A problem common law was used to address in the past
June 18, 2017 12:57 PM   Subscribe

Ethereum is a blockchain, a "decentralized platform that runs smart contracts: applications that run exactly as programmed without any possibility of downtime, censorship, fraud or third party interference." In a phrase, use algorithms to replace contracts, or "code is law." But what if there was a bug that let let someone extract $53 million and walk out with it?
posted by the man of twists and turns (61 comments total) 12 users marked this as a favorite
 
So which kind of money that the people who have money don't consider to be money will turn out to be money -- the money that isn't money that the thief stole from the original people who tried to invent a new money, or the money that other people who tried to invent a new money decided was a better new money than the old new money that the thief stole?

And will any of them ever cash out their money that may or may not be money and receive, well, money?

And how much is any of this worth in tulip bulbs?
posted by delfin at 1:14 PM on June 18 [29 favorites]


This happened like a year ago, right? And then the Ethereum administrators wound up having to do a hard reset of their whole system which fundamentally undermined its credibility as a proof of concept? Am I remembering right?
posted by Sokka shot first at 1:15 PM on June 18 [10 favorites]


"While you would think that the sane, sensible thing to do..." The entire point of the exercise is that there is no room for "sane, sensible" things in a Distributed Autonomous Organization. This is supposed to be pure "sanctity of contract".

These are Galtians being hoisted upon their own petard and they like it not.
posted by pdoege at 1:22 PM on June 18 [15 favorites]


The one thing that I do find fascinating about cybercurrencies, whether they're called ether or Bitcoin or flooz or beenz or dogecoins or quatloos, is how they play upon the very nature of money itself -- something is only a currency if you can convince someone else that it has value, and that others will accept it as such. What makes the dollar more stable than wampum, cowrie shells or your Metafilter favorites count?

The concept of starting something new to test these theories is a lot less nervewracking when it doesn't involve a quarter of a billion dollars of other people's Officially Recognized Money.
posted by delfin at 1:26 PM on June 18 [7 favorites]


An operational definition of "hubris."
posted by Thorzdad at 1:37 PM on June 18 [4 favorites]


There's something wrong when you say, "Lawyers create an elite oligarchy common people cannot understand. Law is written in stilted and confusing jargon, so small business owners find themselves victim to the strange machinations of lawyers and judges. Obviously, we need to replace laws written in English with laws written in Javascript or Haskell or Hoon enforced by an immortal decentralized machine that controls all money."
posted by ikea_femme at 1:50 PM on June 18 [82 favorites]


And yet something like money - mortgages was supposed to go though the county register of deeds every time the holder of the note changed and pay a fee.

To avoid paying a fee a thing called MERS came into existence. If the holder of a note doesn't register with the county isn't that supposed to prevent a claim in the future? How can MERS held paper then have the ability to stake a claim - judicial fiat if challenged?

How exactly does the whole MERS thing not mess up a clear title with anything MERS touched? Fiat? Ignorance of the issues? Some kind of agreement where one party doesn't admit fault and pay a fine and the other side of the agreement reached says 'we are deciding there is nothing actually wrong because, well a fine and its more convenient then a public fight'?

Money and debt ARE a social structure - is it a social good to stick with what 'the law says' and if 'the law' would make a MERS held mortgage to avoid the fee at the register of deeds become a debt no longer backed by the property? Or is it better to make sure the people who placed money in a real estate trust that existed because of MERS are backed by the property in the case of default?

Etherium/Bitcoin/dogecoin/whatever - if the social structure supports other things like MERS when it is something like a decoupled debt to a person's home or retirement income via REITs why not the same support for the [whatever]coin scheme?
posted by rough ashlar at 1:50 PM on June 18


Also, it's interesting how these communities aren't just libertarian, but that strain of libertarianism that argues anything that isn't "true capitalism" is living on borrowed time, which means that since the federal reserve isn't true capitalism, the US dollar and all other popular currencies are doomed. I'm not sure if it's more a sales pitch (buy while your dollars are still worth something!) or a rationalization (Yes, there's not a compelling need for this, but what if everything else died?).
posted by ikea_femme at 1:54 PM on June 18 [4 favorites]


The blockchain "contract" was always silly as a broad proposition.

As a narrow proposition for a contract which is summarized: "Upon the occurrence or non-occurrence of A by B date, C will pay to D an amount of bitcoin equal to E which C already has and will restrict" it can work but ONLY for types of event "A" which both parties agree can be capably determined to have (not) occurred by the blockchain algorithm.

