The Cryptocurrency world has been excited by so-called “smart contracts” that use a blockchain, which are imagined to be revolutionary, disruptive, and darn near magical. This will, some say, enable Decentralized Autonomous Organizations (DAOs), which can replace all those silly corporations and governments with code.
This is certainly a real technology, if not quite as new as enthusiasts might think. But reality is far less magical than fantasies about “disrupting money”, and saner voices are being heard, in the fevered forests of Bitcoinland.
A recent piece by Michael Mainelli in Coindesk explains “Why Smart Contracts Need Shrewder People”. The main point, as I have said many times, is that so-called “smart contracts” are actually “executable contracts”, which means that they are software, code and data. The main difference from conventional contracts is that “smart contracts” may be designed to work without human intervention—for better or for worse.
Mainelli offers a good description of the kinds of contracts that are being created with this technology. For example, “Ethereum hopes to satisfy complex contracts in areas such as betting, mortgages and insurance.” (Call me a dumb old programmer, but these are not especially complex contracts to execute, as far as I can tell. Check out a movie deal in Hollywood if you want to see a complex contract.)
Of course, he correctly suggests that these very routine and formulaic transactions (e.g., selling insurance), employ a lot of people who are doing very little difficult decision making. Perhaps they can be replaced by algorithms, but that’s not an unambiguous boon, is it? (As annoying as insurance sales people can be, they are human and they are my neighbors. Replacing them with some offshore algorithm is not necessarily a formula for better customer experience, or necessarily good for my home town, IMO.)
Mainelli reiterates the point <link>> I made earlier that most interesting contracts are about the real world, and therefore must rely on some kind of data to determine if the terms have been correctly executed. Even if the code is totally correct (which is unknowable), the results are no better than the data—and that can be impossible to even obtain, let alone trust. (One of the good things that human agents do is apply common sense to try to work around unknowables.)
He also comments on the interesting issue of ‘How do I know the code will do what it says it can do?’ And ‘When the code isn’t doing what I wanted it to do, how do I stop it, and if needed, move the problem into the shrewd hands of experts, mediators, arbitrators, and lawyers?‘ Enthusiasts may imagine a world where code is perfect and conditions clear cut, but, as Manielli says, this is often not the case. And, by the way, contracts that are not recognized by any real world jurisdiction are, well, not going to be adjudicated by any real world entity. Who would tie up substantial resources in a contract that might not be enforcable?
His conclusion is that simple, dumb, short term contracts will be the first to be implemented. And he gives examples including “token-based marketplaces” (e.g., coupons or concert tickets), voting processes, identity management (e.g., notarized document exchange).
These are exceedingly “dumb” contracts, and not very sexy. But these are useful and important, so let’s try to show that the technology actually works.
Speaking of “not sexy”, Pete Rizzo writes about “The ‘Unsexy’ Way Earthport is Using Shared Ledgers for ‘Blockchain’ Efficiencies”. He is referring to Earthport, which uses blockchain technology to implement cross border payments. They use Ripple, but they do not use any cryptocurrency—just the ledger and related software.
Using the shared ledger and cryptographic signatures is, indeed, cost effective they report. But this is scarcely an autonomous or a “trustless” system. The parties all know and trust each other, and they deal in “fiat” currencies. The benefit comes from the speed and low cost of the shared, mutually trusted ledger, not from anonymity or cryptocurrency.
Blockchain “contracts” technology is starting to get realistic, if “unsexy”.
My own suspicion is that public key cryptography and digital signatures will prove the most important part of the equation, offering the digital equivalent of a notary service. (No surprise there, we knew digital signatures were important when they first came out in the 80s.) The other pieces of the “blockchain” – a distributed, write once data structure, and digital tokens- are probably not important or practical for most applications.