Lists Home |
Date Index |
From: Didier PH Martin [mailto:email@example.com]
>I probably have the wrong model but, in my mind, a role playing games are
>involving humans playing roles, the machine is there only as a kind of
>scorecard. Most of the system's state modification is based on the behavior
>of the people. Is that what you have in mind? A kind of "matrix" where
>humans helps to resolve some problems? To resolve two proposition about the
>same resource from two different agents (i.e. a document classified by a
>search engine and containing an RDF description) how do I use an RPG to
>solve that problem? Would that matrix interested to solve my problem? Can
>you help me understand your ontology and why you reach such conclusion, I
>lost, I do not know why you came to that conclusion.
Len Replies: I discussed many topics in the last mail. I'll try to
make this short:
1. Some of the players are human. Some are avatar endowed agents.
Both can shape the environment. The avatar agents could represent
ontological viewpoints. Complex, but doable. In living RPGs which
have been used by policy makers since the 60s, a director set the
situations and changed the circumstances for the role players to
reveal different aspects of the conflict. After some number of
plays, most of the facets of the conflict would be illuminated.
2. If the matrix reveals a conflict, any resolution strategy will
have to first determine if the conflict is real or superstitious,
say, connotative or denotative differences.
3. An RPG is not an efficient way to solve the conflict, but then
what is? Again, we could get the humans together and ask them,
but as the debates here and in the TAG have demonstrated, some
terms will have a consensus definition, and others will come
down to 'I am on the TAG, I wrote this or that and know him or her,
and you should ignore all other comments but mine and those
that agree with me'. IOW, authority by proclamation. This
is not too different from the RPG. The list members are playing
a role-based game, assume rights and privileges based on role,
and have strategies for coming to a faux consensus or a real
consensus. The result of that will be a document in which
some terms will be widely embraced, and other sections which
will be largely ignored. It isn't wrong; it is the best the
process can create. Again, the jam bands.
The Semantic Web cannot resolve a set of equally probable
choices except to rig the game so that some choices are
more equal than others. We can know the orbits of the
terms and the attractors that they orbit, but predicting
which choice is correct in a given time slice requires
one to finesse, cheat, step outside the system and fix
a value (what the RFP director does).
In short, is there a way to make the two search engines
play each other? I know that sounds far-fetched, but
most innovative solutions are hard to distinguish from
fantasy except by having an implementation strategy.