Re: card/board games as income source

[WFreitag@aol.com]

> WFreitag@aol.com wrote:
 
>> So the envelope keeps growing, and the rowers keep on rowing, and there's 
>> no earthly way of knowing just how fast that we are going, or which way 
the 
>> river's flowing... (Sorry, started channeling Wonka there.) 

> Indeed, measuring progress towards the goal of Interactive Drama is 
> difficult.  We see gestures towards it, but few compelling results, and 
> nothing to indicate that the problem is 'solved'.

Too true. I only typed the (slightly garbled) lines from the old Willie Wonka 
movie because it was very late at night and the cadence of the phrase "so the 
envelope keeps growing" reminded me of it. (If you've seen the film, you 
might recall that Wilder starts out the poem in a kind of singsong.) Only when I 
went to delete it that I realized it actually said what I wanted to say.

>> So, every time 
>> discussion starts up again on the list, someone feels obliged to explain 
to 
>> us why we're wasting our time with interactive story system technology 
(that
>> is, the very subject this list is supposed to be about) 

> Well, given the slow progress in the field, someone is inevitably going 
> to suggest, "Less of X!  More of Y!"  In the interest of getting things 
> done, I'm usually in the camp of "Less attempts at dynamic content 
> 
generation!  More manual labor!"

Fundamentally, I agree with you. I think I recall arguing the same side as 
you (in favor of "data") in a "data versus process" debate with Chris Crawford. 
I commended Benja's experiments with brute-force branching text, and I've 
mentioned my contemplated card game -- certainly printed cards aren't going to be 
big on dynamically generated content.

But I don't think there's any danger of getting away from manual labor and 
brute force techniques, except perhaps by the most rarefied AI story systems 
concepts (which I haven't kept up on recently). Don't tell Laura that there isn't 
a lot of manual labor in building an Erasmatron storyworld. In fact, there's 
even a lot of manual labor in the Erasmatron software itself -- the built-in 
core character behavior simulations for things like gossip and lie-discovery 
are based on many eplicit rules, not abstract math. For all the sophisticated 
work in the Facade (interactivestory.net) engine, most of the work is clearly in 
writing the beats.

If, however, by "more manual labor" you mean "more actual development, less 
theorizing and/or waiting for someone else to develop better tools for you," I 
couldn't agree more. You did manage to convince me of that much, in our heated 
exchanges a few years ago, even though I still maintain that theorizing isn't 
without value either.

But Thom appears to be saying, basically, "less of everything you're all 
doing and talking about, and more..." more what, exactly? Staging old-hat "edgy" 
multi-communication-channel drama in yet more imitations of the movie "The 
Game?" Making silk-screen prints of people on bicycles? Dentistry? (That's all 
face-to-face between humans, right?) Maybe we can get an answer to that soon.

[edit: or maybe not; just read his response to my last post]

> I just think humans are better and faster at composing dramatic rules 
> systems than computers are.  I call that sort of thing "writing skill."  

Well, of course. Though I think of the problem just a bit differently. 
Composing dramatic rules systems is something I'd think the author-programmer would 
do. The hard part, for computers (and often, human writers too) is 
instantiating. Even when you've got a formula (such as Dramatica, or a brand name romance 
style guide, or My Life With Master) telling you that in this chapter the 
hero should suffer a setback that will bring him emotionally closer to the mentor 
character, not everyone on every try can write something good to fill it in 
-- and a computer's hopeless at the task. What, exactly, should the setback be? 
A lot of common sense knowledge about the real world and how real people 
react is required to be sure that the instantiation is even plausible, let alone a 
good fit to the rest of the story. And as soon as you say "common sense 
knowledge" in a computer system, you know you're in trouble. Erasmatron has made 
some great progress on the character-reaction side of the problem, but it's not 
solved. Ahab nails a doubloon to the mast: an event and image of great power 
in its context in the story. But if you have a "nail doubloon to mast" verb in 
a Storyworld then even if the right moment for it happens to come up, Ahab 
won't be standing by the mast or he won't have a hammer with him. The 'tron isn't 
designed to "push" toward specific events happening. And if you generalize it 
to "[character] nails [something] to [something else]" it's even worse: you 
now need a whole slew of additional rules to prevent the whole cast from going 
around nailing teacups to cannonballs at odd moments.

(There are people, as far as I know, still working on the AI "common sense 
knowledge base" project, which has been ongoing for many years. Perhaps that 
will help. Meanwhile, I think event libraries like Dramaton's might still have a 
place.)

> I don't have any problem with people trying to amplify their writing 
> skill using computers.  

It's called a word processor :-). Okay, I suppose Dramatica might also fit 
that description.

> I just think trying to get the computer to do it 
> entirely is damn hard.  Achievable within the next 20 years, but I think 
> we can get pretty close to strong AI in that timeframe also.  We're 
> going to need a lot of computing resources, and a way to manage the 
> overwhelming complexity of those resources.  A more "biological" 
> approach to computing.

If anything, I think you're optimistic on both achievements. However, work 
with genetic algorithms (that's pretty darn biological) had yielded some 
surprising results (enough to call into serious question the "computers cannot be 
creative/inventive" canards).

