GNS, AI, and psychology.
Alfryd:
Ron- sorry, I missed your post there. Fortunately, much of what I said above is, I hope, applicable to your own remarks.
Quote from: Ron Edwards on August 21, 2010, 10:23:16 AM
Timo, to follow up on your AI idea, I tend to think that all three "sectors" you describe are present in all role-playing, and often as priorities of technique, rather than expressing any of the Creative Agendas.
If I were to extend the logic of your post into something I'd be more likely to agree with, it'd approach not only the substance of cognitive effort, but the relationship of the actor with everyone else involved. It would focus more on group dynamics of approval or excitement, and then seek repeated cycles, and finally, examine the content and activity within those cycles.
I submit, or speculate, that all three things you describe would be present at peaks of the cycles, or in very successful play, throughout the cycles, and that other things - rather specific ones - would pop out as CA-identification for that group.
*coughs* *Morgan*
Well, in the same sense that (virtually) all role-play involves some degree of capital-E Exploration (Sim,) and some form of adversity or problem-solving (Game,) and some amount, however limited, of 'storytelling' or protagonism (Nar,) then yes, all three processes would be frequently invoked during any role-play experience. But the basic idea of GNS theory, insofar as I understand it, is that individuals often have marked preferences for a particular mode or blend of modes, in terms of which they prefer to exercise primary control over first, or on which side they're most willing to accept compromises, or which they get the most enjoyment out of wrestling with, or which they consider to be means to other ends(?)
Quote
Does any AI research concern group dynamics and relationships among interacting AIs?
Almost certainly, but my knowledge of the subject is pretty limited- I took a university course on the subject along with some reading in my spare time, and given the breadth of the field we only touched briefly on inter-agent communication or language synthesis. I'll see if I can look something up, but your guess may be as good as mine. :)
Motipha:
*grin* I was about to say. I'd like to claim that I'm that savvy and insightful, but I cannot claim the OP. I really was just wondering how Aaron's comment applied.
I still stand by the fact that it's a great story though.
Ron Edwards:
Hi Morgan (sorry about the name mistake; I'm easily confused),
My take is that you are falling into some common misunderstandings of the body of ideas developed here.
1. Creative Agenda is a bigger topic than internal processes of a single participant.
Your claim here,
Quote
I mean, if there's some kind of truth to the comparison with the categories of machine learning I gave, then none of the three modes are inherently dependant on social dynamics at all. An AI agent working in complete isolation can plan courses of action to attain specific goals, build up a model of their environment by exploration, and infer new formal rules based on prior observations, without ever consulting another agent or even requiring a theory of mind.
... is flawed for the purposes of talking about Creative Agenda. Such individual or private functions are necessary to the activity, just as one's personal physique, psychology, training, and actions based on these are necessary to the activity of playing baseball ... but playing baseball is a social phenomenon, relying on communication, responsiveness, and above all the observably expressed priority (I do not mean speech, I mean actions while playing) of baseball and its reinforcement among the group members during play itself.
See What is Creative Agenda? It's not a great thread. The initiator of the thread, Fred/Vaxalon was in attack mode rather than inquiry mode, and one participant, Nathan (Paganini) was fundamentally confused despite posting with a tone of authority. However, I do manage to articulate there that a Creative Agenda occurs socially, and that whatever may be going on individually and internally is, for purposes of definition, relevant only to the social expression and reinforcement.
2. Brain anatomy and physiology
I'll leave your ideas on the individual prioritization of the modes based on brain physiology to the researchers in the topic. My questions to them would concern whether cognitive content is energetically limited, and whether gross architecture has been shown to play a determinant role when an individual trades-off among possible priorities.
3. Sociobiology and world-building
I'd like to start by tagging your sociobiological musings as speculative, which is not itself a bad thing, but you're missing an important point: that the kind of cognition called "mapping," or in present terms "world-building," is common among organisms and the human version is not especially distinctive in many of its parameters. We probably inherited most of our cognitive capabilities rather than evolving them in the history of our one species. This is relevant to the present discussion because our brains, or rather the human mental functions, carry out such mapping at a constant, basic, no-attention-switch level. Arguably we can do hardly anything else without it. I don't see any evidence suggesting that we ever stop.
