« Should the government peg the S&P 500? | Main | Chess is now a young man's game »
Intelligent agent modeling
I am more optimistic about intelligent agent modeling than is Tyler. For one we already have an important, convincing, and Nobel-bestowed variant of intelligent agent modeling, namely experimental economics. Experimental economics uses one particular type of intelligent agent, the type based on...genetic algorithms. True, the intelligent agents used in I-A models are typically not as sophisticated as the agents used in experimental economics but they are rapidly improving. (Moreover, such agents are already important economic actors in their own right in limited areas, e.g. portfolio insurance, and they will continue to become more important as time continues.)
I see bringing experimental economics and I-A modeling closer as an important goal with potentially very large payoffs. Here, for example, is my model for a ground-breaking paper.
1) Experiment
2) I-A replication of experiment (parameterization)
3) I-A simulation under new conditions
4) Experiment under the same conditions as 3 demonstrating accuracy of simulation
5) I-A simulation under conditions that cannot be tested using experiments.
Now that would be a great paper. I-A agent modeling is already very useful for modeling contagion, peer effects, and highly non-linear environments. It will become even more useful when combined with experimental economics in a way that demonstrates the equivalence of the two types of intelligent agents.
Posted by Alex Tabarrok on December 31, 2008 at 07:39 AM in Economics, Science | Permalink
Comments
Hmm.. see chapters 3 and 4 of Douglas North's Understanding the process of economic change. People are not computers and human cognition does not work with 'genetic algorithms'. Pretending people can be accurately characterized as fleshy number chrunching machines will I believe add very little to the understanding of human behavior. The reason why IA modelling is becoming popular is not scientific but status potential: it's a apparent substitute for math skills cq analytical intelligence. IA is getting popular at our sociology department, for example.
Posted by: JSK at Dec 31, 2008 7:58:29 AM
Sure that would be a great paper, but Tyler explains well why that isn't going to happen. Agent models may be rapidly improving, but the are still a long long way away from capturing enough human complexity to predict well.
Posted by: Robin Hanson at Dec 31, 2008 8:09:27 AM
On complexity/suitability, it depends on what you're trying to model. There's lots of intelligent agent modeling and simulation going on now in wireless networks in a field called cognitive radio and we're modeling exactly what will be fielded.
There intelligent agents control the operation of say WiFi routers or cellular base stations and implicitly or explicitly compete for resources (e.g., spectrum or power budget).
Beyond Tyler's equilbria, we also end up caring alot about both parameterization effects (as Alex alludes to, but parameterization also includes variations in reasoning capacity and information availability), but also convergence and stability. As we pass through a recession and experience commodity price spikes and collapses, it should be clear that even in traditional economics we should care at least as much about how we get to (if we get to) an equilbrium as the existence of an equilibrium.
So what Alex proposes is a) being done, b) being done within a game theoretic framework, and c) looking at more than he supposes.
However, it's being applied to model intelligent agents in interacting machines (wireless devices) as opposed to models of interacting humans, which has the benefit that the modeled agent can be the real agent (just move software from simulation to the machine). There's a longish tutorial on the topic at this link with some simulation examples / discussion as well.
Posted by: Jody at Dec 31, 2008 8:28:41 AM
"I am more optimistic about intelligent agent modeling than is Tyler."
You're in good company. Here is Frank Hahn on the subject:
"The Next Hundred Years", The Economic Journal, Vol. 101, No.404. (Jan., 1991), pp. 47-50.
"I am pretty certain that the following prediction will prove to be correct: theorising of the 'pure' sort will become less enjoyable and less and less possible.
[...]
Instead of theorems we shall need simulations, instead of simple transparent axioms there looms the likelihood of psychological, sociological and historical postulates.
[...]
In this respect the signs are that the subject will return to its Marshallian affinities to biology. Evolutionary theories are beginning to flourish and they are not the sort of theories we have had hitherto. In particular, biologists have always known that, say, the giraffe was not inevitable. There are many routes evolution could have taken even in stationary environments. But wildly complex systems need simulating.
[...]
Evidently one could go on in this vein for a long time. My contention is not that twentieth-century sheds no light, nor indeed that its methods will not continue to provide some illumination.
But it is my prediction that the latter will increasingly be found to be too faint in the search for answers to questions which have quite naturally arisen from twentieth-century theoretical developments.
Not only will our successors have to be far less concerned with the general (leave alone the 'generic') than we have been, they will have to bring to the particular problems they will study particular histories and methods capable of dealing with the complexity of the particular, such as computer simulation. Not for them the grand unifying theory of particle physics which seems to beckon physicists. Not for them, or at least less frequently for them, the pleasures of theorem and proof. Instead the uncertain embrace of history and sociology and biology. Unfortunately, as recent work by biologists shows, these subjects are still in a state in which they can learn from our own past work rather than teach us new tricks."
Posted by: Alex at Dec 31, 2008 8:41:44 AM
This, of course, assumes that experimental economics doesn't suffer from the same or worse flaws as I-A modeling for anything more complicated than a dollar auction.
Posted by: Ted at Dec 31, 2008 9:00:28 AM
Does the GMU Econ program interact much with the Social Complexity program at GMU's Krasnow Institute? I would think the faculty of the two departments could produce a lot of interesting research along the IA topic.
Posted by: Sam at Dec 31, 2008 9:02:55 AM
Another variation on what is suggested here by
Alex T. is being done in some places already is
to have experimental subjects interact with
computer simulation programs in various ways.
One place this has been done, and which has
produced papers published recently in JEBO,
is CeNDEF at the University of Amsterdam, founded
and directed by Cars Hommes, who has coauthored
important papers with that old complexity theorist,
William A. Brock.
