Hi Phil, > But on systems, you say this research institute idea will let people > become nearly clairvoyant about how people will behave. Um, you suggested a research institute, not me. I would see this kind of project as largely modeling work to be done, with a strong focus on skillful applied people and domain experts.
A strategic simulation system is pretty well understood cognitive tool, but I admit I do imagine something bigger. Indeed, an architecture that can be used to model populations on a multitude of dimensions. For use from everything from urban planning to national security. Full GIS, automatic parallelism, optional object orientation, physics simulation features, clever constraint fitting and optimization, etc. Automated model validation with an expectation of huge underlying compute power. Not just numerically but also semantically, building on projects like http://www.opencyc.com. A while back DoD had a project called HLA (High Level Architecture) that was along these lines but I think it never really went anywhere. My main objection to that was that it was too concerned about federated simulation, i.e. pulling in legacy simulation codes. Also needed is a growing array of database resources; Everything from Lexus/Nexus to detailed map, scholarly works, census data, and whatever else would be available to classified users -- all integrated by some kind of unified query system. A big workbench where analysts can work efficiently to try to pin down algorithms for human behaviors and institutions they see or read about. Ideally there would be some open source core package developed for the public good that would seek to support many different kinds of users in academia, business and government. Perhaps that could be done at some invented institute, but probably better to actually try to ensure it gets done by hiring a credible contractor. With that momentum to get started, users could develop more open or proprietary modules and databases to develop a rich ecology for modeling the human world. Could such a thing make decision makers clairvoyant? Of course not. But it could pull everything together in one place and help the people that support those decision makers look at a problem at a range of scales, and consider alternatives systematically. Marcus ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
