News for hackers. Stuff that matters.
Re:Blockchain study finds zero success rate and vendors don'... by nostrademons on Friday November 30, 2018 @07:25PM (#18574081)
They make sense when the primary vulnerability isn't the person doing the recording but the possibility that the data might be tampered with or misrepresented before it gets to the consumer.

Consider, for example, the problem of restaurant reviews. We know that one person's review of a restaurant isn't going to be 100% accurate, because it's subjective. Socially, we've figured out ways of getting around that, like soliciting reviews from many different people, looking for reviews from people with tastes like ours, looking at the specific details they post, etc. They're also not generally incentivized to lie, because they've already eaten there and their review isn't likely to influence their service in the future.

All this falls apart if, say, the platform responsible for collecting these reviews has a business model where they are incentivized to bury negative reviews and manipulate scores upwards, but only for restaurants that have given them money. It'd be a much more trustworthy system if the reviewers could receive currency directly for the information they provide, provided by the potential customers who benefit directly from that information. Alas, basically nothing on the web works that way.

If only current blockchain implementations had acceptable performance to handle terabytes of data on mobile phones...

Re:Blockchain study finds zero success rate and vendors don'... by theoh on Friday November 30, 2018 @07:53PM (#18574218)
What makes your proposed system immune from abuse?

"reviewers could receive currency directly for the information they provide, provided by the potential customers who benefit directly"

The greatest benefit goes to restaurants with positive reviews, which are thereby incentivized to game the system.

Blockchain might have solved the problem of trustless consensus with proof-of-work, but you are now introducing a new situation in which any number of fake, anonymous contributors can distort the review system. If the review reader pays the review writer directly, upvoting their own phony reviews doesn't cost the corrupt restaurant anything.

Re:Blockchain study finds zero success rate and vendors don'... by nostrademons on Friday November 30, 2018 @08:11PM (#18574307)
It's not immune from abuse. It's immune from abuse of the particular sort I describe.

I'm not so naive as to think that all problems are magically fixed by blockchain, but given the number of posts on HN and elsewhere who complain about the power of big tech companies and assume they're tampering with or misrepresenting data for their own profit, it seems like people consider this problem to be important.

Re:Blockchain study finds zero success rate and vendors don'... by theoh on Friday November 30, 2018 @08:23PM (#18574361)
Blockchain can't fix the problem of lack of trust, where it is trust that is desired.

The trustless consensus idea is inherently abstract and paranoid, for better or worse. But life in human society requires constant acts of concrete trust and mutual consideration.

The cypherpunks who invented the blockchain don't seem to have explored the idea taken to its "logical" conclusion of ripping out all trusted entities and (somehow) replacing them with exchange of anonymous cryptographic tokens. The blatant truth is that doing away with trust altogether would paralyse every aspect of civilization.

Re:Blockchain study finds zero success rate and vendors don'... by spookthesunset on Friday November 30, 2018 @08:12PM (#18574313)
Why not just use a git repo for reviews? You can clone it and all that good stuff...

What, specifically, does a review system gain with The BlockChain?

And who pays for it? Blockchain’s aren’t cheap!

Re:Blockchain study finds zero success rate and vendors don'... by coryrc on Friday November 30, 2018 @07:52PM (#18574213)
Also doesn't address astroturfing or spam, which are more of a problem without a gatekeeper.