jury/reputation/voters

Previous topic - Next topic

zack

This post is a work in progress, it may change as info comes.
Zack is offering a 0.05 bitcoin prize to anyone who can prove that walled gardens are better before 1AM California time on Dec 12.
2 ways of implementing this have been described

1) Walled garden
*new juries have the same distribution of coins as an old jury, minus the people who choose not to accept the free reputations offered to them.
*the possible distributions that can be created are limited by the existing distributions and which people are willing to forgo their free reputation.
*the first crowd-sale for the first jury is special, there is never another crowdsale.
2) Free market
*new juries are completely owned by the person who paid for their creation.
*This reputation can then be distributed in any way possible.
*every crowdsale for every jury is the same.

zack

Augur is designing a walled garden so that the participants in the crowdsale will get maximally valuable coins.
The walled garden causes major unsolved security issues.
If we did it like a free market instead so everyone has equal power, then the security issues vanish.

In a free market anyone could show up at any time and compete with the participants from the crowd-sale.
If we do the walled garden, then competition wont happen, and the participants from the crowdsale act as a semi-permanent nobility.

They are breaking the software in an effort to make the crowd-sale more profitable, because they want to make more profit in the near future.

psztorc

I think, in this cryptocoin-space most of all, the "free market" phrase is quite misused. Free markets work because entrepreneurs are free to combine capital -however they like- to offer goods and services to customers. However, the customers themselves don't need any freedom other than the freedom to go to a competitor. Google doesn't give me a choice of search-algorithms, or even search-options (unless I specifically ask for a few simple ones)...it just finds what I want.

Satoshi's innovations allow us to freely combine code to make new pieces of software, but the software itself is highly restrictive: Bitcoin doesn't create digital scarcity as much as it prevents double-spending. Soft forks (which "add" functionality) are literally defined by restrictions.

So, unlike in entrepreneurship / development, and unlike in personalize-able-software (game-hotkeys, desktop background, social media profiles), I regard every single choice given to the user as a failure of the protocol.

In fact, over the last few days, for Whitepaper 1.4 (the last and almost-certainly final or semi-final version), I have been going through all the choices about fees, prices, branching to make absolutely sure they are replaced by high-quality feedback loops.

My current thoughts are for 1 initial distribution, copied into two competing branches, listing-fees as a step-function of decisions already added, and Voters using a >phi vote to [1] set the "Branch Rules" text for the far future, and [2] propose and agree to Branch into two daughter-branches. These decisions are to maximize stability and security in both the technical and economic domains.
Nullius In Verba

zack

https://www.youtube.com/watch?v=r8JopZWlvtw
The first 20 minutes of this talk is relevant. He explains how bitcoin is neutral to any contract mediator, and the benefits.
Why would truthcoin not benefit from this same neutrality?

Quote from: psztorc on December 11, 2014, 05:08:08 PM
I regard every single choice given to the user as a failure of the protocol.

If this statement was true as a general principle, then it would imply that the walled garden is better than free market.
Is it true in general? What else would this statement imply?
Letting people spend arbitrary amount of their money is bad, since they have too many choices. We should restrict them to only spending 10% at a time.
Letting people bet on arbitrary English text decisions is way too many choices. We should restrict them to a list of pre-approved things to bet on.
Letting people give text string names for reputation and prediction markets is bad, since it is too many choices. Instead we should use the unique hash to identify such things. Like transaction IDs in bitcoin.
Letting anyone bet in either direction is no good, since there are so many choices. Each person should be restricted to only betting in one direction.

I don't think "every choice given to the user is a failure" works as a general principle. It is similar to Occam's razor, and if you add an "except" in their, I think we can make it into a general principle.

psztorc

By user, I was referring to the "employee" input (voters/authors, not traders). It is clear from the context (branching decision) and from the distinction I drew between entrepreneurs and customers.

Users only need to choose what to buy, choosing how to make it is the entrepreneur's job.

If you want to maximize the blockchain's flexibility, just open a completely blank C++ terminal. With that you can build anything you want. But is useless to the user.
Nullius In Verba