The VAST majority of contracts, of course, rely upon assessments of events which cannot be made by a blockchain algorithm AND involve sums of money which the contract payer either doesn't yet have or can't afford to lock up until payable AND cannot lawfully or practically be denominated in a blockchain currency.
posted by MattD at 1:55 PM on June 18 [26 favorites]


We're going through another GPU shortages right now due to a cryptocurrency - this time it's Ethereum. Newer AMD GPUs are skyrocketing in resale price and impossible to find because they're particular efficient at mining Ethereum. To the point where the $300 CAD GPU I bought 2 months ago is now selling used for $500-600 CAD on eBay. I'd considering selling mine if I didn't actually use it every day.
posted by thecjm at 1:57 PM on June 18 [2 favorites]


I find this subject easier to understand if you remove the jargon-
cryptocurrency = coupon
blockchain = database
contract = automated payment
posted by bhnyc at 2:14 PM on June 18 [12 favorites]


Is it better instead of "Lawyers create an elite oligarchy common people cannot understand. " one instead goes to out how (some, most?) Judges treat lawyers different than common people in their courts?

Now, most of the time the not-lawyers walk into the courtroom 'not understanding the rules of the courtroom' and of course they get beaten badly. But humans are human and as they are "pack animals" and holding that bar card means you are part of a group the other is not. And "not in my group" people ARE treated different than "in my group" people are.

Rather than learn the rules, document the actual rule-breaking and documenting the utterly crappy job the profession does of self-regulation and the broken nature of the court system and try to come up with a fix within the system these efforts of 'the blockchain' are their answer. Other expressions of "the system is broken" would be people who are associated with Black Lives Matter saying 'you can't get justice'. And perhaps both BLM and blockchainers are right - there is no fix to the system.

If the money and justice system was working - there would be no traction for the blockchain or [whatever]coin because there would be no perception of need. Crisis seems to drive people to looking at alternatives like in Venezuela.
posted by rough ashlar at 2:26 PM on June 18 [1 favorite]


Smart contracts for launching nuclear weapons. I'm sure we will all sleep easier.
posted by Pembquist at 2:32 PM on June 18 [1 favorite]


But humans are human and as they are "pack animals" and holding that bar card means you are part of a group the other is not. And "not in my group" people ARE treated different than "in my group" people are.

More like folks without a lawyer don't know what the heck they're doing and just waste everyone's time. That's not a big problem in something like small claims, but if there's real money involved you're just being obstreperous. You can't just walk into an operating room and start cutting.
posted by leotrotsky at 3:04 PM on June 18 [11 favorites]


But humans are human and as they are "pack animals"…

I'm sorry, but that's nonsense.
posted by Anticipation Of A New Lover's Arrival, The at 3:11 PM on June 18 [12 favorites]


yeah, humans aren't "pack animals" they're social primates.
posted by Homo neanderthalensis at 3:25 PM on June 18 [7 favorites]


Even as a person who is very pro-technology, who is himself a software developer, everything I have seen about Ethereum thus far seems like the worst parts of the software developer mindset. To wit:

"Oh look, here's a complicated problem (contracts) with an imperfect solution that was come to over centuries of legal evolution, which relies on a specialized knowledge to assist with the understanding and execution of the solution. Obviously a better solution would be for me to assume I can solve this with software that will in no way ever be subject to bugs or poor design. Please pay no attention to the fact that now I'm just changing which specialized knowledge is required for this task, and the knowledge I've replaced the old one with is the one I specialize in."
posted by tocts at 3:39 PM on June 18 [32 favorites]


The idea that laws work like software code is the underpinning of smart contracts, and as such they are a deeply stupid notion.
posted by Pope Guilty at 3:50 PM on June 18 [29 favorites]


I think this is an article reporting on the event that Sokka shot first mentioned. Just wanted to clarify it's not a new theft.
posted by ntk at 3:54 PM on June 18 [1 favorite]


A blockchain is like a DB, but on the order of 100,000,000 times slower and with on the order of 10,000 times more power consumption, something like that. You don't have to trust anyone, though!
posted by hleehowon at 4:45 PM on June 18 [4 favorites]


As soon as you decide "I'm going to use a computer to solve this problem" you begin to distort the parameters of the problem into something digestible to a computer. You start to hack off or ignore the things that don't fit easily into code.

Bad as it is, the legal system can take into account the context in which the contract is taking place. Programmers often fail to account for every exception or edge case for people's names; why would we assume they could replace something as complex as contract law?
posted by JDHarper at 4:47 PM on June 18 [22 favorites]


I'm confused. Why would any user purposefully continue to operate in a block chain with a known exploit?
posted by cyphill at 4:51 PM on June 18 [6 favorites]


Disclosure: I hold on to ETH as a speculative asset. I am waiting to see real world use cases emerge.

I thought these articles are all well written and cover the nuances of both the idealism present in the community and the reality (as everyone here has noted, software =/= laws). My understanding of Ethereum was it was there to facilitate trustless interactions surrounding things like digital identity, supply chain management across multiple corporations/borders, and to be Bitcoin 2.0, all at once.

Only Bitcoin 2.0 has come to fruition. That's increased the value of my investment though!
posted by thebotanyofsouls at 4:54 PM on June 18


We're going through another GPU shortages right now due to a cryptocurrency - this time it's Ethereum. Newer AMD GPUs are skyrocketing in resale price and impossible to find because they're particular efficient at mining Ethereum.

