[sandworm101]

How many of the TheDAO Curator members are lawyers?

Contracts are agreements that are meant to be legally enforceable. The enforcer has always been the King, a local governmental authority and a third party. The very concept of a contract assumes the neutral third party. That third party is to interpret the contract, identify potential scoundrels, nullify illegal contracts and generally make sure everyone isn't playing games. Smart contracts seek to sidestep that ancient structure by replacing the neutral third party with an inflexible machine. Good luck with that.

Contract language is also meant as a manifestation of intent. Smart contracts seek a perfect manifestation, dismissing all notions of imperfect knowledge or misunderstanding. Typos rarely matter in real contracts. Intent can trump language where appropriate. But in smart contracts typos are everything. Good luck with that too.

[tdaltonc]

I'll make a prediction about where this is going longterm. Smart contracts will take off despite their rigidity but we will start writing them to include common law arbitration protocols. The arbiters will initially be 1 or more human but as the block chain begins to evolve it's own common law(s), fuzzy machine arbiters will emerge to fill the "incompleteness".

[sandworm101]

>>... to include common law arbitration protocols.

That's a great example of the sort of ambiguity that an arbitrator must deal with. Do you mean to refer to "the common law" as in the body of law descending from the brits, or do you mean the common law principals of precedent and authority? Or do you mean only the concept of precedent by which past decisions under similar fact patterns are used to inform decisions regarding new fact patterns. I have some assumptions based on your choice of language and my best guesses as to your cultural background, but good luck finding a machine capable of such understanding. If we crack that, then much of our culture is moot and the robots can be left to run things.

[tdaltonc]

I mean a "concept of precedent by which past decisions under similar fact patterns are used to inform decisions regarding new fact patterns." Without the need for territorial jurisdiction there will be many competing (but still distinct) common law groups. Signatories would pick a lineage at signing (npm install JAMS).

We'll start with human only arbiters, (panel of three from JAMS for example) then moved to mixed panels, then totally mechanical arbiters will handle most disputes once they're good enough.

[TheSpiceIsLife]

then totally mechanical arbiters will handle most disputes once they're good enough.

I'm not seeing it. Who's going to enforce the outcomes of machine dispute resolution? I mean, who's going to even want to sign up to the machine dispute resolution? Not me, no way, that's the freaky dystopian future we're supposed to be trying to avoid.

[MrRadar]

> I mean, who's going to even want to sign up to the machine dispute resolution? Not me, no way, that's the freaky dystopian future we're supposed to be trying to avoid.

Hell, that was a plot point in a 1987 episode of Max Headroom, "The Blanks", where a "blank" (someone who has erased themselves from all government databases, which is a crime in the world of Max Headroom) protests after she is tried by a computer that she has a right to be judged by a human but the prosecutor points out that as a "blank" she "has no rights" (implying that if she was not a "blank" she would have such a right).

[tdaltonc]

Business wants a reliable and predictable legal system. They don't care that the judge has a pulse. If the machine looses it and starts making terrible decisions, people would just start using a different arbitration group. You could probably even update old (well written) contracts to accommodate that.

[sandworm101]

Business wants a just and fair legal system.

Reliability and predictability are extraordinarily easy. For example: he who pays the most to the judge wins. That's very reliable, very predictable. It can be automated. Decisions can be taken instantaneously with absolute and perfect certainty as to who should be the victor. But it isn't just or fair, nor is it friendly to business.

[GFK_of_xmaspast]

Business wants a legal system that will rule in their favor.

[anonbanker]

I cannot wait until I can do UNIDROIT[0]-compatible contracts in Rust using Parity[1]. building an implementation of UNCITRAL Model Law[2] would be the biggest market disruption I could think of. Imagine being able to do ex aequo et bono (think Judge Judy/small claims rules) arbitration; even with human (flawed/biased) arbitral tribunals, this would be amazing.

ROSS is killing the need for lawyers, and this is the path to killing the need for judges. Software is eating the world, and i couldn't be happier.

0. http://www.unidroit.org/english/principles/contracts/princip...

1. https://ethcore.io/parity.html

2. http://www.uncitral.org/uncitral/en/uncitral_texts/arbitrati...

[lukesed]

>then totally mechanical arbiters will handle most disputes once they're good enough.

It seems to me that any such arbiter would need to have the practical ability to interpret the human intent behind imprecise/incorrect language and author/fix code. I think that would be a major advance in computer science and the most important application would be writing software.

