Re: Content (Legos and Potato Heads)

[Bob <mantic@mbo.net>]

> >   I do not believe that it is necessary for end users to scribble all
> >over the sim software's content in order to customize it and shape new
> >content. Because talents vary this could even be prohibitive to the
> >player who cannot scribble well. I am of the opinion that engineered
> >customization tools should provide all users with building blocks in the
> >fashion of Role Playing Games, Mr Potato Head and Legos. Such elements
> >have an exponential effect on variety, but are more manageable.
> 
> Legos are a particularly interesting metaphor because Legos actually
> represent a whole continuum in themselves, ranging from parts that are not
> representative of anything and can be used in a great many ways (e.g. small
> yellow blocks), to parts that are high fidelity representation of things with
> little flexibility in how they can be used (e.g. molded green tree branches).
> Users can choose for themselves which types of parts to use, and whether to
> build the highly representational models on the box covers or mix the parts
> for creative innovation. One _can_ scribble with Legos, and that's not a bad
> thing.
> 
> It's interesting to note that interactive storytelling concepts usually start
> out with the notion that all their "pieces" are similar, but when put into
> practice they usually end up with a mix of more mutable general-purpose
> pieces and more specialized pieces.


	When I was a kid there was much less variety among lego pieces.
Primarily colour variety and a few different sizes were provided in the
different kits, and very few if any high fidelity specialized parts.
Over 30+ years the basic system has been steadily expanded and the
company has made a fortune at every step. Scribbling is possible to a
degree but always limited within the lego system; these limitations seem
appropriately analogous to the limitations of computer resources.
Unfortunately, the polygon triangle is a little too basic a block.


> The issue of scribbling is very different depending on whether or not the
> world is shared with other users. Many text MUDs allow users to scribble to
> their hearts' content in their own proprietary spaces, but no changes at all
> in public spaces or other users' proprietary spaces. Graphic MUDs have stayed
> closer to the Mr. Potato Head approach.


	I have seen at least one experimental effort at a graphical MUD engine
that was attempting to allow user modification of space. The engine used
a cubic spatial division to define the world and players would be able
to modify the space within controlled cubes in various ways. For
example, digging a maze of tunnels out of a mountain or building castle
walls. The prospect in itself didn't seem like a lot of fun (my
opinion), although it would have been a nice supplement to a world of
many systems.


> For simple single-player interactive stories, persistent effects on the world
> are not an issue except insofar as they affect the current story.
> "Scribbling" in this case translates not into "spoiling the polish of the
> world" but into "spoiling the polish of the story instance." In the ideal
> case, reasonable well-meaning choices should never result in a bad story, but
> no system should be expected to maintain story quality in the face of
> deliberate let's-screw-with-the-system behavior such as the hero killing
> people for no reason.
> 
> >   Wedding these kinds of content customizations to simulated environments
> >offers other possibilities. Perhaps a player might create and play a
> >character awhile, and then decide to leave that character in the world.
> >By using a rules based neural net learning system in addition to other
> >tables and systems such user-created denizens would be very dynamic,
> >shaping the experience of other players. And this wholly on the basis of
> >simulation.
> 
> Wow, this sounds very ambitious. Aren't there way too many potential input
> bits? I could see things like if attacked, run away or fight back, but
> resolving competing drives (e.g. I'm hungry, I'm also being pursued, so
> should I keep running or stop to eat?) isn't a strong point of neural
> networks. To say nothing of all the important interactions that are primarily
> conversational.


	By defining all such variables in tables a neural net would be capable
of distinguishing a great variety of situations, and make decisions
based on the degree of hunger and degree of danger of the moment. I
admit that it would be a complex system to handle with NN learned
behaviour systems alone.

 
> On the other hand, since the neural net is self-learning, nothing would be
> lost from the user experience in trying it out. As long as the ones that
> don't work well are removed from the world promptly. (Now there's an idea,
> define the neural net with a genetic algorithm, breeding the most successful
> previously created neural net bots to create the new learning net schema for
> each newly created player character. The fitness function could simply be
> autonomous survival time [or perhaps EPs gained during autonomous survival
> time could also be figured in], selecting in favor of those that can learn to
> protect themselves from monsters and player-characters and to play the game
> at a minimal level such as obtaining food when hungry.)
> 
> - Walt


	Yeah, that idea intrigues me to no end. However, I would be happy with
an imperfect, limited neural net propped up by more static systems.
Genetic approaches often breed themselves into (boring) evolutionary
dead ends. They are also counter to player participation. Perceived
flaws in reasoning inherited from a player aren't entirely a bad thing
-- the player's creative participation is no less rewarded if the
resultant creation is less than perfect. 

								--Bob