Messages

"Rep" or "Reputation" is the currency that is owned by the members of the oracle. If you want to participate in the oracle, you have to buy rep from someone who already has it.

Augur's security model says that the rep owned by the oracle has to be worth more than all the outstanding shares in all the markets that the oracle judges over.

As Paul Sztorc talks about in this essay: http://www.truthcoin.info/blog/pow-cheapest/
The cost of locking up 1/2 of all the coins is very expensive. It is as expensive as the interest rate applied to the locked up coins.

My hope is that it is possible to use rep worth 1 to judge over markets worth 100. A > 1/100 ratio.
Is my dream impossible?

Here is the reasoning Augur is using the justify their > 1/2 ratio:
If the oracle lies, the value of rep will drop to zero.
The oracle will only lie if the profit exceeds how much they lose in rep.
The maximum profit is smaller than the value of all the outstanding shares.
You can only steal one side of a bet, bets that are very unequally probable are most profitable to attack.

It looks like augur is prepared to use dominant assurance contracts to gather the funds to pay oracle.
Their oracle can get paid both ways.

Augur has a very simple solution to this problem.
The person who proposes a decision to the oracle needs to pay money to the oracle.
Even if there are zero trades and zero trading fees, the oracle still gets paid to judge on the outcome.

There is a whole class of altcoins that Hivemind can't coexist with.
That is the same thing as saying: "There is a whole class of attacks that Hivemind is vulnerable to".

If you used dominant assurance contracts instead of trading fees, then you could coexist with any other blockchain.
You would welcome and encourage the existence of "parasite contracts". It allows your blockchain to specialize on judgement, and someone else can specialize on trading.

One major way we could improve trading: We should have batch auctions with uniform pricing.

This is discussed in the blog post "Oracles are the Real Smart Contracts".

The argument in "Oracles are the Real Smart Contract" seems to be about turing completeness, but the attack I am talking about doesn't require turing completeness at all. Any cryptocurrency that lets you make bets based off bitcoin/hivemind's state is good enough.

the altcoin would not have sufficient mining power.

Mining power doesn't matter. A government sponsored blockchain could be the attacker.
Their authority doesn't come from mining, it comes from law and enforcement and the format they accept taxes in.

For example, if Mexico decided to move the peso onto a blockchain, and they taught their blockchain to be aware of bitcoin and hivemind's state, then anyone in Mexico could make bets without paying trading fees.
The market cap of Mexican peso is high enough to be a problem.

Ethereum is popular today, but don't get too distracted by it. Anyone can make an altcoin.

Yes, you do need meatspace information, to set the Schelling Indicator.

One of us is confused about something.
Most types of crowdfund do need meatspace information. Because we usually crowdfund something in meatspace. Like having a new song written by a famous artist. We only pay the artist if the song actually gets written.
One type of crowdfund does not need meatspace information. If the crowdfund is a success just by virtue of the money having been raised, then we don't need any meatspace info. The blockchain already knows how much money was given to the crowdfund.

We can raise money to pay the oracle without collecting any trading fees.

A numeric example:
You and I have a channel on hivemind, we each put $1000 into it. We make 100 bets over a year-long period, each with a $1 fee, so there are $1900 left in our channel, all of it is on your side, because you kept winning bets more than me.

We each lock $100 into an ethereum contract promising to settle the hivemind contract at a sequence number of 1, and with the final balance of $1949 on your side, and $50 on my side.

The total amount of time $200 was locked up on ethereum is as long as it takes to close the channel on hivemind.

Case 1, we are both honest. on hivemind you get $1949, and I get $50, on ethereum we each get our $100 back.
Case 2, exit scam:
After we make the agreement on ethereum, I try to finalize the hivemind contract at the step where I had $950 and you had $1049, (after I had lost the first bet).
The ethereum contract locks our $200, and refuses to give it back.
You provide the guilt transaction. Hivemind gives you $1900, and it gives me $0.

Because altcoins exists, and because you collect trading fees, lightning network introduces a flaw.
Pairs of traders can participate in hundreds of trades, then close their channel as if they had only participated in one trade.
This way, they don't have to pay as many trading fees.

They lock coins on the ethereum side as a promise not to bring each other to trial.

If the oracles aren't over-paid, they will steal.

Ethereum already has bitcoin relay. If hivemind was very popular, they would participate without paying fees.

The oracle needs to be over-paid, or else they will steal.

