Comment by pierre | Hacker News Reader

Comment by pierre | Hacker News Reader

If you are intrigued by the idea that law can and will be represented as computer code / data, here are a few link to go deeper : * British national act as a logic program (1982): the paper that is the based of most current effort in this domain https://www.researchgate.net/publication/234805335_The_Briti...

* Standford codex LSP initiative that try to standarize the format in which legal rule will be encoded https://law.stanford.edu/publications/developing-a-legal-spe... (as far as I know the initiative is still going, don't hesitate to contact oliver goodenough if you want to get involved)

* OECD recently published a full repport on current initiative in the domain https://www.oecd-ilibrary.org/governance/cracking-the-code_3...

* There was a workshop at last ICAIL where a lot of people showed their progress on the topic (link to all demo video in this gdoc : https://docs.google.com/document/u/0/d/1-7cJ0KsYzQ8IOY3L_bYX...)

* MIT started a journal on the topic ~1 year ago https://law.mit.edu/

If you want to work on this topic DM me :)

> and will be

strong doubt.

laws are by definition supposed to be „soft” and not hardcoded. see the smart contract hacks.

this simply doesn’t happen in law, because the spirit plays a large role, and the spirit is something that cannot be coded

I firmly believe that it will be. However it won't be anytime soon because this entails a transition similar to going from rule by king to rule by laws.

The law is (nowdays) is backed by writing, and since I think that computing is a next step in the evolution of writing (along with printing press) one day---likely in dozens of generations (hundreds of years)---the law will be backed by computing.

All these technologies change the nature of a human's consiouness; it will take hundreds of years. Look at Walter Ong's work for better made arguments along these lines https://en.wikipedia.org/wiki/Walter_J._Ong

You're missing the point that flexibility of interpretation and "execution" is a very much desired feature.

And you're missing the point that these things exist for humans

+1 Law is human-to-human agreement about a lot of vague soft stuff. Code APIs are computer-computer agreement.

Some exceptions to the above are: 1. Financial contracts (see ISDA derivatives). They're written with a big "human" document upfront and then there's a "notification addendum" attached to each use of that contract. 2. Master Sales Agreements (MSAs): The first MSA is a human-to-human agreement. Everything after that is order-forms. And negotiating the MSA requirements is very very human (risk, trust, effort, cost, benefit & promises). Order forms are pricing decisions that can be "automated" (especially around annual renewals if within budget without red flags).

Yeah, the more black-and-white domains tend to be financial and commercial – and in those domains, there have been some pretty cool illustrations: http://complaw.stanford.edu/complaw/readings/sla.pdf

> these things exist for humans

so does law.

As do all laws. You have a point about contracts though, assuming we admit contracts between non-moral parties to actually be a legally enforceable thing. If your would-be law/contract/??? can't get both sides in front of a court, one ought to disregard its terms.

One position I've heard is that cultural concepts that are used in legislation or contracts ("reasonable effort", "reasonably foreseeable", "undue risk") may always be assessed subjectively by humans, but that the logical structure of rules and conditions ("any of the following", "none of the following", "two or more of the following") may not be.

There are lots of drafting disputes about things where someone wrote something like

> a and b or c

leading lawyers to argue about whether this should be read as ((a and b) or c) or (a and (b or c)).

There's no reason that this kind of ambiguity should have been permitted to exist in the first place.

I think my understanding of this is related to my understanding of Lojban, which tries to avoid all syntactic ambiguities but explicitly says that it's a non-goal to avoid ambiguities related to the cultural meaning of words and concepts. Like if you say something is "medieval" or "fun" or "convenient" or "beautiful" or "fair" or "postmodern", Lojban doesn't try to make the truth-conditions for your statement objective with regard to what these concepts do or don't refer to. But to the extent that that's agreed between two people, they should then agree on what a particular sentence using these concepts means or doesn't mean.

Though I do envision that when people use better tools for avoiding parsing ambiguities in legal texts, they will still argue (and there will be many legal philosophers insisting) that they should still be permitted to argue that something was still a drafting error, because the (only permitted) interpretation under the drafting formalism is manifestly unfair or unreasonable and could not have captured their true intention.

Yes thank you. And the same misundestanding comes up when people thing Goedel's incompleteness results precludes mechanical theorem proving.

In these classic papers, lawyers excitedly discover formal logic. Worth a quick scroll just for the diagrams, which wil make you laugh.

https://digitalcommons.law.yale.edu/fss_papers/4519/

https://repository.law.umich.edu/articles/29/

https://drive.google.com/file/d/0B8axPLLxhjQFWGc0ellBYTF3RnM...

I feel like this argument is akin to saying that the measurement ruler will never take off because medieval worksites use "soft" local measurements. I think the long-term view, in my opinion, is that of course contracts will be clearly delineated with definite terms. We may still need juries to determine facts and application of these contracts to facts, but we should not have to try to understand the intent of the contracting parties- that's what the contract is.

Adding to this list:

* https://catala-lang.org/ -- DSL for translating/annotating legal texts

* https://arxiv.org/pdf/2011.07966.pdf -- A Modern Compiler for the French Tax Code

* https://www.youtube.com/watch?v=EshxZVMURt4 -- Formal Methods and the Law

Thanks Pierre and Paul for the links!

For folks who happen to be unreasonably interested in this stuff: there are research engineer positions open at the Centre for Computational Law at Singapore Management University where the bulk of the R&D is happening, in partnership between Legalese and the university.

TL;DR: Move to a tropical island, get paid to write open-source software, and explore the arguments being made in this thread in way more detail than you dreamed possible. DM me on Twitter, @mengwong

Reply on news.ycombinator.com