Robin Hanson is right, of course, that prediction
remains the name of the game, and we shall just
have to wait and see. These sorts of models have
been able to replicate real world phenomena that
more standard traditional models have not, as I
argued in the other thread on this. But, prediction
remains as elusive and difficult as ever (and the
simple models touted by Tyler, do not necessarily
do all that much better, hence so much of the
economist-bashing we see going on right now all
over the place).
Happy new year, everybody.
Posted by: Barkley Rosser at Dec 31, 2008 9:11:47 AM
JSK,
Just because people aren't genetic-algorithm intelligent agents doesn't mean that we can't use them to model how people behave in markets. A model is simply a simplified approximation of reality. The advantage, as I see it, is that I-A models are both simpler (in their mathematical description) and have the potential to be more accurate than general equilibria models. I-A models also can be used to see under what conditions equilibria can change, which is somewhat difficult to do otherwise.
Posted by: quanticle at Dec 31, 2008 9:39:15 AM
"Agent models may be rapidly improving, but the are still a long long way away from capturing enough human complexity to predict well."
I think Tyler's criticism wasn't that agent models are insufficiently complex. Rather his point was that he doesn't really think complex models are necessarily better because in all complex models you have to make a large number of assumptions. And as he put it you generally get out what you put in. In other words models just end up reflecting the intuitions and biases of the modeller.
To Tylers criticism there are several other related criticisms I would add:
1) more complex models tend to overfit and thus are often worse at prediction since they have higher variance. This is the reason why many in econometrics prefer parsimonious models.
2) more complex models result in more possible points of failure. Every assumption in the model is another point of failure.
3) The ugly truth that nobody will admit is that computer simulations strongly encourage "curve-fitting" by practitioners. Imagine a grad student. He has a paper due in 2 weeks. He runs his IA simulation and he gets garbage results. So he tweaks some parameters and reruns the simulation. Again crap. Tweaks and reruns. Now he gets plausible results but not what he wants. So he tweaks and reruns. After a large amount of tweaking he publishes his paper with some interesting results but he never mentions all the tweaking he did. Some of his parameters are embedded in his code and he doesn't even report them. For other parameters he comes up with some plausible reason why he chose them but after the fact. His code is unpublished.
3) is the main reason I don't believe in global warming GCM models.
Posted by: assman at Dec 31, 2008 10:06:07 AM
Just as a side note, Genetic Algorithms are just one way of modelling adaptive behavior - there is plenty of other methods from very simple (reactive behavior, simulated annealing) to more complex (e.g. Bayesian forecasters).
Besides there is a large body of literature that use GA (and other of the aforementioned methods) in Economics. For example, if I understand correctly your proposition for the paper, something along those lines was done at Caltech - search for Turing tournament.
Posted by: EP at Dec 31, 2008 10:18:46 AM
"Does the GMU Econ program interact much with the Social Complexity program at GMU's Krasnow Institute? I would think the faculty of the two departments could produce a lot of interesting research along the IA topic."
There is some interaction but not enough yet. Richard Wagner and Rob Axtell, along with a number of graduate students, have helped to bridge the two departments. Look for more crossover in the coming years. I have been trying to get the Austrians and the complexity folks to see each others viewpoints, and there are a number of other heterodox-complexity economists who interact with both camps.
Posted by: liberty at Dec 31, 2008 12:37:09 PM
Prof. Tabarrok a comparison of experiments with human vs artificial agents was done in the paper by Gode and Sunder (1993) JPE that I mentioned previously. This was within the context of double auctions. The artificial agents behave as humans when they are constrained. True few work like this has been done since. In CeNDEF experiments with human subjects on expectations are been developed in order to provide artificial agents with more credible expectations rules.
The idea that is around these days is about doing hybrid experiments at the same time with humans and agents, and even in far away countries or cultures I must say.
Nobody within the complexity/ABM community would deny that if your research question is answered with standard tools you should use them. What they emphasize is that boundedly rational agents purposively acting do not fit conventional modelling tools. Emergence or evolving market structures cannot be understood either with those tools. SO if you care about time in economics, capital heterogeneity, disequilibrium, fat tails and price volatility, you should be paying attention to these unorthodox tools.
HAPPY NEW YEAR!
Posted by: Pedro P Romero at Dec 31, 2008 12:44:09 PM
DEAR JSK,
IA is also getting popular in REAL scientific fields such as biology, physics, chemistry, etc.
Analytical intelligence in economics is something akin to cornering the market on 1950's mathematics
Keep perfecting that buggy whip---
I prefer Economics 2.0
Posted by: jsk2.0 at Jan 1, 2009 12:32:07 AM
this is an interesting discussion. however, it looks like all the intelligent agents are busy playing chess in the next thread, so i wouldn't expect much to get done here.
Posted by: babar at Jan 1, 2009 10:02:44 AM
租車
Posted by: 租車 at Apr 23, 2009 10:09:25 PM
租車
Posted by: 租車 at Apr 23, 2009 10:11:43 PM
Mabinogi online gold
Mabinogi money
Mabinogi Gold
cheap Mabinogi gold
buy Mabinogi gold
2moons dil
2moons gold
buy 2moons dil
2moon dil
cheap 2moons dil
Flyff gold
flyff Penya
flyff money
buy flyff penya
cheap flyff penya
cheap flyff gold
aion gold
buy aion gold
cheap aion money
aion money
cheap aion gold
Dofus kamas
buy dofus kamas
cheap kamas
dofus kama
dofus gold
dofus money
Knight Online Gold
Knight Gold
Knight Noah
Knight Online Noah
Posted by: aion at Jul 7, 2009 3:04:34 AM