[jrpt]

You're being pedantic. Contracts explicitly specify the arbitrator, such as AAA or JAMS. More info at http://arbitrationnation.com/arbitrationnation-roadmap-when-...

[sandworm101]

Yes but that is a flesh-and-blood arbitrator. We are talking about coding the arbitration process into a machine. Being pedantic is exactly what happens when one tries to express cultural norms in the exact terms used by mechanisms.

[jrpt]

If you could capture the entire logic in code, it wouldn't be arbitration, it'd just be a feature of the smart contract.

The way you'd encode an arbitration clause into a smart contract is making the arbitrator an oracle. That makes it a part of the smart contract, but it's still resolved by an external arbitrator.

[kwikiel]

That's fair point.

""So I’ll start there, and imagine that there are semi-trusted ‘oracles’ that compete to be the most reliable and trustworthy verifiers of contracts. People involved in contracts choose N of them, and then require that contract conditions be validated by one or more of them before the contract pays out. Pick more than one so no single oracle can steal the contract’s funds, but less than N in case some of them go out of business or just aren’t around to validate contracts when it is time for the contract to pay out."

http://gavintech.blogspot.com/2014/06/bit-thereum.html

So you can "just" write contract in some language so that: - it will cover most of cases without human intervention - if there will be bug with code oracle will read (as human) content of terms to get idea about contract intent (basically dispute)

But it seems that good marketing and some really interesting technology can switch focus from asking basic question - "we are trying to solve existing business problem or just create cool tech?" And yet when i read blogposts about ethereum i see more and more complicated contraptions (ruby goldberg machines in fact) and reason why this whole thing is even build is not clear.

[naasking]

Right, the point is that the arbiter can evolve independently of the contract logic (via establishing new precedent). The contract would just specify a unique, unforgeable designator for the arbiter to use.

[tdaltonc]

Am I correct in thinking that a smart contract has to be open source? Because there might be good reasons to keep the arbitration oracle's source code secret.

[hammock]

Pretty clear to me that he's talking about creating/evolving new common law, in the vein of the living body of common law descending from the Brits

[sandworm101]

>> in the vein of the living body of common law descending from the Brits.

Does the speaker recognize the various forks of that body of law? The US/Canada/Australia and others split many years ago and are now developing their own versions. I assume the OP is speaking of them collectively, but try asking an American whether Canadian cases should be relevant to US decisions.

[umanwizard]

The US has 50 versions, not 1

[sandworm101]

It;s actually much more than the 50 states. There are the territories and the feds. Then there is the precedent used by various boards and committees such as the NLRB. Canada and the UK have similar splits (ie scotland) but the biggest and most relevant divide is that between nations.

[tdaltonc]

Yep. This is what I meant.

I'd add that there could be many competing common law lineages and that contractees would selecting one when they draw up a contract.

[tdaltonc]

For example:

either party to this contract submit a signed request for arbitration within the escrow period of this contacts then

1) a panel of 3 arbiters from the New Atlantis Common Law Arbitration Group will be selected at random.

2) 50-ETH will be set aside for court fees.

3) The panel can execute any 1 of 5 events by submitting 2 of 3 signed tokens. If the panel cannot arrive at a consensus, one token will execute at random.

Many competing standard arbitration packages will develop.

[TheSpiceIsLife]

If the panel cannot arrive at a consensus, one token will execute at random.

Hang on, this isn't the casino. Contracts don't usually have a 'random outcome' clause, do they?

I think we better take this to the courts.

[sandworm101]

>> Contracts don't usually have a 'random outcome' clause, do they?

Like in a betting shop? When I lay my chip on "red" then I am entering into a contract with the casino. A random event can be a term in a contract, but you are correct in stating that in the case of a dispute we do not allow a coin flip on the "who is right" question.

[tdaltonc]

Whether the random number is pulled before the hearing (picking just 1 judge instead of 3) or after (by picking one vote token at random) seems the same to me.

[TheSpiceIsLife]

Maybe I'm missing what you're getting out? It seems like you're suggesting dispute resolution should have a "fuck it, let's flip a coin" option?

[tdaltonc]

I'm saying that it kind of already does. Judges are picked for a case partially at random.

[kolinko]

That's why smart contracts will be really lousy for things that traditional contracts are used for.

