Right so it's done by people, following a set of policies, all of which ultimately are subordinate to written laws. So like what's the alternative you're suggesting?

Smart contracts enable this mechanism to be pre-agreed upon and have the authority automatically and securely deferred to the pre-arranged arbiter. This improves transparency, reduces transaction costs, and starts the process of commoditizing oracles/arbitration.

So instead of a lawyer or arbitrator we get another kind of arbitrator, or an "oracle" which is at root is a human making decisions and putting them into a digital system of some sort.

Both of which, like every other extant thing, are subject to review by the courts if their actions make people sufficiently unhappy.

And, though now it's confusing, opaque, and unreliable, this system will be cheaper and more transparent because...

Because it limits the scope of the arbiters responsibilities and capabilities, and precisely specifies their purpose.

Those benefits you're describing are the result of a writing a contract, not of embodying the contract in a digital construct [edited for clarity]

Sure. What is your point?

EDIT: I guess i'll respond to your edit with another edit. I understand your point, that you can say the same things using a traditional contract, but they don't have the same force. For example, the smart contract can automate and execute the disbursal of funds based on the attestation of some pre-defined oracle. While it's true that this oracle is not decentralized, it is now isolated and scoped. It's role is proscribed to solely this - attestation, and everything else is actuated around that oracle, in a secure and trustless way. I believe that the ability to circumscribe and scope responsibilities in a computational way has valuable benefits, insofar as it commoditizes and modularizes the processes involved. I think of it like a shipping container. Shipping containers commoditize and modularize shipping. Smart contracts commoditize and modularize adjudication.

So what does the smart contract bring that the regular written contract doesn't?

All of my responses in this thread have thus far been enumerating them.

As yodon explained, those benefits are handled just as well by ordinary contracts. I've still never got a straight answer from anyone about what smart contracts _actually_ bring to the table.

See my edited response to yodon. I'll also add, that they free up innovation in a valuable way. Take Augur, for example. If you're not familiar, Augur is a decentralized prediction market. In order for a non-crypto startup to create something like a prediction market, they'd need to develop serious banking relationships with banks. The banks could simply refuse to work with them, charge them exorbitant fees, and otherwise encumber the process.

However, since Augur is built on the Ethereum network, they didn't need to ask anyone's permission, and they don't ever need to have custody of anyone's funds. This is an example of 'permissionless innovation', to use the buzzword of the day. Permissionless innovation is understood to be one of the most important drivers of economic dynamism, and right now any sort of application that touches the financial sector is heavily burdened by the fact that our financial system is predicated on trust. Smart contracts solve that problem. This is a very subtle point, but it is extremely significant, in my opinion.

> The analog is Visa. What would ordinarily be a huge lawsuit is settled by a private commercial arbitration process--namely, a chargeback--in case of dispute.

Which remains a process with human-based wiggle room.

Yes. And that is a feature in my view.

You should read some of Larry Lessig's work on this topic. If every law was enforced strictly to the letter, the world would be a very different place than it is now. You would get a ticket for going 0.1 mph over the speed limit. You couldn't spend $1 over your credit limit on your card. Your landlord would penalize you for being 1 day late. Every cancellation policy at every hotel would always be rigidly enforced.

The point is lowering the cost in the common case while still providing recourse in exceptional circumstances. That, in my mind, is still a huge net positive.

Visa is a payment network, not an issuer (like Discover or AMEX). They don't issue credit cards to consumers, they just act as the underlying technology in the five-party model connecting merchants to consumers and their banks and the merchant's bank. Chargebacks are disputed with the issuing bank, not Visa...

I know this. But Visa sets the rules of the game. They make the rules, everyone else participates.

Visa isn't really a technology company. That's how they started, but these days, I think of them as more of a "facilitation platform" responsible for getting everyone into one legally sound network and defining common standards for things, everything from data interchange to how long customers have to request chargebacks, how long banks have to respond, what are valid reasons for revoking an authorization, etc.

My point is that the courts make the decisions, not the state of a blockchain or any other computer system. Since a blockchain is just like any other computer system with regard to legal capabilities, using it is a net negative because of all the well-known problems of a blockchain system.

Except the merchant loses the chargeback disputes the majority of the time. Chargeback is great if you are a consumer and a nightmare if you're a business.