Elon Musk on complicated ideas

[psztorc]

(1:08)
https://www.youtube.com/v/MJHTY0gWOGw?start=600&end=668&version=3&autoplay=1

I really enjoyed hearing this. Truthcoin is very frequently explained as "SVD". Which is really one of the least important, most modular and replaceable pieces of Truthcoin. Nowhere near the most important.

Here's my list of some of the other pieces which just happened to "luckily" fall together:

2] The second coin-type, which allowed people to "buy into" something, and allowed the protocol to take something away from lazy-laborers or malicious-laborers (making cheap talk expensive, making laziness expensive). It's function as equity counters the previously-uncounterable "exit scam" incentive.
3] The boiling-down of our definition of "true" to a definition that is actually useful (what most of a group of people will believe in the future, adjusting for inconsistency).
4] The concept that three things were conceptually (and literally) the same number: reputation-coins/equity in the laboring corporation, influence in the laboring corporation, and weights in the weighted-covariance of the Vote Matrix.
5] The insight from mechanism design that certain groups of features (incentive compatibility, revenue neutrality, individual rationality) are -mathematically- mutually exclusive, and one must break, and that the fact that real world investors payed trading/brokerage/SEC fees implied that RN was the one feature to force to break, and it should be broken in Truthcoin the same way it is broken in the real world: by overcharging traders by a consistent multiple.
6] That fact "the Decision/Market Author always has complete control over the clarity of the question" implied that the Authors would have to be cut into the mechanism's revenues. To offset this influx of cash, Authors would need to -upfront- discharge (pay) a fixed amount of cash.
7] Some basics of finance, for example that the system would be completely unhindered by "slow" check-ins (ie, by the fact that a DJIA market would operate perfectly, even if it "only" checked-in with reality a "mere" four times a year [or less]). This implied that "Decisions" could indeed all be resolved at the same time...implying larger, rarer, "vote matrices" (a matrix which, finally, could be reinterpreted as a dataset on which to use SVD / multivariate methods).
8] That the reporting system, which ordinarily would not scale at all (everyone reports on everything?!), could in fact be intelligently managed (fee-limited quantity of Decision-slots, Branching with a "harmless fork" of the VTC) to scale in something approaching viral biology.

[zack]

I have talked with a lot of different people about truth coin. It is interesting how different people will get excited by different aspects.

* anonymous gambling for illegal markets
* predicting the future
* funding public goods
* sports betting
etc

I cannot wrap up all the interesting implications into a sound bite.
SVD is not as exciting as other aspects of truth coin.

[psztorc]

I was talking about the design problem that I tried to solve.

But I do also agree that people enjoy the applications for completely different reasons. Sometimes, I'll be explaining it to someone one day, then the next day I'll begin the same explanation and it will end in a completely different place.