Paying the oracle a trading fee wont work, because any altcoin can participate in gambling without paying the trading fee.

One way to solve this is to crowdfund a payment for the oracle before the oracle agrees to provide judgement over a decision.
That way oracles will only judge over things that pay a high enough fee.

Check out reptrader on reddit. Maybe you can sell there.

To fund the oracle, we want a dominant assurance contract that succeeds if a certain amount of money is raised.
We don't need an oracle to create this type of PM.
The blockchain doesn't have to learn any meatspace information. It knows how much money was given to the crowdfund.
I think you can already do this type of crowdfunding with bitcoin.

If we make a PM that doesn't have an oracle, then it will be safe from FRP.
We can use this PM to bootstrap funding for oracle-powered PMs.

Couldn't anyone make an altcoin to commit these same attacks that you describe for sidechains?

Paul Sztorc made this video about how blockchains can interact destructively.
https://www.youtube.com/watch?v=2OOKgTSrITs&list=PLw8-6ARlyVciMH79ZyLOpImsMug3LgNc4

His central argument: when miners allow the creation of a sidechain, they need to use the same caution as a soft-fork.
Support for this argument:
* Since oracles are paid by trading fees, sidechains create a free-rider problem. Participants on the sidechain can bet in hivemind markets without paying fees.
* It is possible to make a sidechain for stealing bitcoins from the main chain.

His arguments apply to alt-coins in the same way they apply to sidechains.
It is impossible for bitcoin miners to stop someone from creating an altcoin.

At this point it would seem that blockchain prediction markets are an impossible goal.
Anyone can make an altcoin to ruin it for everyone.
Read a little more, hope is not lost.

=== Focusing on the "free rider" problem.

Paul has previously claimed that hivemind will solve the free rider problem in general. He talks about it on page 14 of the applications document: http://bitcoinhivemind.com/papers/3_PM_Applications.pdf

Here is a meme illustrating Paul's contradictory claims.


Flying Fox is being designed under the assumption that Truthcoin dominant assurance contracts do solve the free rider problem. So we don't collect fees from gamblers at all. Instead, a dominant assurance contract is used to raise money to pay the oracle to make a new market.

===Focusing on the "sidechain steals 1% of coins" problem.

If bitcoin was a flavor of delegated proof of stake consensus, like Flying Fox, then the sidechain attack to steal 1% of the coins would fail.
If some miners started participating in an attack like this, the users would stop delegating power to those miners.
Ambiguity isn't enough to protect the miners.
Users will demand the miner include a transaction from the designated frozen list. Failure to comply means the miner will lose their power.
Even if there is only a 10% chance that a miner is participating in an attack, that would be enough justification for people to stop delegating power to that miner.

Paul has made a very convincing argument for adding DPOS characteristics to blockchain consensus.

I don't trust the price on that website because they don't have any markets. Even gatecoin is dead now.
I have consistently been selling my rep at a lower price than that, and I am not finding as many customers as I want.

I can't afford to pay anything above 0.003 for rep, because I already have too much, and am struggling to sell.

A reporter is someone who owns votecoins for a branch.
Each branch has a different flavor of votecoins.
If you want to be a reporter for a branch, then you buy some votecoins for that branch.
Votecoins are similar to reputation from the Augur project.
Augur had a crowdsale to distribute their reputation.

For a branch to decide who won a bet, it asks the votecoin holders.
The votecoin holders participate in a game designed to extract the truth from them.
This game is more accurate if there are lots of people owning small amounts of rep rather than a few people owning a majority.

Since the Augur rep is worth over $100 million, I would be very surprised if Paul Sztorc decides not to have a crowdsale.
Hivemind has the bitcoin community behind it, which is bigger than the ethereum community.
The world record for a crowdsale was $113 million.
Hivemind could raise much more than that, which would bring a lot of attention to the project.

That is such a large amount of money to refuse.
But, his motivations in building truthcoin so far has been ideological rather than financial.
He wants us to be wise and honest.
If the votecoin crowdsale somehow interfered with his goal of improving the world, he might avoid the crowdsale.

I plan to have a crowdsale to distribute the votecoins for Flying Fox.

edit: Paul doesn't actually have rep. sorry.

I'm pretty sure Paul Sztorc has over 10,000 rep.
Unless he already sold them.

I would pay 10 bitcoin for your rep.

All you have to do is write a comment here saying you want to sell to me, and write a bitcoin address that you control, I will handle everything else.