This is because Ethereum was designed to be resistant to ASIC mining. For the past 2-3 years Bitcoin mining has been utterly dominated by customized ASIC chips designed to do one thing and one thing only - mine Bitcoin more efficiently and faster per watt/clock cycle.

This has made it nearly impossible for the average user to be a functional or efficient part of bitcoin network without burning their own money on electricity or hardware to support the network.

So in the case of Ethereum/Ether they designed it to be computationally difficult to optimize for use with custom processors like ASICs to keep the mining power and network support in the hands of regular users with regular computers to help prevent it from falling into the increasing mining centralization that Bitcoin has been experiencing, where big money investors and speculators are setting up truly massive ASIC based mining operations and consolidating mining/hashing/voting power. (Which, incidentally, is what is essentially fueling the complicated segwit vs. big block arguments that Bitcoin Core, Classic and Unlimited have been fighting over for about the past year or so.)

So for Ethereum/Ether the most efficient way to mine Ether is via plain old general purpose computing and a whole lot of it, which is why people are buying every suitable modern GPU they can get their hands on, because even at inflated prices all it would take to recoup their investment is a few Ether tokens at current USD values for ETH. (Which, incidentally, is skyrocketing in price.)

Something else to consider about Ether as a token/currency (not Ethereum the blockchain technology) is that it's theoretically or functionally pegged in value to bitcoin. Ether isn't supposed to have an intrinsic value. As I understand it, it's supposed to be a slightly fungible token that can be "spent" by buying hashing/computing power on the network. Say, to mine Bitcoin. Or to provide cryptographic services for a web or other application. You can redeem Ether to provide these services, whether or not you mined the Ether, bought it outright or earned in on returned value through a smart contract.

This is essentially what Bitcoin was experimenting with as a blockchain, in that instead of the value of a bitcoin being pegged to gold it was (theoretically) pegged to proof of cryptographic work. Except as far as I know there wasn't really any way to tap into that proof of work for anything useful outside of Bitcoin's blockchain.

But with Ethereum, theoretically you can use Ether to fund, say, time on a rather large distributed supercomputer to do actual work.
posted by loquacious at 5:23 PM on June 18 [4 favorites]


since the federal reserve isn't true capitalism, the US dollar and all other popular currencies are doomed

BUY GOLD MINE ETHCOINS NOW
posted by tobascodagama at 5:25 PM on June 18


I'm confused. Why would any user purposefully continue to operate in a block chain with a known exploit?

The exploit wasn't within the Ethereum code itself, but with the code in the DAO. The DAO is essentially a program or app that ran on the Ethereum network. If you didn't buy into that particular DAO and put your Ether tokens into it, you were essentially unaffected.

People are likely remaining in Ethereum Classic for a number of different reasons. They either believe that it will succeed despite the fork. Or they preferred the un-altered version for ethical or personal reasons. Or they like the price value as a speculative investment and they're hoping to ride low prices to high prices.

As I understand it - another thing that can happen when a blockchain splits like this is that someone can theoretically be left with valid coins/tokens on both chains, except now they are two different kinds of coins/tokens and can't be traded or spent within that blockchain.

There's nothing to stop someone else from perceiving value in either of those chains and buying them with a different coin or fiat currency from the user that holds them. So this speculation is definitely fueling support for Ethereum Classic as well as many other lesser-valued and market-capped coin systems.
posted by loquacious at 5:37 PM on June 18 [3 favorites]


So this exploit happened a year ago, and after that Ethereum blew up into the must-have cutting edge investment that my co-worker is trying to convince me to buy?
posted by clawsoon at 5:37 PM on June 18 [1 favorite]


From those articles, it looks like the bug was in the contract code itself, not the underlying Etherium code, which hadn't been clear to me before.

Which I guess it's good that the underlying system is OK, but does highlight the problem with putting your money in to software that needs to be bug-free on release because it is running on an immutable distributed system.
posted by ckape at 5:48 PM on June 18


yeah, humans aren't "pack animals" they're social primates.
posted by Homo neanderthalensis at 3:25 PM on June 18 [+] [!]


Takes one to know one, apparently!
posted by mwhybark at 5:49 PM on June 18 [2 favorites]


It was hot, the night we burned Chrome.
posted by runcifex at 5:56 PM on June 18 [7 favorites]


So this exploit happened a year ago, and after that Ethereum blew up into the must-have cutting edge investment that my co-worker is trying to convince me to buy?

He's probably talking about Ethereum (vanilla, main, whatever) and not Ethereum Classic. If so, yes. Again, the exploit isn't against Ethereum itself, but a program called a DAO that ran on top of it. The DAO was so large, though, that at the time it would have undue influence on the value of Ether.

Lots of people are jumping into buying Ether right now because they missed out on cheap Bitcoins. Someone I know has essentially tripled their money in about a month by going all in on this and asking to be paid for work in either BTC or ETH, but he's taking his BTC and buying ETH with it through exchanges.

