Thanks Erika and Helen, "the issue isn't the accuracy of the bombs you have but how you use the bombs you have, indeed if you should use your bombs at all"
I liked this. : ) R On 03/11/2011 13:33, Erika Jean Lincoln wrote: > Two Ted talks featuring Malcolm Gladwell, demonstraits ideas about "the > model" one on the norden bombsight, modeling warfare > and the other on spagetti sauce, modeling consumer goods > > http://www.ted.com/talks/malcolm_gladwell.html > > http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce.html > > Erika Lincoln > Electronic Media Artist > Winnipeg/Manitoba/Canada > http://www.lincolnlab.net > > > ----- Original Message ----- > From: helen varley jamieson<[email protected]> > To: [email protected] > Cc: > Sent: Thursday, November 3, 2011 6:57:27 AM > Subject: Re: [NetBehaviour] Why Models Are Always Wrong > > this guy is an economist who seems to use models to prove their > falibility (& has other great concepts like WMEs - weapons of mass > exploitation): http://02ae523.netsolhost.com/Home.htm > > > h : ) > On 2/11/11 11:03 PM, ruth catlow wrote: >> Hi Rob, >> >> I have been pondering and wondering along this line of thought for a >> while too. >> >> We watched Copenhagen a couple of nights ago - a play by Michael Frayne >> that takes as its basis, the mystery surrounding the wartime meeting of >> Niels Bohr and Werner Heisenberg. >> >> In the play Heisenberg says something like... >> >> 'applying maths to people is always interesting. 1+1 can equal so many >> different kinds of 2.' >> >> I note in all large institutions a move towards measuring things because >> they can be measured - the data is then often presented as if there is >> only one kind of '2'. (the equations lend great swagger potential). >> >> Diagrams (I suppose some of these are just another expression of >> equations) are just as dangerous. >> >> I got myself- continue to get myself into all kinds of trouble taking >> network topologies too literally. >> >> After all most social networks are not flat. A node is not a node is not >> a node and the links between nodes have different resistances and >> qualities (again whether social or material). >> >> > >> >> Even if you have a perfect economic model with perfect data, calibrating >> the model doesn't have a single answer. So even in a perfect world, >> economic models can and will give incorrect answers.. >> >> < >> >> My feeling is that economists understand this very well and that their >> models are systems for gaining and maintaining temporary advantage- just >> keep moving fast, inventing new refinements of the tools that only you >> have access to. >> >> is it that scary? >> >> : ? >> Ruth >> btw ps. if anyone has a copy of the Copenhagen play I'd really like the >> exact quote. >> >> >> >> On 31/10/2011 20:18, Rob Myers wrote: >>> As a child I could never work out what a + b = c meant. I mean what >>> should a and b and c *be*? Of course the answer is that they have any >>> number of possible values. That's the power of equations. But when you >>> are trying to fit equations to reality, it can be a problem: >>> >>> http://www.scientificamerican.com/article.cfm?id=finance-why-economic-models-are-always-wrong >>> >>> Even if you have a perfect economic model with perfect data, calibrating >>> the model doesn't have a single answer. So even in a perfect world, >>> economic models can and will give incorrect answers. >>> >>> Bear in mind that both the financialized, high-frequency-trading, casino >>> capitalism side of the economy and the >>> government-management-of-the-economy side are based on ever more complex >>> models. >>> >>> And I think there's a more general problem that this illustrates. >>> Presumably it's not just economics but philosophy, politics, theology, >>> aesthetics, any kind of quantitative model of reality will suffer from >>> this problem. >>> >>> Which is a scary thought. >>> >>> - Rob. >>> _______________________________________________ >>> NetBehaviour mailing list >>> [email protected] >>> http://www.netbehaviour.org/mailman/listinfo/netbehaviour >>> >> _______________________________________________ >> NetBehaviour mailing list >> [email protected] >> http://www.netbehaviour.org/mailman/listinfo/netbehaviour >> > _______________________________________________ NetBehaviour mailing list [email protected] http://www.netbehaviour.org/mailman/listinfo/netbehaviour