But on the other hand, there are some simple - yet important - things that can be implemented in smart contracts pretty easily. Say, futures, or stock options, or certain kinds of insurance. For these things smart contracts are great because they offer extremely cheap enforceability and speed.

This is a textbook blue ocean, or innovator's dilemma. The new tech is much worse than the old one in some cases (enforcing intent), but far better in some other ones (speed, enforceability, global access)

[gamblor956]

Typos rarely matter in real contracts. Intent can trump language where appropriate. But in smart contracts typos are everything. Good luck with that too.

The City of Cleveland and Frank McCourt would like to have a word with you about your novel theory. In real contracts, typos--like an errant comma--can be significant and completely change the meaning of the language. Cleveland lost the original Browns because of a typo; McCourt lost the Dodgers in part because of a typo.

[sandworm101]

In those cases the court found the intent manifest in the language. These were sophisticated parties with legal advice. Even if they court didn't see intent, it is within the courts' power to force such persons to follow the language if for nothing else than to serve as example to other negligent contract drafters.

A court will only grant that which is asked for. So for a typo to matter, one of the two parties before the court must be claiming that is isn't a typo.

[kolinko]

There are also cases of people selling cars (and other stuff) ultra cheap on ebay, because they misplaced a comma when specifying a price.

It was obvious that the price was supposed to be $20k, not $20.000, and yet the seller lost.

[sandworm101]

Prices for the sale of consumer goods are subject to some special rules, mostly to prevent bait-and-switch advertising. But actual typos happen all the time and are corrected. There are a great many instances of airline websites getting prices horribly wrong. These contracts are not normally honoured and wouldn't be enforced by a court (40,000$ tickets for 50$ sort of things). Much depends on the specific facts and whether the price is obviously a mistake.

[Mikhail_Edoshin]

I once read a story told by a young U.S. lawyer who happened to come from one of West European countries and thus had a specific cultural background. They were in the middle of finishing a deal and when he was re-reading the contract, he noticed that he made a typo and turned $1.5M they were to pay at some stage into $15M. And the contract was already signed! In horror he went to the business owner and confessed. It was a revelation to him when the owner took it very lightly and told him something like: "It's OK, I'll just phone them and we'll have this fixed."

(And it wouldn't stand a chance in court either, because a simple calculation would show this couldn't possibly be a right sum for a sane person to pay for whatever the contract was about.)

[wolfgke]

> Contract language is also meant as a manifestation of intent.

I challenge you to create a formalization of the intent and create a better programming language for smart contracts that includes intent. I could imagine quite well that intent could be formalized to some kind of "firewall" or "sandbox" rules for the smart contract, i.e. that if the execution of the smart contract violates these rules, it will, for example, be canceled by the system.

[sandworm101]

I would counter with the concept that any valid interpretation of intent must come from one with cultural understanding. Machines do not have cultural understanding. They cannot identify language/intent clearly in violation of cultural norms. Before attempting the intent code, I'd first need to see a machine capable of understanding why Shylock's pound of flesh was an illegal contract.

[jsprogrammer]

Machines don't understand. They simply compute. It is up to the programmer to have them perform the appropriate computations.

[naasking]

The fact that machines don't understand doesn't entail they can't understand given a suitable model for what constitutes "understanding". We simply lack such a model.

[jsprogrammer]

We can define words to mean whatever we want. The machines under consideration are algorithmic, defined by simple logical rules that allow the next state to be derived from the previous. Any understanding within a model comes from the algorithm, not the machine.

[jcoffland]

If the machine knows Asimov's three laws of robotics then it will know that Shylock's pound of flesh cannot be had.

[mrob]

Asimov's Three Laws aren't a replacement for the whole system of human values. His robot stories are full of robots doing strange things that conflict with human intuition but are aligned with the Three Laws. Asimov himself recognized that the laws were not sufficient, and added the Zeroth Law as a workaround:

"A robot may not harm humanity, or, by inaction, allow humanity to come to harm."

But this law is vague and difficult to interpret. The Laws of Robotics are not a formal specification of ethical behavior. From the point of view of a fiction author this is no problem. The ambiguities allow for more exciting stories. But when real human lives are at stake, it's a serious problem. "Friendly AI" is a non-starter if we can't even define what "friendly" means.

[Mikhail_Edoshin]

I think the whole idea of these stories was to find ways these seemingly simple and straight laws can come into amazing deadlocks.

