John, (01)
I'm still on my track. (02)
> the real world has to deal with a continuous infinity of possible
> variations and combinations. (03)
In case of a legal problems the "continuous infinity" is just a discrete
number (it is probably a "big number" for ordinary people but maybe not too
big for computers). To say if computers can or cannot work with this number
of combuinations we have to estimate it. I am sure that the number of
possible variations and combinations in the legal system (inspite of real
continuous infinity) is not too big for the power of networked computers.
Any legal story and a dispute is coming into the legal system by words sayed
by people involved into the legal dispute with the rate as few bytes per
second. It is not a big deal for computers to work with it. (04)
> For a discrete system of rules that apply to a discrete set of elements,
> consider the laws of chess. Those laws are so precise that they can be
> implemented on a computer and be carried out precisely.
>
> But when two humans are playing across a table at a tournament,
> disputes can arise whether one player deliberately or accidentally
> touched a piece, whether one player was doing something to distract
> another player, etc. (05)
"Chess computer" has changed the "chess world", so why not a "law computer"
(a "legal ontology") could change the "legal world"?
Chess computers win chess grandmasters . Who cares if the legal games will
be won by the computer program? (06)
> Huge numbers of disputes can arise about whether a ball or a human
> body part was inside or outside a boundary line, whether one player
> deliberately or accidentally hits, bumps, interferes with another.
> These situations are very far from clear, and the rules of the
> game, by themselves, cannot account for all possible interactions. (07)
How does a real "human" lawyer resolves such unclear issues? The program
could do a similar job. It could be an interactive dispute around similar
cases. An unprecedented cases must be discussed by humans, of course, and
then be added to the system. (08)
> Then take the laws for just a small subset of a community: the
> traffic laws. The number of conditions is much smaller than the
> full set of laws, but just consider all the possible variations
> in the world:
>
> 1. Static configurations of roads, highways, intersections,
> stop signs, traffic lights, obstructions such as trees,
> bushes, parked cars, curbs, railings, buildings, etc.
> (09)
Not a big deal. Even millions or billions variations can be quickly
estimated for a real case. A lot of roads is already in the Google Map. Some
more legal semantics added there and it could be a good traffic lawyer :) (010)
> 2. Dynamic interactions with other cars, trucks, pedestrians,
> animals, sunlight and other distracting light sources, etc. (011)
I am sure that for millions of stories the there will be a lot of similar
cases. The number of considerable parameters could be pretty small. (012)
> 3. The condition of the vehicles, their brakes, steering,
> tires, etc, and the liability for unsafe conditions
> caused by the manufacturer, service station, owners... (013)
All conditions are checked by police (or whoever it must do by law) and
entered in the system as limited number of words, photoes and schemas.
Ccomputers can work with lots of such objects much better than humans. (014)
> 3. The condition of the vehicles, their brakes, steering,
> tires, etc, and the liability for unsafe conditions
> caused by the manufacturer, service station, owners...
>
> 4. Weather conditions, road conditions, and the responsibility
> of individuals or governments for causing or fixing them... (015)
Access to the other information sources could help (auto-sercvices, wheather
info centers, so on). These points just support a benefit of the
"computer-lawyer" as a "human-lawyer" is not able to collect and estimate
all related information for the limited period of time to proceed the case.
The networked computers have a benefit to communicate immediately with
different data souces (which could be also ontologized) (016)
> Even trying to think of all the possibilities is mind boggling.
> It's impossible to anticipate all of them, and even more difficult
> to decide what to do about them and record them in the legal code. (017)
What are the mistical human skills which allow human lawyers to do this job? (018)
Sometime in the past people thought that machines cannot win chess games
against humans.
Now it is a reality.
Sometime in the future machines will protect us in a most legal situations. (019)
We have now a unique chance to invest our efforts in this future by
creating, say, a trafic law ontology as a case (or implement another simple
legal code) because (I suppose this based on a common sense) the laws of
Justice must be open for public (as the "open data") and formally
understandable (by resoners). (020)
Yuri (021)
----- Original Message -----
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
To: <ontology-summit@xxxxxxxxxxxxxxxx>
Sent: Tuesday, February 01, 2011 5:23 PM
Subject: Re: [ontology-summit] [Making the Case] Elevator Pitch (022)
> On 2/1/2011 4:43 PM, Yuriy Milov wrote:
>> The only place
>> where things are supposed to be managed by discrete categories absolutely
>> is
>> a legal system - otherwise people cannot folow laws if the laws are not
>> written by letters and if the legal comments are not understandable by
>> common sense (common logics?).
>
> The categories of the law may be clear, but their application to
> the real world has to deal with a continuous infinity of possible
> variations and combinations.
>
> For a discrete system of rules that apply to a discrete set of elements,
> consider the laws of chess. Those laws are so precise that they can be
> implemented on a computer and be carried out precisely.
>
> But when two humans are playing across a table at a tournament,
> disputes can arise whether one player deliberately or accidentally
> touched a piece, whether one player was doing something to distract
> another player, etc.
>
> Then consider games with continuous motion of people and physical
> objects, such as basketball, baseball, football, soccer, etc.
>
> Huge numbers of disputes can arise about whether a ball or a human
> body part was inside or outside a boundary line, whether one player
> deliberately or accidentally hits, bumps, interferes with another.
> These situations are very far from clear, and the rules of the
> game, by themselves, cannot account for all possible interactions.
>
> Then take the laws for just a small subset of a community: the
> traffic laws. The number of conditions is much smaller than the
> full set of laws, but just consider all the possible variations
> in the world:
>
> 1. Static configurations of roads, highways, intersections,
> stop signs, traffic lights, obstructions such as trees,
> bushes, parked cars, curbs, railings, buildings, etc.
>
> 2. Dynamic interactions with other cars, trucks, pedestrians,
> animals, sunlight and other distracting light sources, etc.
>
> 3. The condition of the vehicles, their brakes, steering,
> tires, etc, and the liability for unsafe conditions
> caused by the manufacturer, service station, owners...
>
> 4. Weather conditions, road conditions, and the responsibility
> of individuals or governments for causing or fixing them...
>
> Even trying to think of all the possibilities is mind boggling.
> It's impossible to anticipate all of them, and even more difficult
> to decide what to do about them and record them in the legal code.
>
> John
>
> _________________________________________________________________
> Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
> Subscribe/Config:
> http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
> Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
> Community Files: http://ontolog.cim3.net/file/work/OntologySummit2011/
> Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011
> Community Portal: http://ontolog.cim3.net/wiki/
> (023)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2011/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011
Community Portal: http://ontolog.cim3.net/wiki/ (024)
|