The upside of this has been he has essentially been getting 2-3 times his contracted wages by running some mild arbitrage both against USD and other coins. Almost every weekend BTC has been taking a price dip on Fridays due to people selling off and taking profits before the weekend, likely because they're trading at work. And since he gets paid on Fridays - in a flat USD amount - but in whatever the current value of BTC is against USD.

And he wants to get paid with temporarily weak BTC. because come Monday evening or so the price of BTC has been going back up to last weeks mid-week high or better, effectively multiplying his paycheck. He also has multiple "hot" wallets linked to a debit card, so he can pretty much spend it like cash at any POS terminal. It's pretty damn slick and I'm more than a little envious.

And it's working. For now. And he has no misgivings about the fact it could all come crashing down, so he's diversified. (And, if anything, both BTC and ETH have been more stable than gold, and it doesn't suffer the same long term inflationary rates that the USD does.)

Sure, we could be talking about tulips, here. Or Beanie Babies. Or the stock market. But we're not.

The only functional similarities between those things and cryptocurrencies is that the only real currency is trust. That trust includes speculation, or trust in either the system as a stable enough store of value or an increase in value.

The mechanical differences between a blockchain and, say, buying gold or investing in stocks or tulip futures are actually quite vast.

This technology isn't going away any time soon. It's also probably going to replace centralized/fiat banking for a lot of people.

This is going to cause huge problems for governments and banks alike who have had a monopoly on currency for some time now. The only way a government could stop it now was if they banned cryptography outright, writing cryptographic code or private cryptographic exchanges between individuals.

There isn't really any feasible way to determine if a block of cryptographic text is a bitcoin transaction, the HTTPS traffic to a legitimate bank or an HTTPS enabled porn site.

And blockchain/cryptocurrencies are already affecting you whether you realize it or accept it or not. It's already being used to fund stuff software projects, music/art projects and other real world, tangible things. It's also likely that it's already being used to evade income taxes and transfer large quantities of what we generally call "money" across borders without the use of banks at all.
posted by loquacious at 6:03 PM on June 18 [5 favorites]


This is going to cause huge problems for governments and banks alike who have had a monopoly on currency for some time now. The only way a government could stop it now was if they banned cryptography outright, writing cryptographic code or private cryptographic exchanges between individuals.

Short of restrictive and oppressive regimes governments have never had a "monopoly" on currencies. The only reason we in the United States use the USD is because you have to pay your taxes in it. I assume this is the same in a lot of countries. There's absolutely nothing that Bitcoin or any other cryptocurrency changes about this status quo.
posted by Talez at 6:45 PM on June 18 [2 favorites]


loquacious: Governments don't need or want to ban these currencies. They need and want to tax them. As far as I understand how distributed block-chains work, that's pretty simple: Government discovers whoever saying "I accept payment in bitcoin (or whatever) at this wallet ID (or whatever)". Government looks up that wallet, which they can always do, because distributed verification, and says "Okay, everything going into this is sales income and is subject to sales tax (if applicable) and income tax". Government writes whoever a large bill. It is now up to whoever to prove the government wrong.

The reason this doesn't happen is because cryptocurrencies are a drop in the bucket compared to what happens in real-estate, food service, investing, etc... and there are are resource constraints that mean the Government prioritizes different things.
posted by Grimgrin at 6:51 PM on June 18 [7 favorites]




loquacious: Governments don't need or want to ban these currencies.


Some governments.

India, however, well, some things to creep you out and turn you into a quasi cyber liberatrian crypto anarchist whatnot:

1. All Indians are in a database that includes their biometrics.
2. The government is trying to choke off the entire cash economy.
3. The government requires you to show ID to get any SIM card including prepaid.
4. Ditto with satellite uplinks.

And finally: the government is in the hands of people with a prior track record of using databases to single out members of an ethnicity for mob violence.

How long before Gujarati Muslims, or low-caste residents of Uttar Pradesh or some other people are designated for exclusion from the normal Indian economy, and wind up very much in need of an avenue like Bitcoin to survive? I really hope the next few years show me to be a raving lunatic and a chicken little in this regard.
posted by ocschwar at 7:20 PM on June 18 [2 favorites]


loquacious: Governments don't need or want to ban these currencies


S.1214 is already pushing in that direction.
They don't need to outright legally ban it, they just need to hamstring it.

And proving who actually owns, controls or is responsible for an account can be difficult, as is identifying whom to send the bill to for taxes based on a simple wallet address. An offline wallet that isn't linked to a reporting exchange service is essentially disconnected from all identifying information.

And if that wallet is stored outside of the US and you're entering the US with the plain English passphrase to that wallet, is that entering the US with or without declaring those funds?

Is it already in the US if that value is stored on a blockchain which primarily physically exists in the US due to the nature of the majority participation in that network?

Less informed and more reactionary representatives have already made noises about attempting to ban or put heavy controls on cryptography. I would not bet against the idea that they would use cryptocurrencies as a wedge to attack and ban personal cryptography.

