Hacker News Reader

Comment by Tenoke | Hacker News Reader

I like decentralized services as much as the next guy, but here are some problems off the top of my head:

1. Running a node could easily be made illegal in most jurisdictions (abiding criminals), which is problematic for multiple reasons (there is distinction between helping the market operate and merely browsing it).

2. A big enough bug is enough to get everyone involved in trouble, and even if it is easy to fix the flaw everyone's identity up until that moment would be compromised (bitcoin for example has had many problems, which were later fixed, except that the stakes weren't as high)

3. During disputes, the arbiter can side with whoever offers them the biggest fee.

View on HN

> 3. During disputes, the arbiter can side with whoever offers them the biggest fee.

Then he would be left a negative review. Arbiters will be chosen based on fairness.

Sure, but making new identities is cheap in such an environment. Not that I am saying there are no fairly easy solutions, that could be implemented to prevent this, or that this is a big issue.

Sure, but making new identities is cheap in such an environment.

Yes, but the real question is how easily can trust be manufactured?

No, it definitely isn't. It's not enough to have zero negative reviews, you also need to cultivate hundreds of positive reviews.

How do legitimate people get their first positive review?

Exactly. A little social engineering by the police could quickly erode the trust in the network.

Police could set up a ring of their own accounts, facilitate fake transactions amongst each other, leave positive reviews, take down one large player, then shut that ring down. Imagine hundreds of these rings within the network.

That's possible if they're using only network-global review and reputation system.

If they're also relying on WoT-like reputation system where peers rate peers, and reputation's calculated from those chains of trust, massive sockpuppet attackers won't be able to quickly erode trust, as they'd generate that trust only throughout their own circle, having minor impact on others. They'll have to play by the rules for a long time to spread connections through the network.

No idea how architects handle this, but my guess is, this could be done either by relying on external reputation (i.e. knowing the arbiter outside of the network) or by network somehow rewarding buyers and sellers who take a risk to chose "newbies" as arbiters.

The arbiter is going to be left with a negative review no matter whether they were fair or not. There's often some measure of uncertainty in arbitration - you can barely ever prove beyond all reasonable doubt that one side attempted to scam the other.

Basically, you're going to have to trust yourself to see which arbitrators are scammers, based on reviews, some amount of which are likely to be fake. This doesn't really help the whole situation of avoiding scammers, just moves it around slightly.

Maybe add a second arbitrer that overlooks the whole case and decides if the first arbitrer has been fair?

And how do you trust the second, when you can't trust the first?

In the real world, there is no such thing as "trust" at all.

On the contrary - in the real world, trust is unavoidable.

Disregarding the obvious familial trust, I trust the manufacturers of the food I buy not to poison it. I trust that the gas pump is actually pumping gas. I trusted the bank teller to deposit the cash I gave her into my bank account and not just put it in her pocket (a trust her bank shared as well.) I trust my employers to pay me and my employers trust me not to steal their IP.

Of course, in each of these cases, there exists a system to enforce regulations and punish infractions when trust is broken, because while trust has to work in many social and business transactions, people cannot always be trusted not to cheat, lie and game the system. And yes, this means one still has to trust that system to an unavoidable degree.

This prototype had a single arbiter per transaction, but there is no reason why you couldn't have N arbiters on each transaction, for as high an N as you want.

Imagine having, say, three arbiters for each transaction, and two of the three have to side with one side in order for the money to be moved. This cuts down a lot on the potential problems, especially if the arbiters were kept anonymous from one another, thus preventing collusion and making sure that each investigation is independent from the others.

Reply on news.ycombinator.com