Although your speculations about the priorities are not illogical, and I do agree strongly with the idea that the selective context of human behavior was and is composed of other humans, I think you're falling into the trap of adaptationism - specifically, that for variation to be present (i.e. observable at present), the individual variations must have been optimal at separate historical points. Selection is frequently mixed up with one of its possible outcomes, fixation (100% presence of a single variant), whereas in real creatures, fixation does not appear to be the most common outcome of selection. Furthermore, flexibility among certain options (or spectra of options) may itself be the outcome of selection, rather than any single point on it being "the fitness" winner.
But again, I think you are focusing way too much on world-building = Simulationist. I know that my essays point to the idea of Simulationism being "Exploration squared," but later discussions finally uncovered more content. See my comments in Ignoring the subjective? about it (see the embedded link for the prequel and Constructive Denial? for the useful follow-up). These threads finally managed to extricate the definition from fictional content and to focus on the correct zone of interest, the social and creative priority.
4. And the big problem, individual typology. You wrote,
Quote
Well, in the same sense that (virtually) all role-play involves some degree of capital-E Exploration (Sim,) and some form of adversity or problem-solving (Game,) and some amount, however limited, of 'storytelling' or protagonism (Nar,) then yes, all three processes would be frequently invoked during any role-play experience. But the basic idea of GNS theory, insofar as I understand it, is that individuals often have marked preferences for a particular mode or blend of modes, in terms of which they prefer to exercise primary control over first, or on which side they're most willing to accept compromises, or which they get the most enjoyment out of wrestling with, or which they consider to be means to other ends(?)
Your "basic idea, insofar as I understand it," needs to be completely revised. I cannot state more strongly that marked individual preference is not the core concept of the ideas (technically the Big Model). It's a historical observation, and of great practical importance (hence why I originally emphasized it) but not a defining feature.
And furthermore, you have it backwards. In role-playing, I am emphatically not saying that Sim, Gam, and Narr are all present (and hence it's a matter of percentages), but rather than Exploration must be present, socially, and in that context only one such priority is viable at a time. It seems to me that you might be thinking in terms of what some people called "the GNS Triangle," which was an intellectual fallacy.
I think it would be useful as well to consider the difference between Techniques and Agenda. You may be missing that fundamentally important paragraph in GNS and other matters of role-playing theory:
Quote
For a given instance of play, the three modes are exclusive in application. When someone tells me that their role-playing is "all three," what I see from them is this: features of (say) two of the goals appear in concert with, or in service to, the main one, but two or more fully-prioritized goals are not present at the same time. So in the course of Narrativist or Simulationist play, moments or aspects of competition that contribute to the main goal are not Gamism. In the course of Gamist or Simulationist play, moments of thematic commentary that contribute to the main goal are not Narrativism. In the course of Narrativist or Gamist play, moments of attention to plausibility that contribute to the main goal are not Simulationism. The primary and not to be compromised goal is what it is for a given instance of play. The actual time or activity of an "instance" is necessarily left ambiguous.
This issue was made easier to understand when I defined the layers of the Big Model, especially in terms of Techniques as components of play, whereas Creative Agenda is a unifying principle rather than a component (or made of them).
See also Frostfolk and GNS aggravation and Frostfolk, carrying on for a good example of someone else grasping the difference.
In conclusion, I am not claiming that AI has nothing to offer our understanding of role-playing or other human activities (although the payoff to date is mighty low given all the rhetoric), and I definitely do think that sociobiological perspective has much to offer (although its adoption in non-biological disciplines is usually crappy). But I think that the first step would be to understand what I mean by Creative Agenda and then to see whether and how AI theory may apply.
And it strikes me that I do not want to slap you down, which this post seems like it might be bordering on. In the interest of following up on the part I find most interesting, the whole deduction/induction issue, check out zplay - liberating Sim and embarrassing Exploration!. Although the overall thread is totally obsolete and rife with gibble-gabble, Jason Lee's comments show how induction, deduction, and abduction are integrated within Exploration (i.e. all role-playing) regardless of Creative Agenda.
Best, Ron
Alfryd:
Quote from: Ron Edwards on August 23, 2010, 10:55:37 AM
Hi Morgan (sorry about the name mistake; I'm easily confused)...