> I really hate the word 'rich' as applied to either technology or 
> content.  It's a meaningless gestural word, much like we were taught in 
> grade school that 'very' is a meaningless word.  (Usually.)  Microsoft 
> is the worst offender, using "rich applications, APIs, and content" in 
> near continuous drone, like chocolate that's supposed to be good for 
> you.  So, anybody who talks about things being "rich and engaging" ends 
> up sounding like a Microsoft Tele-Evangelist to me.

Sorry about that. It was a case of my using personal jargon without defining 
it in the post. "Richness" is part of my guiding schema of five 
usually-desirable qualities of recreational virtual worlds (along with "objectivity," 
"accessibility," "attractiveness," and "responsiveness" -- quote marks are to warn 
that those are idiosyncratically defined personal jargon words too. By "rich" I 
simply mean "able to hold a visitor's interest for an extended period of 
time." As with all of the five "qualities," richness is a goal, not a prescription 
for how it must be met. And it's entirely dependent on the visitor, that is, 
on what holds any particular visitor's interest. Richness can come from game 
play, narrative (interactive or otherwise), fun activities (think Disney World 
or Vegas), interesting fellow participants to chat with, or even just having a 
lot of pretty images to look at (though the latter is at present a very 
inefficient way to achieve it, unless you're a porn site).

"Richness" _achieved via_ visitors' involvement in interactive narrative is 
my specific virtual-world-design goal.

An no, it's not very specific, or very profound. I didn't mean it to sound 
like it was.

>> requires the ability to interact with its visitors so as to 
>> generate the stories THEY personally want to tell.

> The job of telling the story can be assigned to various people or 
> things.  At one extreme, you assign it all to the player.  They tell 
> "boring fish stories" to each other, which are deeply engrossing to 
> them, but to nobody else.  At another extreme, you assign it all to the 
> author.  Really that's a movie, no interactivity at all.  (Well, 
> excepting that some movies are "mentally interactive," in that they can 
> be left open-ended, and the audience completes the film with their own 
> meanings.)  A third extreme is to assign it all to the computer.  At 
> present, this produces either trivial results or random noise.  Neither 
> are terribly interesting to a human.
> 
> So we have this triangle: author, audience, computer.  Many strategies 
> can be tried within this engineering triangle.  Some of our religious 
> arguments are no more than people advocating different points in the 
> triangle.  Anybody into Barycentric coordinates?  :-)

An excellent assessment. Let me just clarify one thing for the benefit of 
other readers: I gather that by "telling the story" you also mean "generating the 
story," not just dispensing a story that's already written in advance. I 
think that's what you mean, because we all know that a computer can play a DVD, or 
railroad a player through an adventure game with a good story built in -- but 
that doesn't count as "telling the story" in the sense you're talking about.

You have to make that distinction carefully, because it gets more complicated 
from there. Watching a movie isn't quite "all-author" as far as _telling_ is 
concerned because at the time the audience watches, the author's already 
packed it in. The movie film itself is telling (but not generating) the story, and 
as you said there's no interactivity at all (except in the "mentally 
interactive" sense you spoke of, which I hope we can just accept as inherent in all 
communication and therefore doesn't need mentioning most of the time). But the 
(already done) _generating_ is still, as you said, all-author. The plastic film 
and dye doesn't, itself, create anything.

However, what do we make of cases where the author is present at the 
performance, and there is two-way communication between author and audience, but the 
author is doing all the heavy lifting creatively? Two examples: first, a 
tabletop role playing game in which the players call their characters' actions while 
rarely doing anything unexpected -- "There's a door? We open it. There's a 
monster attacking us? We try to kill it. There's a shadowy guy in a bar telling 
us about a great treasure in a cave? We go look for it. The path splits left 
and right? We go, um, right." -- while the GM spins their choices into a 
well-plotted adventure. Second: an improv performance, where the audience is asked 
for bits and pieces, but it's clear that between the audience ("Pick a place!" 
"Pick a TV show!") and the performers, it's the performers' creativity and 
skill that's being exhibited.

Most would say that this moves us a bit away from the "all-author" vertex 
along the author-audience vector. And I tend to agree. But we should at least 
consider the possibility that these are actually still all-author; that some 
minimal threshold for meaningful decision-making by the audience hasn't yet been 
reached.

Now suppose the author makes a film with four different endings. At a certain 
point the film stops and the audience, using buttons built into their theater 
seats, votes on which ending will be shown. Different people with different 
philosophies will plot that differently in the triangle. Some of the arguments:

- The audience is making a choice (but only one) and is otherwise just like 
an ordinary movie, so it's near the all-author vertex but displaced a short 
distance along the author-audience axis.

- The audience is interacting, but they can't be interacting with the author, 
because the author's not there. They're only interacting with the mechanism 
that counts the votes and controls the projector. Therefore a bit of the story 
generation has been ceded to the machine, which is enough to move the spot a 
bit toward the computer vertex (with the vote-counting device representing the 
computer).

- But the vote-counting machine is dumb, and cannot possibly be creating 
anything. So the previous argument must be wrong. (Thus begin some of those 
religious arguments you mentioned. This particular one leads eventually to an 
argument over whether or not any computer, obeying its programming, can ever be said 
to create anything, which is more an argument about the meaning of the word 
"create" than about computers.)

Anyway, it is a useful framework for such discussion.

- Walt