[dclowd9901]

Intent is simply the will of the party manifest. If each party presents their will, there is no misunderstanding. Conflicts of will can be treated like merge conflicts and resolved in the meat space.

[alkimie2]

(I'm not a member of The Dao, although I do have some Ether. I switched from engineering to law in the early 90's.)

I'm sure there are some lawyers who got involved--I was tempted to get some experience with the concept early on. I decided not to primarily because the basic concept of crowdsourcing decisions where the ownership interest determines who has most of the voting power and the major holders are likely to be new at it strikes me as, frankly, a silly idea.

But even if all the participants were lawyers or represented by lawyers there are still going to be bugs. Pick up any set of service terms for just about anything and there will be logical flaws. It's just human nature. There is no such thing as a perfect contract.

To me the most interesting thing about this episode is that The Dao included limitations imposing an time lock on withdrawal that gave the community some time to consider what to do and the pros and cons re the whales in the Ethereum community doing a hard fork in the name of justice.

[sandworm101]

I don;t think lawyers could help with bugs, but their absence could be telling. Most of the lawyers I know wouldn't go anywhere near this sort of thing. A board of directors, or any other governing body, without any lawyers makes me suspect those that were asked fled the project. Or that the organizers deliberately didn't approach any lawyers because they knew what they would say. Anyone with a background in contracts, specifically dealing with contracts gone sour, would see red flags all over this concept.

[pash]

> Contracts are agreements that are meant to be legally enforceable.

You have hit on an important point, but like most commentators here and elsewhere, you seem entirely to have failed to recognize its significance. In the American and English tradition of common law, a contract is an agreement made with the intention that it be subject to legal enforcement [0]. Not every agreement is meant to give its parties recourse to legal remedy, and hence not every agreement is a contract. That is, not every agreement is subject to contract law [1].

It is of little significance that people in the Ethereum community (or in the broader crypto-asset community) have taken to calling their programs "smart contracts"; the name signifies nothing. In the main, and particularly in the case of the DAO, it has been made clear, ad nauseum, that these programmatic agreements are not meant to be subject to legal enforcement; in short, it has been emphasized from the start, and at every intermediate step, that smart contracts are not contracts in the legal sense. Indeed, the major motivation for the development of programmatic agreements (i.e., "smart contracts"), has been to supplant enforcement de jure by enforcement de machina.

The law allows for agreements that do not have the force of law behind them, and naturally the law says little about such agreements. For this reason alone, absent separate provisions which purport to invoke legal enforcement of the intent behind the code that implements a smart contract, it is perfectly reasonable, legally speaking, to argue that prima facie these agreements are not subject to contract law.

Thus the independence of programmatic agreements from legal constraints is legally plausible; this is a very different situation from the legally absurd arguments put forth by some crypto-zealots that, for example, transfers of cryptocurrencies are not subject to laws restricting money transmission.

Now, might a judge decide that any particular programmatic agreement—or "smart contract"—is in fact a legal contract? And that therefore the judge, not the code, ultimately determines who gets what? Sure. Judges tend to decide that they get to decide, when there's any question about it. But it's not unreasonable to imagine that a judge might really say, "No, this is not a legally enforceable contract. It says so right on the box." And if that happens, then what the code says goes.

0. https://en.m.wikipedia.org/wiki/Intention_to_be_legally_boun...

1. Beyond the determination of whether the agreement is legally a contract, I mean. In particular, I mean to point out that a so-called "smart contract" might well be legally deemed not to be a contract at all, and therefore not subject to the provisions of contract law that give precedence to the parties' intent in the agreement over its literal interpretation. In other words, if a programmatic "smart contract" is not legally a contract, there is no legal reason that the law should favor any outcome other that what the code's execution ordains, or that the law should say anything on the subject at all.

[wyager]

Nothing wrong in principle with having a mechanical third party. You just have to actually write your contracts correctly.

[arcticfox]

It seems simple enough to include a fallback mechanism to human judgment in times of extreme error. Maybe requiring 90+% consensus from members, or something like that. Obviously, TheDAO didn't have that, and explicitly precludes any fallback to traditional systems in its documentation.

[TheSpiceIsLife]

and explicitly precludes any fallback to traditional systems in its documentation.

Which is, of course, nonsense. You can't document your way outside the law. Not in any country I'd want to live or conduct business in anyway.

If enough of any one persons money is involved you can bet there are lawyers right now trying to working out who to sue.