No worries. :)
Quote
... is flawed for the purposes of talking about Creative Agenda. Such individual or private functions are necessary to the activity, just as one's personal physique, psychology, training, and actions based on these are necessary to the activity of playing baseball ... but playing baseball is a social phenomenon, relying on communication, responsiveness, and above all the observably expressed priority (I do not mean speech, I mean actions while playing) of baseball and its reinforcement among the group members during play itself.
Again, I agree fully that role-play is a social phenomenon and, so, obviously, when discussing specific CA in the context of role-play social forces are very important. The point I'm making is that the same basic agendas need not be limited to activities that are defined as social.
I mean, just to give a simple example, SimCity is a highly successful video game which many people play with no clear win-conditions, no detectable story, and no external social expectations whatsoever. To a first approximation, it's a solo experience, but nonetheless oddly compelling. (People often make the assertion that SimCity and related games are about 'tackling challenges', but closer inspection belies that- the game's baseline mechanics get steadily easier as you go on, rather than throwing escalating challenges to match the player's skill or resources. It's mostly about deriving enjoyment from watching the interplay of internal cause-and-consequence.) Similarly, although Narrativist examples of video game design are essentially nonexistant (because we don't yet have the AI to recognise and accomodate questions of theme and protagonism,) there are no shortage of devotedly single-player video games which show an abundantly clear Gamist agenda, either in terms of their design, how players tackle them, or both. (It's also interesting to note how many players of the Sims or Crusader Kings seem to take active pleasure in spinning dramatic narratives out of their in-game experiences, and this, too, is an essentially solo activity.)
I just think that placing analysis of social interactions ahead of the private motives and instinctive abilities of the players is, in a sense, putting the cart before the horse. If players didn't have innate preferences that frequently outweighed or were at least comparable to social expectations, I don't see how GNS disagreements could arise at all. And if those modes or preferences are observable in situations which are about as non-social as human activities can get, is it really useful to restrict discussion about them to social settings?
Quote
I'd like to start by tagging your sociobiological musings as speculative, which is not itself a bad thing, but you're missing an important point: that the kind of cognition called "mapping," or in present terms "world-building," is common among organisms and the human version is not especially distinctive in many of its parameters. We probably inherited most of our cognitive capabilities rather than evolving them in the history of our one species. This is relevant to the present discussion because our brains, or rather the human mental functions, carry out such mapping at a constant, basic, no-attention-switch level. Arguably we can do hardly anything else without it. I don't see any evidence suggesting that we ever stop.
Oh, everything I'm saying here is just amateur speculation- feel free to "slap me down" if the situation calls for it.
I fully agree that environment mapping is present in many other organisms and no doubt represents the outcome of a long evolutionary process, but precisely the same can be said for methods of social networking and action planning. We decide on things to do to achieve long-range goals on a constant, basic, no-attention-switch level. We infer insights about large-scale structure across many data points on a constant, basic, no-attention-switch level (it's arguably fundamental to sensory processing.) But organisms were taking action, and sending eachother 'messages' of some form, long before they had any kind of abstract knowledge about their environment.
I also agree that "arguably we can do hardly anything else without it". In the same sense that you can't have an imaginary situation without Exploration, you can't have 'intelligent' cognition without environment mapping. You need to know how the world works before you can manipulate it to achieve desired ends, and any patterns you infer need to be checked for consistency with prior knowledge.
However, I feel that the specific psychological impulse associated with what could be called Simulationism is a kind of detached pleasure in relating and observing cause and consequence, relations between facts, in searching for 'objective' understanding for it's own sake, without particular regard for any immediate practical purpose. Broadly speaking, a blend of skepticism and curiosity: "Why does this happen"? or "what would happen IF?"
I will try to read up on the threads you cited and muse on the subject some more, and I do appreciate your commentary on the subject of variability and selection. Perhaps rather than 'historical points' it would be better to talk about 'relevant variation in adaptive pressures'? I'll have to think about it some more.
Quote
Quote
For a given instance of play, the three modes are exclusive in application. When someone tells me that their role-playing is "all three," what I see from them is this: features of (say) two of the goals appear in concert with, or in service to, the main one, but two or more fully-prioritized goals are not present at the same time. So in the course of Narrativist or Simulationist play, moments or aspects of competition that contribute to the main goal are not Gamism. In the course of Gamist or Simulationist play, moments of thematic commentary that contribute to the main goal are not Narrativism. In the course of Narrativist or Gamist play, moments of attention to plausibility that contribute to the main goal are not Simulationism. The primary and not to be compromised goal is what it is for a given instance of play. The actual time or activity of an "instance" is necessarily left ambiguous.
