5 ms·
> What are the killer smart contracts? Security tokens (i.e. actual equities that happen to exist according to a blockchain rather than according to several st
by derefr 8y ago
> What are the killer smart contracts?
Security tokens (i.e. actual equities that happen to exist according to a blockchain rather than according to several stock exchanges, but which are still managed by government regulatory agencies, just through the regulator's presence as another smart contract on the same blockchain) seem fundamentally sound. Even governments are getting behind them.
Also, non-fungible asset tokens. Better known as "deeds." (Right now they're mostly used for virtual collectibles, but fundamentally they're the same "possession of X is ownership of the thing X represents" objects that fit the government's model for e.g. real-estate law. They're just digital instead of paper.)
> How does the price reflect the price of computation?
Nobody's claiming it does, I don't think. The price is definitely mostly speculative value.
What I would assert is that it's not 100% speculative value. Even with every single speculator out of the system, the value of ETH would not crash to zero; it would crash to some very small nonzero amount.
And that would be just fine, as long as the value stabilized there, since you can still use an ETH that's worth pennies in USD—but is stable there—as a medium of exchange for buying and selling these other digital financial instruments. (And, in fact, an ETH that's worth only pennies is much better from the perspective of actually being able to deploy and run novel smart contracts.)
> Has someone made a more efficient EVM? A less efficient one?
There are quite a few such projects. Pretty much everyone building "a new blockchain" these days is doing so by forking the Ethereum codebase and adding their own features (e.g. Hyperledger Burrow), or by building something that's "heavily inspired" by Ethereum (e.g. Neo.)
There are also Ethereum side-chains aiming to optimize Ethereum computation by moving most of the non-trust-requiring parts of a computation to infrastructure that's not quite so decentralized—Golem (https://golem.network/ https://golem.network/), for example, which is something you interact with on the Ethereum blockchain itself, but where the resulting computations from this interaction happen partially off of Ethereum, and therefore are cheaper than an entirely on-chain transaction.
Of course, most people don't really need that inversion; they just build their product so most of it happens centralized anyway, and only the part that needs to be mass-auditable and not under single-party control gets put on Ethereum. This is, for example, how CryptoKitties is built: it's a regular web-app, where the only thing about it that's on a blockchain is the record of what user owns what kitty-UUID. Every fact about that UUID is just on CryptoKitties' servers, but those facts aren't really under contention—only the ownership is. That's the idiomatic way to build an app that stores data on a blockchain at this point.
- riskneutral 8y agoJust so you know, your entire description of how ownership of stocks is tracked is completely incorrect. I would think one has to first properly understand the existing financial system before proposing an overhaul that is supposedly an improvement.
- derefr 8y agoTo be clear, first, I should mention that the "security tokens" that everyone cares about, are meant to be a ledger for tracking the cap-table equity issuances and modifications for private companies; they aren't the equivalent of common-class stock of publicly-held companies. The key advantage of them over the current model is that they make it much easier for an accredited investor (probably a partner at a PE or VC firm) to invest in a private company. But, let's ignore that distinction and talk about "security tokens" generally, because they theoretically can be used as an equivalent of public-issue stock certificates (even if nobody really cares or is excited about using them this way, right now.) I compared security tokens to "equities", and that word has a set meaning in finance: the equity class of investments—i.e. shares of companies. But really, there is no precise concept that is a 1:1 mapping for the thing that a security token translates to, which is probably why you're upset about my description of "how equities work." You can loosely compare a security token ledger-entry account balance to a physical stock certificate held by a person. But when you want to be precise, you have to get into the fact that security tokens are multi-jurisdictional—since they represent the cap table of a company or SPV directly, and because regulators of security tokens require full KYC to apply, both domestic and foreign investors must appear directly on the table, without proxy representation. Which is to say, if you want to sell your security token in the foreign market (i.e. on a securities exchange that operates under foreign regulation), then you must list the very same security token, rather than creating a jurisdictional proxy token, as one would if they wanted to list e.g. a public American company on the Canadian TSX. From this perspective, the concept of "how is equity ownership currently tracked" looks a lot less like a centralized regulatory scheme per equity, and a lot more like a distributed system where {exchange, regulator} pairs reach eventual consensus over ownership. (And this distributed paradigm has real consequences, both today and historically, despite law using the simpler single-regulatory-regime-per-equity model. For example, multinational corporations have shareholder meetings where they vote on things. How are these votes counted? Well, mostly, you have to let the state of the trades keep on marching on a bit after the vote, and then look at the historical ownership according to each exchange at the moment of the vote, and use that to weight each vote. The "preliminary" results of a shareholder vote can—and sometimes have been!—"rolled back" because the eventual consensus of the exchanges as to the ownership-at-point-of-vote was different than what the company thought it was when running the vote.)