The historical track record of the US government and how much it gets involved with any financial activity at all is rather bleakly clear on this point. The intense friction and regulation of our banking system is the laughing stock of most of the rest of the world, including many previously third world countries.
posted by loquacious at 7:38 PM on June 18 [2 favorites]


loquacious: At least in Canada, the government doesn't have to prove it to send you a tax bill. They can just assert it, assess the tax, and it's up to you to make out a credible case to disprove it. Failing that, you owe taxes.

This is before we get into the fact that the revenuers have powers of investigation, search, and seizure in the enforcement and collection of taxes that are so broad as to be quite literally unconstitutional to use to build a criminal case.

As for jurisdictional issues, the only questions are "is it worth something? and do you control it?" If the government can get a yes to the first and a strong probably on the second, the government can (usually) tax it.

I'm focused on the tax dimension because that's what I know, and, bottom line that's what I think matters. Libertarians like to suggest that cryptocurrencies will cause governments to collapse because they can no longer collect taxes. And that seems to be part of the attraction of the project. I think this is another situation where the issue they're trying to address in software is not the issue that exists in the real world.
posted by Grimgrin at 8:24 PM on June 18 [4 favorites]




The VAST majority of contracts, of course, rely upon assessments of events which cannot be made by a blockchain algorithm AND involve sums of money which the contract payer either doesn't yet have or can't afford to lock up until payable AND cannot lawfully or practically be denominated in a blockchain currency.

But that won't be true for long. Did you sell me the car? This blockchain from the DMV says you did, and now only my blockchain key will turn it on. Will you really have the money to pay for it next Tuesday? That blockchain from your HR team says you will, and that you've already committed to give that money to a trusted escrow agent. Is that money mine now? The money's blockchain says it is, and the Master Computer is so confident of that it's going to let me spend it now, not next Tuesday. Did you try to screw me? PayPal's BlockBuddy's blockchain guarantee has me covered, and in a few seconds their assassins will be knocking on your door to blockchain your organs for a wealthy Brazilian benefactor.

There's something wrong when you say, "Lawyers create an elite oligarchy common people cannot understand. Law is written in stilted and confusing jargon, so small business owners find themselves victim to the strange machinations of lawyers and judges. Obviously, we need to replace laws written in English with laws written in Javascript or Haskell or Hoon enforced by an immortal decentralized machine that controls all money."

If you trust the machine, then there's nothing wrong with this at all. 'If the machine says something is yours, it indisputably is, and the machine will enforce this for you' sounds much better to lay people than 'well, it depends on what lawyers you can't afford say about the matter, and even if you win, you might not get what's yours anyway'.
posted by obiwanwasabi at 10:57 PM on June 18 [1 favorite]


Ughhh. This is like the perfect example of why courts exist and law is not a simple matter of inputs and outputs:

Did you sell me the car? This blockchain from the DMV says you did, and now only my blockchain key will turn it on.

What if the seller misrepresented the condition, legal status, physical location, or other attributes of the car? What if one or both of the parties was not able to consent to a contract (under duress, underage, drunk)? One or all of these would likely (IANAL) make the contract of sale voidable (and well should), but the blockchain won't know anything about that.

That means in these cases it has to get brought in front of a judge of some kind, which means that any such transaction could be brought in front of a judge, and we're back to square one.
posted by thegears at 3:28 AM on June 19 [12 favorites]


Did you sell me the car? This blockchain from the DMV says you did, and now only my blockchain key will turn it on.


Unfortunately half of the miners decided they don't like how you smell and forked the currency and now it belongs to someone else.
posted by empath at 3:40 AM on June 19 [3 favorites]


Unfortunately half of the miners decided they don't like how you smell and forked the currency and now it belongs to someone else.

I know you're being facetious but there's a difference between blockchains, proof of work, and mining. Any government/commercial use of blockchains will most certainly not involve miners.
posted by Talez at 6:56 AM on June 19


So for Ethereum/Ether the most efficient way to mine Ether is via plain old general purpose computing and a whole lot of it, which is why people are buying every suitable modern GPU they can get their hands on, because even at inflated prices all it would take to recoup their investment is a few Ether tokens at current USD values for ETH.

If buying a video card was a reliable way to make a profit, why would NVIDIA sell them? They'd make more money taking them straight from the factory to their own Ether-mining farm.
posted by straight at 7:07 AM on June 19 [1 favorite]


If you trust the machine, then there's nothing wrong with this at all.

So your target audience for this scam is people who have never used a computer before?
posted by straight at 7:08 AM on June 19 [6 favorites]


If buying a video card was a reliable way to make a profit, why would NVIDIA sell them? They'd make more money taking them straight from the factory to their own Ether-mining farm.

Because it's marginal, because AMD cards have the compute efficiency to make it profitable, because it's adhoc, because it's inconsistent, because the market can't bear the extraction of the amounts needed to make it worthwhile.

