> but generally when a person is trying to evade some legal authority, it's probably not for reasons good for society
Reminds me of the "nothing to hide" argument, that only someone trying to commit a crime would need or want protection from the legal system. History paints a different picture.
In the case of smart contracts, where both parties (if I understand it right?) agree that the code defines the contract itself, it seems like saying "... but I made a mistake" (or it has an error) would be very hard to prove. It would be like if you had written in a 3 year no-questions-asked return period into a car contract, rather than a 3 day one, and then tried to litigate when someone actually used that.
This is not really generalizable. Arbitration clauses for example waive one method of recourse in favor of another method of recourse, both already accepted by the courts. I expect you couldn't replace that with trial by combat and expect it to hold in court, but you might be able to argue that a different resolution process with historical precedent would hold - at least you could test it in court.
https://www.pewforum.org/2013/04/08/applying-gods-law-religi...
Important to note that the courts already recognize and/or have precedence for this, and also there are limits to what you can shift here. Finally if the resolution process is not acceptable to one party there are ways to bump it back to the courts.
None of this applies to "smart contracts", which aren't contracts in the legals sense anyway. There is no (yet?) agreement by the courts that this is a valid resolution method, and any issues of contractual law brought before a civil court will be resolved by legal principles, not source code. I suspect that the best you could hope for in the current setting is that the source code + associated communications speaks to intent.
There's no reason to believe that it is impossible that US courts and/or legislation would at some point give some legal status to a smart contract beyond above, but that's not the case today as far I know.
Or something like that. I doubt that's how it will go down here, but it will take a human judgement call or agreement-- not a coded contract-- to resolve this.
> but I made a mistake" (or it has an error) would be very hard to prove
I don't think it would be that hard to prove that the implementation of a smart contract doesn't match the clear intent. The bigger issue isn't the disputability of the contract, but the difficulty of identifing who you would take to court.
The parties might "agree", but who cares? If they are in the US, for example, and the two parties have a smart contract that breaks US contract law, one of the parties can file a lawsuit and attempt to get their money back. You can't go to the judge and say "sorry, the code is the legal authority here".
1. A normal contract (legal or illegal) that requires outside enforcement in the first place to force parties to comply.
2. A smart contract (legal or illegal) that executes itself without outside enforcement and can be overturned later (not literally overturned, but a subsequent transaction can be forced) by meatspace mechanisms
On (2), sure we can find ways to have the execution fail. In fact, anything where there is not fully escrowed payment/collateral/etc. can fail to execute properly if the other side does have not what it needs to deliver/or does not make it available on chain.
Of course, people are very careful prior to entering those contracts, because they know how big of a headache it will be if enforcement is needed!
> In fact, anything where there is not fully escrowed payment/collateral/etc. can fail to execute properly if the other side does have not what it needs to deliver/or does not make it available on chain.
Well, sure, if you write a contract that makes it possible for one side not to pay up, then that might happen. Having software run escrow is basically the whole point..
If you want smart contracts to be only applicable to very narrow sets of problems so be it, but otherwise you need to be able to allow, for example, unsecured lending and highly uncertain payoffs at T0 (staying in the finance domain)
Ok..sure, but I think it's sort of pedantic to bring up a class of contract that, obviously, nobody in this thread is talking about. It's a bad example anyway; even if you are being pretty careful it's simply not a risky transaction and therefor out of scope for complicated enforcement mechanisms like smart contracts.
I don't think smart contracts are very well suited to unsecured lending, at least not with available software. There would be no incentive to pay it back without some mechanism to force collections. Collateralized loans, however, is a great use case that exists already.
This is early stage tech, the scope is pretty small. I don't think anybody is arguing the contrary.