News for hackers. Stuff that matters.
Re:Law for Computer Scientists (2020) by mo_42 on Sunday January 28, 2024 @04:07PM (#39169813)
When I was studying CS, I had to take some courses outside of CS and I took law. It was pretty fascinating and I even thought once or twice of switching.

I now have the very nerdy perspective that law is the operating system our socially run on. Laws are small snippets of code similar to a predicate in Prolog. We apply them once the conditions are fulfilled.

Re:Law for Computer Scientists (2020) by Terr_ on Sunday January 28, 2024 @04:28PM (#39169985)
When you include common-law and look at the broader cycle, I'd argue the legal system is more like a JIT optimization rather than source code: We formalize and streamline lots of existing implicit social rules and shared expectations, things that exist-before and can operate-without any formal legal text... Just not as well.

In other words, the relationship is more like recipes versus cooking. People cooked foods first-- possibly many many times--and then formalized them later with documentation and rules to help make an outcome transferable or fast.

However, it would be a mistake to think that just because recipes are ubiquitous, they are a prerequisite for the process.

Re:Law for Computer Scientists (2020) by geraldhh on Monday January 29, 2024 @10:09AM (#39177082)
thou nobody ever got put in jail because he defied the recipe
Re:Law for Computer Scientists (2020) by Terr_ on Monday January 29, 2024 @01:08PM (#39179821)
King Henry III disagrees [0], but arguably that's both.

[0] https://en.wikipedia.org/wiki/Assize_of_Bread_and_Ale

Re:Law for Computer Scientists (2020) by agumonkey on Sunday January 28, 2024 @05:12PM (#39170339)
Hehe, I felt this slightly a few times in the recent years.

Economy is also a large scale self adjusting optimizer, logistical neural network ?

Re:Law for Computer Scientists (2020) by Reflecticon on Sunday January 28, 2024 @07:41PM (#39171528)
Lawyer here and even though I know just a little about programming, I also always think that laws, contracts etc are like code running on society. Only when people fight over it and go to court, you run it on a "computer" but we only can test it beforehand running on human imagination. I'm very excited about chatgpt and it's capabilities to finally have a test environment for clauses where we can ask of ambiguity of terms, cases we haven't considered in the law/agreement and contradictions within the document. If we could have an ide for lawyers, that would be great too.
Re:Law for Computer Scientists (2020) by bitwize on Sunday January 28, 2024 @07:53PM (#39171615)
Laws are less like program code than they are like declarative constraints. And society has to solve for how to operate within those constraints. That's what makes making good law so difficult; it's like prompt-engineering a model and only getting results back, for good or ill, years or decades later. And when you do make a mistake (which is often) it can take that long to fix it!
Re:Law for Computer Scientists (2020) by Zambyte on Sunday January 28, 2024 @09:29PM (#39172078)
Program code can be declarative constraints. As said elsewhere in this thread, Prolog is an example of such a system.
Re:Law for Computer Scientists (2020) by hyperthesis on Sunday January 28, 2024 @08:21PM (#39171760)
The law can turn out to be different from what everyone thought. A higher court can reverse it. Even in the highest court, stating the definitive law, justices often dissent.

So the idea that the law actually exists, and we just have to find it, doesn't seem correct. Given that judges were usually barristers before, it's more like: one barrister confidently argues that the law has always been X, the opposing barrister confidently argies it's always been Y, the former-barrister confidently argues that it has always been Z - and that one always wins.

Re:Law for Computer Scientists (2020) by eru on Monday January 29, 2024 @02:46AM (#39173793)
Btw, not all jurisdictions even care about precedence. So courts can't reverse anything.

(Though in practice precedence seems to have about equal value in common law and civil law jurisdictions. In common law, the court just has to nitpick enough to find a difference between the current case and the precedent, to argue why the latter should not apply. In civil law, courts routinely defer to precedence as a matter of convention and habit.)

Re:Law for Computer Scientists (2020) by corethree on Sunday January 28, 2024 @10:30PM (#39172400)
Yeah except a jury votes on it and it's all up in the air whether all the predicates are actually enforced. Even the "predicates" themselves are open to interpretation and enforcement of a verdict is also sketchy. It's not even close to CS or an OS, more akin to the output of an LLM.

Which makes sense if you think about it.

Re:Law for Computer Scientists (2020) by eru on Monday January 29, 2024 @02:44AM (#39173777)
> I now have the very nerdy perspective that law is the operating system our socially run on. Laws are small snippets of code similar to a predicate in Prolog. We apply them once the conditions are fulfilled.

That's a very government centric point of view.

Compare https://slatestarcodex.com/2017/03/16/book-review-seeing-lik...

See also the exokernel papers like https://pdos.csail.mit.edu/6.828/2008/readings/engler95exoke... or via https://en.wikipedia.org/wiki/Exokernel

Traditional operating systems try to help programs securely share resources, but also want to provide abstractions away from the raw hardware. An exokernel approach limits itself to just the secure multiplexing, and delegates abstractions to another mechanism much better suited for the task: user space libraries.

To close out the tangent: I wonder how you would bring this metaphor back to laws? I guess it would have something to do with Subsidiarity? https://en.wikipedia.org/wiki/Subsidiarity

Re:Law for Computer Scientists (2020) by gcanko on Monday January 29, 2024 @11:02AM (#39177840)
Law is basically man made math. You have laws which can be seen as axioms and lawyers essentially use the existing axioms to prove to a judge a certain conclusion