Take your pick.
posted by Talez at 8:06 AM on June 19 [1 favorite]


I mean I don't mine because I don't like engaging in speculative markets, preferring to look at the future applications of blockchain technology, but there's a gulf of difference between someone in their backyard or in rural China with very little overhead and a company doing things "properly".

That's where there's profit to be had.
posted by Talez at 8:07 AM on June 19


Reading what people who have no idea how contracts or contract laws work, or the problems associated with them, write about "smart contracts" must be what it's like for programmers to read non-technical people writing about computers as if they are literally magical boxes.

What if the seller misrepresented the condition, legal status, physical location, or other attributes of the car? What if one or both of the parties was not able to consent to a contract (under duress, underage, drunk)?

Exactly. The issues that cause contract disputes are usually exactly the kinds of things you can't put down in code: representations and warranties, indemnities, definitional matters, contextual issues, etc. If you have a services contract and something as simple as a dispute over the quality of the deliverables comes up, how would a smart contract handle this in a way that's any different from a regular contract?

It's very hard not to roll my eyes at the breathless, near-religious talk from non-lawyers on this topic. I attended an interesting CLE recently talking about the future of smart contracts, and the consensus seems to be in line with what MattD says above: there are a few limited types of agreements where smart contracts are an appropriate, and even superior, solution, ones with very clear, simple, and definable terms and metrics like a shipping contract. In the short term you will likely see these sorts of agreements become automated. More complex agreements will likely become a hybrid where the parts that can be automated are incorporated into a larger, human negotiated/written contract. There may come to be a stratification where high-volume, low-value contracts are largely automated while low-volume, high-value ones, and agreements involving unique or novel elements, continue to be written by humans.

The intense friction and regulation of our banking system is the laughing stock of most of the rest of the world, including many previously third world countries.
loquacious

Sheer nonsense. The rest of the world is not some anarch-capitalist libertopia where bankers run free from the evils of regulation. If the US system is a laughingstock, it's because our regulations are so disjointed and ad-hoc.
posted by Sangermaine at 8:15 AM on June 19 [7 favorites]


Sheer nonsense. The rest of the world is not some anarch-capitalist libertopia where bankers run free from the evils of regulation. If the US system is a laughingstock, it's because our regulations are so disjointed and ad-hoc.

If I didn't know any better I'd have thought they're referring to the whole domestic transfers between US bank accounts being a cavalcade of shit all the way down.
posted by Talez at 8:27 AM on June 19 [1 favorite]


How exactly does the whole MERS thing not mess up a clear title with anything MERS touched? Fiat? Ignorance of the issues? Some kind of agreement where one party doesn't admit fault and pay a fine and the other side of the agreement reached says 'we are deciding there is nothing actually wrong because, well a fine and its more convenient then a public fight'?

I really wouldn't hold up MERS as an example of success in abstraction.
posted by praemunire at 8:51 AM on June 19




If buying a video card was a reliable way to make a profit, why would NVIDIA sell them? They'd make more money taking them straight from the factory to their own Ether-mining farm.

This is exactly what some scammy ASIC-based mining equipment providers have been doing in the past.

They'll take pre-orders for some hardware, and then they'll delay shipping the actual hardware before employing it in their own mining operations before sending the used hardware on to the buyer, extracting maximum profits.

One of the things that has happened is that they delayed shipment until that version of the mining ASIC/hardware setup was nearly obsolete and unlikely to recoup the investment costs through short term mining, just in time to deploy their own next generation hardware... that was also pre-sold and paid for, and delay that generation from shipping to consumer miners.

It's basically the same thing banks, credit card companies and paypal does by holding on to your money as long as legally possible to earn as much income from it as they can before finally refunding you for an error, verifying an interbank transfer or other reasons to hold up the funds.

If I didn't know any better I'd have thought they're referring to the whole domestic transfers between US bank accounts being a cavalcade of shit all the way down.

This is precisely what I'm talking about. That and international bank transfers and attempting to bank outside of the US at all.
posted by loquacious at 3:41 PM on June 19


Any government/commercial use of blockchains will most certainly not involve miners.

A lot of those use cases would be better served by standard centralized services.
posted by empath at 4:05 PM on June 19 [4 favorites]


this is why lawyers won't be replaced by computers anytime in the forseeble future. human disputes necessarily involve nuance, and when we pretend they don't and try to abdicate the nuance in some attempt at futurism, it only bites us in the ass. in almost any conceivable dispute, at least one party will want a human arbiter making a subjective decision as to what is right.
posted by wibari at 4:53 PM on June 19 [4 favorites]


wibari: in almost any conceivable dispute, at least one party will want a human arbiter making a subjective decision as to what is right.

I've been realizing over the past few years that one of the primary sources of political power - as important as raw force - is a reputation for making fair decisions in complicated cases. It's a lot quieter in the historical record than the armies and battles, but it's fundamental. The longer and broader the reputation for fair judgment, the more willingly that taxes are paid and soldiers serve. Doesn't matter if it's King Solomon or the Common Law, the Confucian gentry or the magistrates of Rome. As goes the consistent delivery of fair judicial decisions, so goes the state.

