3 ms·

Meh, not so fast. Here's the Stanford CS professor who runs the Stanford Center for Computers and Law[1]: "One technical problem with Computational Law, famili

Meh, not so fast. Here's the Stanford CS professor who runs the Stanford Center for Computers and Law[1]: "One technical problem with Computational Law, familiar to many individual with legal training, is due to the open texture of laws. Consider a municipal regulation stating "No vehicles in the park". On first blush this is fine, but it is really quite problematic. Just what constitutes a vehicle? Is a bicycle a vehicle? What about a skateboard? How about roller skates? What about a baby stroller? A horse? A repair vehicle? For that matter, what is the park? At what altitude does it end? If a helicopter hovers at 10 feet, is that a violation? What if it flies over at 100 feet? The resolution of this problem is to limit the application of Computational Law to those cases where such issues can be externalized or marginalized. We allow human users to make judgments about such open texture concepts in entering data or we avoid regulatory applications where such concepts abound. A different sort of challenge to Computational Law stems from the fact that not all legal reasoning is deductive. Edwina Rissland [Rissland et al.] notes that, "Law is not a matter of simply applying rules to facts via modus ponens"; and, when regarding the broad application of AI techniques to law, this is certainly true. The rules that apply to a real-world situation, as well as even the facts themselves, may be open to interpretation, and many legal decisions are made through case-based reasoning, bypassing explicit reasoning about laws and statutes. The general problem of open texture when interpreting rules, along with the parallel problem of running out of rules to apply when resolving terms, presents significant obstacles to implementable automated rule-based reasoning." [1] http://logic.stanford.edu/complaw/complaw.html http://logic.stanford.edu/complaw/complaw.html

anonbanker 11y ago
What you state is essentially banking on the limitations of AI programming, on the hopes that what keeps us "human", and therefore, important as a cog in the machine of law/commerce, will never be discovered or replaced. That might take longer than 15 years, but it's definitely within our lifetimes. and it'll be a beauty to behold.
- hobbyjogger 11y ago
  I didn't state anything. That's a direct quote by the guy who runs what is, by leaps and bounds, the most advanced project applying AI to legal analysis. They're doing some incredible (and incredibly interesting) things, but even they don't see AI taking over courtrooms any time soon. Think about the two problems Professor Genesereth noted above (even in the context of the most laughably simple example imaginable). These problems aren't just of a different magnitude compared to the problems we are working on with AI right now--they're of a different character entirely. Do you know of any AI programs that can generate analogies to every possible thing/situation/event and choose the most "well-suited" among those alternatives based on a holistic, "all important things considered" balancing approach? Do you know of any AI programs that are able to rank human values/morals in a heirarchy, in a way that would be acceptable all (or even most) of us? Until the answer to those questions is "yes", you're not going to see AI sitting behind the bench in an American courtroom. A brief aside: the common law is used as a lazy evaluation strategy[0], whereby we can defer filling in the full content of any idea/word/phrase/whatever until it actually becomes a problem. So the legislature can pass a law "no pets allowed" instead of spelling out a full and comprehensive list of every imaginable item to be banned: "No Mesocricetus auratus [golden hamster], Mesocricetus brandti [Turkish hamster], Mesocricetus newtoni: [Romanian hamster]...and so on for the next 10,000 pages." AI could easily handle some of this filling in of content but not so much items like "reasonable search" or "malicious intent" or even, apparently, "vehicle." Those types of decisions often fall, not to logical deduction, but to induction from incomplete, conflicting and messy human theories about goodness and justice and truth. Just as most wouldn't outsource their decisions about who to love or what god to believe in, they're not going to let AIs make these types of decisions (assuming, of course, that AIs are even eventually capable of doing so--which remains an open question). [0] https://en.wikipedia.org/wiki/Lazy_evaluation https://en.wikipedia.org/wiki/Lazy_evaluation
  - dragonwriter 11y ago
    Quite simply, there is no consensus on the right rules to use to apply the existing body of case, statute, Constitutional, etc., law -- which is why judicial appointments are a contentious, political thing. Without such a consensus, even if all the AI-related problems are solved we couldn't design AI judges -- because the requirements would not be well defined.

anonbanker 11y ago

What you state is essentially banking on the limitations of AI programming, on the hopes that what keeps us "human", and therefore, important as a cog in the machine of law/commerce, will never be discovered or replaced. That might take longer than 15 years, but it's definitely within our lifetimes. and it'll be a beauty to behold.

hobbyjogger 11y ago
I didn't state anything. That's a direct quote by the guy who runs what is, by leaps and bounds, the most advanced project applying AI to legal analysis. They're doing some incredible (and incredibly interesting) things, but even they don't see AI taking over courtrooms any time soon. Think about the two problems Professor Genesereth noted above (even in the context of the most laughably simple example imaginable). These problems aren't just of a different magnitude compared to the problems we are working on with AI right now--they're of a different character entirely. Do you know of any AI programs that can generate analogies to every possible thing/situation/event and choose the most "well-suited" among those alternatives based on a holistic, "all important things considered" balancing approach? Do you know of any AI programs that are able to rank human values/morals in a heirarchy, in a way that would be acceptable all (or even most) of us? Until the answer to those questions is "yes", you're not going to see AI sitting behind the bench in an American courtroom. A brief aside: the common law is used as a lazy evaluation strategy[0], whereby we can defer filling in the full content of any idea/word/phrase/whatever until it actually becomes a problem. So the legislature can pass a law "no pets allowed" instead of spelling out a full and comprehensive list of every imaginable item to be banned: "No Mesocricetus auratus [golden hamster], Mesocricetus brandti [Turkish hamster], Mesocricetus newtoni: [Romanian hamster]...and so on for the next 10,000 pages." AI could easily handle some of this filling in of content but not so much items like "reasonable search" or "malicious intent" or even, apparently, "vehicle." Those types of decisions often fall, not to logical deduction, but to induction from incomplete, conflicting and messy human theories about goodness and justice and truth. Just as most wouldn't outsource their decisions about who to love or what god to believe in, they're not going to let AIs make these types of decisions (assuming, of course, that AIs are even eventually capable of doing so--which remains an open question). [0] https://en.wikipedia.org/wiki/Lazy_evaluation https://en.wikipedia.org/wiki/Lazy_evaluation
- dragonwriter 11y ago
  Quite simply, there is no consensus on the right rules to use to apply the existing body of case, statute, Constitutional, etc., law -- which is why judicial appointments are a contentious, political thing. Without such a consensus, even if all the AI-related problems are solved we couldn't design AI judges -- because the requirements would not be well defined.