|
xii There is little value in a magical machine that spews out scenarios that are merely descriptions of some possible state of the world; we need to be able to understand how such developments might occur and what their implications might be.
xiii Our second experiment... required even more iteration and contemplation because the structure that we began with was "static," a set of possible situational attributes. This proved inadequate to the purpose of MSG [Massive Scenario Generation], and we quickly concluded that the appropriate approach from such a starting point was to construct quickly the sketch of a dynamic system model, even though not enough time was available to do so well. Once a relevant system-level "influence diagram" had been sketched, we could move to a first-cut dynamic model that was capable of both generating diverse scenarios and providing enough context and history to enable the scenarios to be more like significant causal stories... our conclusion... was that the method showed significant promise. The key, however, was to recognize the importance of constructing a model early on, even if it could be only at the level of influence diagrams and initial rules of thumb. Taking that step changes the entire direction of scenario generation and provides a core that can be enriched iteratively through hard thinking, brainstorming, gaming, and other mechanisms.
p.3 The first step in strategic planning's divergent thinking is perhaps the most important: breaking the shackles that bind us to canonical [JLJ - authorized; recognized; accepted] images of the future. The best known planning methods for doing so involve scenarios. The word scenario has diverse meanings but is best understood as a postulated sequence of events with some degree of internal coherence, i.e., events associated with a "story."
p.6 The central theme that has emerged in this work at RAND on CBP [capabilities-based planning] and uncertainty-sensitive planning is the need to emphasize strategies that are flexible, adaptive, and robust - i.e., FAR strategies - rather than strategies tuned to a particular expectation of the future.
p.6 Strategies with simple built-in rules for adaptation are often superior to strategies based on a particular concept of the future.
p.10 A Scenario Generator
After obtaining a model, the process calls for a computer device, a "scenario generator," to create an ensemble of scenarios sampling the space implied by the specification of an experimental design. The resulting scenarios are not simply a re-expression of inputs, because the specification may require considering all plausible combinations of certain variables, only some of which have previously been imagined explicitly. Further, allowance may be made for "wild-card" events. [JLJ - very useful.]
p.15 More generally, abstractions play a key role when we attempt to make sense of complexity... Although detail is sometimes necessary, comprehending issues is best done with a more abstracted view of the problem... we know about the importance of using stories to convey concepts that transcend the stories themselves. We also know the importance of visualization, experiential learning (as in war-gaming), and puzzle solving.
p.19 if we seek to understand the results at a deeper level, what we actually seek is a simple theory, not just some visual patterns. Arguably, such a theory would take the form of either a relatively simple and understandable mathematical equation or a relatively simple computer model.
p.19 Motivated metamodeling is the name that has been given to an alternative approach... In this approach, one draws upon an understanding of the problem domain to hypothesize a reasonable, albeit approximate, analytical structure... This is a new name for an old method, one often used by scientists.
p.45 In reality, the same response would probably have a number of different effects. depending on contextual details (including essentially random factors).
We responded to this dilemma by translating influence-diagram relationships into uncertain model relationships.
p.49 As with human gaming and simulation generally, one often learns something insightful that afterward - but not before - seems intuitively obvious.
p.50 the experiment strongly confirmed the notion that one should take a model-building approach from the outset... Causality is essential to the generation of stories, explanations, and insights, but causality relationships are often difficult to uncover
p.51 the ideal endpoint would be not an answer machine... but rather a sound model for at-the-time MSG applications that could help guide planning under uncertainty.
|
