd definitely need a way to keep people from building the trust and running off with a big scam. That's where it's tough to kind of think of incentives to keep people honest but my one workaround is this: if information about who you can and can't trust somehow flows between individuals in localized networks, and this information can be cut off from bad nodes, it seems to me this is like a good incentive to stay honest because it's difficult to build trust relationships again in a new account and difficult to transact when you don't have any or bad information. It's sort of a network effect that creates high switching costs. This sounds rudimentary but it would kind of be like if everyone defriended you on Facebook, except instead of not receiving pictures and videos you would be barred from trust information and your reputation would be nonexistent on the network, making it difficult to execute financial transactions for you and with you.
Regarding your trust percentages point, I imagine it would amount to something like that but I haven't figured out exactly how to quantify it. It would need to be simple and intuitive for people to understand, and I think it should come from information that exists on the blockchain. For example, the strength of your trust connection could be quantified using number of transactions between nodes, or number of smart contracts successfully completed, and this number could be an input into a localized trust rating. If we were to approach globalized trust ratings, that would involve complexity beyond what I'm able to think about at this point.