Jan, On Fri, Jul 30, 2010 at 4:47 PM, Jan Klauck <jkla...@uni-osnabrueck.de>wrote:
> > This brings me to where I came in. How do you deal with irrational > > decision > > making. I was hoping that social simulation would be seeking to provide > > answers. This does not seem to be the case. > Have you ever taken a dispute, completely deconstructed it to determine its structure, engineered a prospective solution, and attempted to implement it? I have. Sometimes successfully, and sometimes not so successfully. First, take a look at Nova's *Mind Over Money* episode: http://video.pbs.org/video/1479100777/ The message here isn't so much that people create unstable systems around themselves, but rather, that the (present) systems sciences predictably lead to unstable systems. Getting people to act in what may seem at the moment to be ways that are contrary to their interests is a MAJOR challenge. Indeed, much of the AGI discussion on this and other forums concerns ways of *stopping* AGIs from effectively intervening in such instabilities. How can you, the participants on this forum, hope to ever bring stability to our world when one of your own goals is to preserve the very sources of those instabilities? IMHO the underlying problem is mostly too limited of intelligence in most people. They are simply unable to comprehend the paths to the very things that they are seeking, and hence have absolutely no hope of success. You can't write a good Chess playing program unless you have first been a serious chess player. Similarly, I suspect that demonstrated skill in IR is a prerequisite to creating any sort of effective IR program. Hence, I would welcome an opportunity to play on that field, as I suspect others on this forum might welcome. This should be facilitated, and then watch to see which approaches seem to at least sometimes work, and which seem to predictably fail. Once past this, I suspect that the route to an effective IR program will become more obvious. > > Models of limited rationality (like bounded rationality) are already > used, e.g., in resource mangement & land use studies, peace and conflict > studies and some more. > These all seem to incorporate the very presumptions that underlie the problems at hand. For example, the apparently "obvious" cure for global warming is to return the upwind coastal strips to forests and move human development inland past the first mountain range. This approach should turn the great deserts green (again), provide an order of magnitude more food, and consume the CO2 from all of the air and oil still in the ground, plus lots of coal in addition. Of course no one seriously considers this, because it involves bulldozing, for example, most of the human development in America between the Pacific Ocean and the top of the Cascade Mountains. While the rewards almost certainly exceed the cost, the problem is that the corporations who own these developments would commit limitless resources to influence "the best government that money can buy" to stop any such project. > The problem with those models is to say _how_much_ irrationality there is. YES. Some say that my proposal for bulldozing the upwind strips of the continents is irrational, not because it won't work, but because it hasn't been experimentally proven. Once past computer simulations, the only way to "prove" it is to try it. Judge for yourself which side of this argument is "irrational". > We can assume (and model) perfect rationality I don't think so! You may also question this after viewing the NOVA episode above. > and then measure the > gap. Empirically most actors aren't fully irrational or behave random, > so they approach the rational assumptions. What's often more missing is > that actors lack information or the means to utilize them. > In short, they lack a lot of everything needed to make rational decisions, not the least of which are rational questions to decide. Most "questions" in our world contain significant content of irrational presumptions, yet people feel compelled to participate in the irrationality and decide those questions. Any (competent) AGI would REFUSE TO ANSWER, and would first redirect attention to the irrational content of the questions. Steve ------------------------------------------- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244&id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com