To bring back round to the thread of the discussion: If one advantage of blockchains is that they eliminate power and enforce fairness, the very act of successfully enforcing fairness will cause power to accrue to them and the people who control them. That's the irony: The more that you can show that you're immune to power, the more of it you will accrue.
posted by clawsoon at 9:19 PM on June 19 [2 favorites]


A succinct definition of a libertarian:
one who is hell bent on rediscovering why societies have organized themselves the way they have, the hard way.
posted by Freen at 3:46 PM on June 20 [4 favorites]


"It's very hard not to roll my eyes at the breathless, near-religious talk from non-lawyers on this topic. "

One of the things that has been a really stark lesson in the couple years or so for me working with non-profits doing a bit of education program evaluation is just how over-rated the theology of big data is. I've worked with programs that were reporting into massive "data driven" evaluations, and had to pull data from them in return.

There is noise at every step, there are vague claims about outcomes, there are inconsistent definitions and all sorts of insane and skewed incentives.

But the biggest problem is just that quality data collection is incredibly expensive to collect, and (working with another MeFite with a public health background gave a lot of perspective) it would likely cost more to pay for quality data collection than the programs would actually be worth even if they were otherwise providing positive value (e.g. almost all of the value from a field trip would be immediate but the impact could only be known longitudinally, and longitudinal data is hard to get).

In theory, you could come up with a set of evaluative criteria that two parties would agree would govern the resolution of disputes between them after the conclusion of an exchange. That's what we do with contracts, after all. But not to get too high-falutin', it comes down to a fundamental epistemological question: can subjective judgments be transformed into objective judgments, like some sort of fundamental Fourier transform that turns bits into waves? Even with centuries of trying to make decisions as objective as possible, the evaluation of a totality of facts in any given case relies on endless, subtle variations of those rubrics.

I can believe that improvements in both the software and data collection will massively improve our ability to automate many of these decisions, through not only the improvements in technology but also the tendency of humans to bend themselves toward their tools. But the vastness of complication is massively under-estimated by people seeking to build systems to cope with it — it's like the ambition of clockwork automata realized as mechanical Turks. Impressive, but not Turing, you know?

I don't want to come across like I'm denying that this has any potential — it's just that this doesn't seem like it should be aiming to replace lawyers or courts, but rather the mundane back-and-forth of quasi-legal agreements that exist below the level of lawyers. Instead of being a solution for complex litigation, it's plausible to think of them replacing e.g. boilerplate threatening billing letters for a huge swath of people.
posted by klangklangston at 9:25 PM on June 20 [3 favorites]


klangklangston: One of the things that has been a really stark lesson in the couple years or so for me working with non-profits doing a bit of education program evaluation is just how over-rated the theology of big data is. I've worked with programs that were reporting into massive "data driven" evaluations, and had to pull data from them in return. ... the biggest problem is just that quality data collection is incredibly expensive to collect...

I'm working on a health data collection platform right now. I'm convinced that the whole thing would've been more cheaply done with printed and mailed survey forms, and then some Excel manipulations on the backend. The tooling reminds me of where printing was at in the 1500s: You need an experienced journeyman to labouriously set up the machinery, or you need to be using one of the small number of formats which the big printing houses have got down to a routine. Either way, it's expensive, and there's a tipping point where going back to an older technology and hiring a scribe might still be cheaper.
posted by clawsoon at 10:19 AM on June 21 [2 favorites]


In my imagination, smart contracts worked like this:
The smart contract says, "this money is escrowed right now in the system - it has no owner. If I get a (cryptographic) signature from both parties that they're satisfied with the transaction, then I'll release the escrowed money to the seller." Presumably, the seller will sign when they've shipped the goods and the buyer will sign when they've inspected them. The smart contract won't know what you're doing in real life before deciding to sign. But it also says:
"If I get a signature from both parties that they've negated the transaction, I release the money back to the buyer." Presumably, the buyer would sign when they ship the item back, and the seller would sign when they got it back and inspected it. Again, the smart contract doesn't know what you're really doing.
The contract might also specify an arbitrator. In that case, it would expect the arbitrator to sign a thing that doles out the money in a manner decided by the arbitrator.
And then there would have to be a clause that's like, "If we fail to reach a consensus by X date (where a consensus is defined as something signed by both parties or the pre-specified arbitrator), then the system releases the escrowed money to some pre-defined third party instead of either of the original parties." Or maybe the money is shredded or something.

And all the complicated, messy stuff can get sorted out in real life, and the system is agnostic to the criteria that actually make the decision to sign or not.

And maybe there's some checkpoints like there is in a house sale: We expect the buyer to sign that the house cleared inspection. We expect the seller to sign that, I dunno, the buyer got title insurance or something, etc.