Again, I don't disagree with any of this. The point I'm making is that a similar argument could be made with respect to the role and importance of action planning, environment modelling, and rule induction under these circumstances. In Sim play, action planning can be present, but it's either a side-effect, non-competitor or means-to-the-end with respect with the larger goal of maintaining an internally-consistent imagined world. Conversely, in Nar play, environmental modelling can be present, but it's either a side-effect, non-competitor, or means-to-the-end with respect to the larger goal of developing or expressing an underlying emotional theme.
My point is that everything you could say about specific techniques or decisions that, considered in isolation, would lead you to conclude that someone's role-playing is "all three", arguably evaporates when you consider the larger context. A dedicated-Sim player will engage in action planning IFF their environmental model (or SIS) predicts 'that's what the character would do here'. The overall trends and patterns they'll look for and appreciate will be rooted in data available to their character- which is to say, consistent with the environmental model which states that Bob can't magically know what Sally is thinking, feeling or doing unless in a position to observe or overhear it. For a Sim player, the pleasure taken in maintaining a consistent environmental model (if one includes theory of mind as a subset of 'environment') trumps, contrains, and redirects the desire to infer patterns or surmount ambient challenges.
Quote
Although the overall thread is totally obsolete and rife with gibble-gabble, Jason Lee's comments show how induction, deduction, and abduction are integrated within Exploration (i.e. all role-playing) regardless of Creative Agenda.
Thanks for all the links- BTW, I will be sure to look them up and get back to you on this.
I agree that Exploration is crucial to all role-play, and in the same sense, one might naively argue that all role-play is partly Simulationist, and this would be perfectly consistent with the environment-modelling hypothesis. (I'm aware you've explicitly disavowed Exploration == Sim in the past by defining Simulationism as Exploration-to-the-exclusion-of-all-else, and I guess it's mostly a question of terminology, but I'm also not certain it's much more than a tautological declaration: You can't be 'partly entirely' something.) Again, it's possible that's a misconception on my part.
On the subject of the role of induction/deduction/abduction, I suppose that's a valid point. I guess, for example, that rule induction could be used to furnish either (A) mechanical insights about how the world works which could be used to extend the environmental model, or (B) rule-of-thumb heuristics for beneficial action under particular circumstances, which might be related to questions of morality or theme. Abductive reasoning (which I'd incline to call 'informed speculation',) is, I fully agree, something tied up with Exploration, and by extension, tied up with 'curiosity'.
Alfryd:
Ron- I just read through the threads you suggested. There's a lot of stuff there I can agree with.
"Constructive Denial", though. In a sense, I agree this is a useful insight, but in another, I think the term is so general as to be useless. Every CA has to establish rules and procedures that facilitate a particular form of play while compromising or 'denying' others. When the rules of 3E D&D state, in effect, that my Sorceror will never be particularly good with slings regardless of how often she practices with them, that is a 'denial' of a potential course the player might take. It's constructive from the perspective of Gamism (or at least a particular brand thereof,) but denial nonetheless. It's absurd to speak of Simulationist play as being inherently more abstemious in terms of potential choices available to the characters (and I sometimes feel that deprivation quite acutely during Gamist play.)
However, what I feel might be referred to here is the idea that, in Sim play, beyond a certain point, in-game events are supposed to be decided first and foremost by the in-game events that preceded them- that Internal Cause Is King. In that sense, the real players' personal choices are sorta not relevant, because the real players aren't part of the imagined world (which doesn't exist,) but rather a sort of 'distributed platform' for simulating a (mostly?) deterministic cosmos, and social interaction consists of a kind of mutual error-checking with respect to the computed results put forward by each individual.
So I guess, in that sense, I can see a qualitative distinction between Exploration and Simulationism- in that the former requires no active violation of established facts about the world, whereas the latter occurs when the 'lock in' factor of Exploration extends to the point where only one consistent 'vision' of events is really permissable.
Navigation
[0] Message Index
[#] Next page
[*] Previous page