[aianus]

I think Ethereum has the potential to be the BitTorrent to Bitcoin's Napster.

A lot of the killer Bitcoin apps (gambling, drugs, etc.) make it too easy for law enforcement or site operators to steal everyone's money due to their centralization. It's a long shot, but if Ethereum gets really big really fast it's going to be much, much harder to shut down these apps.

[nugget]

Maybe you can answer one persistent question I've had about Ethereum contracts. My understanding is that you can encode logic in the contract so that you and I can, for example, gamble on the outcome of the Superbowl and the contract ''knows'' to ask some trusted authority for the result of the game and then pay out accordingly. Isn't this only decentralized insofar as you and I both trust the trusted authority and that authority isn't ever compromised? Doesn't that authority become the point of attack for hackers? Such that if I lose the Superbowl bet but I can hack the ESPN API at the right point in time, then I can trigger the wrong payout on all contracts and there is no recourse.

[aianus]

It's my understanding that the Augur project solves this issue by decentralizing the reporting.

That is, individuals vote for which outcome is correct in proportion to their balance of 'reputation'. If an individual votes for the correct outcome (as determined by the majority of votes), they gain reputation, otherwise they lose it.

Reputation is valuable because it entitles voters to a share of the platform fees. Therefore there is a strong incentive to play fair in order to grow the platform and the value of one's reputation (similar to miners in Bitcoin).

https://augur.zendesk.com/hc/en-us/articles/208857885-For-Re...

[Natanael_L]

Sounds like somebody can dump in money to skew all the reputation scores.

[aianus]

As in buying a lot of REP and faking an outcome?

You could do that but the idea is that the hit to your REP value from screwing people over will be greater than any illicit profits you may gain.

Similar to why your drug dealer doesn't just point a gun at you, take your cash, and keep the drugs; he can make more money being an honest long-term dealer to you.

[RangerScience]

AFAIK, yes, that is an issue. But it's not like it's a completely new issue. Timely information alteration is a long-standing tradition commonly known as "conning", as in "conman" as in "confidence man".

Under one perspective, the root of blockchain technology is a rigorous way to determine consensus reality. Ethereum is then an evolution on top of that - giving a good way to process that consensus reality.

Theoretically, one could begin implementing sensors of actual reality - is it raining today? who won this sportsball event? - within the system.

The TL;DR is that: Yes, but a) that's not a new problem and b) that's why this is all still under development and experimental, and why other people in this discussion only advise entering this market as an enthusiast.

[j1vms]

Employing a form of statistical threshold on the information that is gathered before making a decision could probably help resolve a lot of the potential for gaming the system.

[phodo]

That is insightful and is the concept of an 'oracle'. There is a spectrum of techniques to address having a trusted, tamperproof oracle... Some better than others, eg checking on multiple public sources and triangulating their results to more robust voting systems. I haven't run across one that is perfect, though. I believe this will get addressed over time.