But now it's looking like they're trying to accomplish much more complicated and comprehensive than this in their contract programs?
posted by Galaxor Nebulon at 9:28 AM on June 23


Cryptographic escrow is probably not a terrible idea? But, yeah, you'd still need to designate an arbitrator in the case that there's a dispute between the two parties, in which case you might as well just be using traditional escrow. But at least the happy path is faster, which is not nothing.
posted by tobascodagama at 9:58 AM on June 23 [1 favorite]


What if the seller misrepresented the condition, legal status, physical location, or other attributes of the car?

You mean stuff that's already listed on registration papers today and would be trivial to include in the chain? 'I'm sorry, but your 2027 Tesla is actually a 2025 model, because the DMV computer tells me so.'

That and, you know, most normal people inspect the car and take it for a test drive first. They don't say 'You know, I think I'll just buy it completely unseen. I can always take them to bloody court later. That always works out well.'

It's very hard not to roll my eyes at the breathless, near-religious talk from non-lawyers on this topic.

Sounds like something horseshoe salesmen said in the early 20th century. Your days are numbered, same as theirs were.
posted by obiwanwasabi at 9:57 PM on July 10


Ugh, my eyes just fell out of my head from all the rolling. Thanks a lot.
posted by tobascodagama at 8:44 AM on July 11


"You mean stuff that's already listed on registration papers today and would be trivial to include in the chain? 'I'm sorry, but your 2027 Tesla is actually a 2025 model, because the DMV computer tells me so.'"

LOL.

So, here in the states, pretty much every state has what's called a "lemon law," where if, after buying a vehicle, the amount of days spent repairing the car in some initial period exceeds a certain limit, the seller must buy the car back. What those limits are depends on the state, but basically every state has one. What jurisdiction the car is in, whether a defect in the car was severe enough to qualify, what remedies are necessary, what further state enforcement action (i.e. fines or confiscation), what the actual time spent in repairs — none of those things are "trivial" to include in a blockchain, especially since many states have limits on what can be contracted in their state — for instance, in Michigan while I was growing up (unless the lunatic right-wingers of the current GOP have changed them), Michigan consumer law trumped (some) contractual law, to the extent that while companies often made you sign contracts ceding your rights to refund or ostensibly placing your contract's jurisdiction in another state, Michigan law controlled. There were consumer rights in Michigan that could not be signed away, because it was ostensibly in the public's interest to see them enforced.

Even pretending that this could all be dealt with through DMV registration records, you're banking on 1) publicly accessible DMV records to confirm this, which has been a dubious assumption in the US since John Hinkley jr.; 2) the reliability of those records — a notable pattern in California is that when used car dealerships resell used cars, they have a significant float time on updating the records, and that includes the liens that underpin them, leading to people being arrested for driving cars they bought because the title hasn't been properly transferred, and 3) a level of detail that simply doesn't exist in actual car transfer papers — there's no information in a private sale on previous repair work that might compromise the sale value of a car, and even though dealers are supposed to disclose that, the way that's enforced is through civil lawsuits. For example, a huge number cars were flooded in New Orleans during Hurricane Katrina, and owners wrote them off. These records are largely based on insurance claims by VIN (vehicle identification number). But in many states, the VIN is only recorded by the state in sales — the repair history is not recorded by the DMV, meaning that people bought these scrap cars, refurbished them a little, and sold them in other states without disclosing that they'd been written off because of flooding (which does electrical damage and is often unapparent in casual inspection). Even then, with engine block serial numbers being the second identifier for most cars after the VIN, if you swap blocks on the car, the repair history may not match the VIN in some important ways — if your DMV doesn't also record the serial number for the engine block (most don't), you could have a refurb engine with problems swapped into your otherwise reliable car — something that, again, lawyers are usually necessary to deal with.

When you start to sift through this, you see that what seem like objective facts often require dealing with multiple subjective sources to ascertain — like that you might have to take your car to more than one mechanic to try to fix a single issue, like my wife and I did with a blown vacuum pump hose — and even then, you'll often need information from multiple sources that has to be weighed in context of local laws. Arguing that this is "trivial to include in the chain" is really naive.

For scale, remember that namespace is not a trivial, solved problem for programming.

"That and, you know, most normal people inspect the car and take it for a test drive first. They don't say 'You know, I think I'll just buy it completely unseen. I can always take them to bloody court later. That always works out well.'"

Not everything that's illegal (or tortious) to sell shows up in a test drive. How would you know the asbestos content of your gaskets from driving it around?

"Sounds like something horseshoe salesmen said in the early 20th century. Your days are numbered, same as theirs were."

Dude, c'mon. If he's a horseshoe salesman (a farrier), you sound like a patent-medicine shill trying to argue through the survivor fallacy — there are more grand new ways of doing things that don't pan out than there are multi-thousand year occupations that die out.
posted by klangklangston at 4:01 PM on July 12 [3 favorites]


« Older Online Mathematics Textbooks   |   A Sociology of the Smartphone Newer »


This thread has been archived and is closed to new comments