thanks for this post, I always feel a bit overwhelmed when it comes to licensing and other legal aspects of software development
Law for Computer Scientists (2020)
31–40 of 57 posts
Re: Law for Computer Scientists (2020)
#32Earlier quoted context omitted.
Been thinking about it, but our industry is a service industry and we prefer to get paid to offer our expertise. If I go out on my own and find myself lacking clients, I might.
I hear you, but at the same time, you're addressing a community where we do a lot of free work. The world wouldn't run without the open source unpaid software that we build. I can't critique you for this choice, as it is quite reasonable. But I can say that such publication can help others (potentially a reward in itself) and can help drive customers to you, who recognize that more nuance is necessary than what can e…
This book may already exist:
https://www.amazon.com/Legal-Singularity-Artificial-Intellig...
Re: Law for Computer Scientists (2020)
#33When 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.
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…
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)
#34When 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.
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)
#35Earlier quoted context omitted.
It's really hard to compare them. Just like it's hard to compare one country to another. So many factors.
The common way people think about common law versus civil law is this: -common law depends more on courts to make and refine legal decisions -civil law relies more on regulators. In civil law countries it's more common for the statutes (governing text) to be longer and go into great detail. In common law countries you see some extremely short laws - like the Sherman act in US Antitrust law is like 2 sentences long. T…
I do take your point, but the Sherman anti trust act is roughly a page to two pages long[0]
Only wanted to provide an accurate take, your overall point I agree with though
[0]: https://www.govinfo.gov/content/pkg/COMPS-3055/pdf/COMPS-305...
Re: Law for Computer Scientists (2020)
#36Earlier quoted context omitted.
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…
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!
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)
#37When 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.
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)
#38Earlier quoted context omitted.
The common way people think about common law versus civil law is this: -common law depends more on courts to make and refine legal decisions -civil law relies more on regulators. In civil law countries it's more common for the statutes (governing text) to be longer and go into great detail. In common law countries you see some extremely short laws - like the Sherman act in US Antitrust law is like 2 sentences long. T…
I do take your point, but the Sherman anti trust act is roughly a page to two pages long[0] Only wanted to provide an accurate take, your overall point I agree with though [0]: https://www.govinfo.gov/content/pkg/COMPS-3055/pdf/COMPS-305...
Thanks for the fact check. I'm writing on my phone so I can't do this justice, but the operative wording of the Sherman Act (or its followup the Clayton Act) is something like: "unfair restraint on trade is unlawful," or something like that. That's the one to two sentences I'm referring to. My point is that it's extremely limited phrasing. In the 100 years it has been in place there have been thousands or court decisions explaining what it means - and refining and defining what that sentence means in the US legal system.
If the act itself is only 2 pages that is a marvel, though. Usually they spend at least ten explaining why they are passing a law and who they are.
Re: Law for Computer Scientists (2020)
#39Earlier quoted context omitted.
The common way people think about common law versus civil law is this: -common law depends more on courts to make and refine legal decisions -civil law relies more on regulators. In civil law countries it's more common for the statutes (governing text) to be longer and go into great detail. In common law countries you see some extremely short laws - like the Sherman act in US Antitrust law is like 2 sentences long. T…
I wonder if this explains the propensity for lawsuits in the US. It's basically our regulation and enforcement mechanism.
Exactly. Yes. That's the idea.
The US legal system relies more on "ex post" legal enforcement - meaning, if you break the law then you get busted and you personally pay the victim. Europe is a more "ex ante" system - they rely on regulators to strictly define what the law should look like exactly and actually requires industry to do very specific things to comply with it. If someone gets hurt the system compensates them from a fund. The person who hurt them doesn't necessarily pay.
That's the theoretical underpinning and difference in our systems. But like I said the systems have a bit of both these days. Eu is flirting with more class actions, and US has more regulatory scrutiny in certain sectors, like California privacy laws for example being very detailed.
Re: Law for Computer Scientists (2020)
#40Earlier quoted context omitted.
Been thinking about it, but our industry is a service industry and we prefer to get paid to offer our expertise. If I go out on my own and find myself lacking clients, I might.
I hear you, but at the same time, you're addressing a community where we do a lot of free work. The world wouldn't run without the open source unpaid software that we build. I can't critique you for this choice, as it is quite reasonable. But I can say that such publication can help others (potentially a reward in itself) and can help drive customers to you, who recognize that more nuance is necessary than what can e…
Fair enough. Thanks. But... When I offer to do this here no one takes me up on it! So maybe the world doesn't need my blog.