Welcome to the new version of European Tribune. It's just a new layout, so everything should work as before - please report bugs here.
Display:
A college friend of mine became the Director of the Bureau of Trade Statistics in the Department of Commerce during the Nixon Administration and held that position into the mid '90s.  One of his primary responsibilities was maintaining the Bureau's computer model of the US economy as adapted to account for international trade.  I was fascinated by the idea of being able to "what if" policy options. With current models only a few people have the opportunity to make such queries.

It would seem that the explosion of computing power should make it possible to put such models into open source code that could run on relatively inexpensive platforms as a distributed computing system over a network or the internet.  Run in stand alone mode such a program would enable individuals to model the effects of various policies and events.  But a more interesting model might be a sort of economic "Second Life" where a multitude of separate agents take actions directed primarily towards the benefits to their particular institutional avatar and the combined effects on the individual avatars and the aggregate model of the economy could be observed.  This might require a greatly more powerful processor or it might be possible to accomplish via distributed processing.

I have a sense that this would likely give rise to useful management information and could lead to gainful employment for a whole new group of specialists.  Were it set up so that sub sets of the users could devise, employ and test new organizations and techniques in separate economy domains it could allow experimental meta-economics.  If these proved promising, they could be re-integrated into the basic group model and the effects of new agents acting on the existing system could be observed.

It might be important to brand, patent and copyright such software and commercialize it so as to make it more difficult to suppress or control by stasis seeking forces, or perhaps I am overly paranoid.  The overarching goal should be to make this as entertaining, exciting and insight generating as possible while protecting the open source nature and the right of all to use and modify the programs.  I do not know how to properly resolve these issues, let alone how to create the models.  But it seems to me that such programs could revolutionize the way we conceptualize and actualize our common culture.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Thu Apr 23rd, 2009 at 08:40:01 PM EST
[ Parent ]
AFAIK, I'm the only one around here who has done something similar.  In my considered judgment this is a non-trival task.  Or, in English, "Abandon All Hope Ye Who Enter Here."

The problem isn't the computing power.  The problem is the software that tells the hardware what to do, when.  To go anywhere beyond a toy you'll eventually have to deal with phenomenological relevance and THAT, my friends, is a real bitch¹.  The usual method is to ignore it by building some flavor of Expert System.  Now the problem with that is well known: anything not specifically programmed is deemed irrelevant which runs you into all kinds of limitations.  The other method, which doesn't work either, is to implement some kind of a randomizer decision process.  In any event, eventually you run smack into the Lyapunov Exponent which can be this:

or this:

or this:

or this:

Depending on what is you're dealing with and how you're looking at it at any particular moment with your chosen methodology at that moment.  Each of the above graphs are a different way of looking at - call it - Information Integrity and each of them are "right," for a given value of right-ness, AND each of them are "wrong," for a given value of wrong-ness.  Because this is a factal dimension the right/wrong versus right-ness/wrong-ness is neither linear NOR Euclidean, nor is the seemingly straight forward right/right-ness and wrong/wrong-ness².  

Just to complicate everything further the relationship of semiots ('words') to semantic (what the word points at, what it "means" on another axis) is 1:Many.  Technically there's no theoretical limit giving 1:∞ but humans get around this through applying arbitrary limitations, e.g., grammar, and verbal negotiating, a computer don't know either.

To quote Drew, "Wheeeeeeeeeeee!"

Throwing a lot of people at the problem, the Open Source route, isn't going to do a damn bit of good IF the goal is a explanatory model.  Even people with decades of experience and adequate funding can't - really - do a proper job.  (Ladies and Gentlemen  of the jury I give you the IPCC Model and Report in evidence.)  And it is very much an open question if the most efficient way forward would be to build a Artificial Life system using genetic algorithms (with or without classifiers?) and let the little buggers build the model for you.  

Because in the final analysis, We don't know how to write a good model of the economy.  Various good and brainy people have tried over the years and bounced every time.  

---------------------------------------------------------------------------------------------------- --------------
¹  I could bore you to death with why.  Suffice to say, when faced with n-dimensional streams of data and the odd datum popping in with varying potential and actual affective and effective implications merely doing the comparison operations takes you out of Real Time user response.

²  To Them Wot Know .... yeah, but YOU try it in 50 words or less.  ;-)

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Thu Apr 23rd, 2009 at 11:35:20 PM EST
[ Parent ]
I agree. But surely this is an argument FOR distributed citizen decision-making, not against? If the 'expert-few' can't model the economy, then why would the 'amateur-many' decisions be less useful?

You can't be me, I'm taken
by Sven Triloqvist on Fri Apr 24th, 2009 at 03:00:02 AM EST
[ Parent ]
Sven Triloqvist:
If the 'expert-few' can't model the economy, then why would the 'amateur-many' decisions be less useful?
What, the market knows best, now?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Fri Apr 24th, 2009 at 04:08:34 AM EST
[ Parent ]
If you want to describe the brain as a market, then, yes ;-)

You can't be me, I'm taken
by Sven Triloqvist on Fri Apr 24th, 2009 at 04:50:06 AM EST
[ Parent ]
AI metaphors are exceedingly bad when taken literally.

What I mean is that "experts can't know as much as everyone collectively" is the key fallacy behind market fundamentalism.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Fri Apr 24th, 2009 at 05:15:44 AM EST
[ Parent ]
Generally AI metaphors are lousy even if you take them as AI metaphors.

</cynicism>


Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Fri Apr 24th, 2009 at 11:06:55 AM EST
[ Parent ]
That's probably what I meant to say :-)

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Fri Apr 24th, 2009 at 11:14:07 AM EST
[ Parent ]
i think what no-one ever says is that you can't preview, prognosticate, or prophetise consistently about something as inherently nebulous or unpredictable as the sum of millions of individual choices.

it's a kind of Everest no-one has climbed, though many gaze at the summit longingly, and many have fallen off its slippery faces...

anyone that really consistently pulled it off would be able to name and number their income. with all that computers can do to ease our toil, it's understandable that man should want to use them to find the rosetta stone of economics.

until humans become completely linear creatures (too many heading that way imo), those who want to guarantee their winnings merely have to tithe a chunk to the lobbyists, pressuring politicians to hamstring regulators. it can get messy though, which is why everyone would like to find an algorithm to make them easy squillions, like a blackjack 'system'.

back to dreaming of the view from everest!

"We can all be prosperous but we can't all be rich." Ian Welsh

by melo (melometa4(at)gmail.com) on Fri Apr 24th, 2009 at 02:25:10 PM EST
[ Parent ]
Welcome to My World©

i think what no-one ever says is that you can't preview, prognosticate, or prophetise consistently about something as inherently nebulous or unpredictable as the sum of millions of individual choices.

No you can't.  What you can do is construct a decision-making environment that simulates¹ what might happen when this, that, or the other comes down.  Now, unregenerate cynics equate "simulate" with "pulling it out of the air."  I prefer to define it as "a systematic procedure resulting in a very informed guess."  :-)

Fortunately, there are certain patterns and breakouts - statistically, psycho-socially, and others - to human behavior that can help boundary the problem.  And, with a little bit of intellectual effort and humility, it's possible to get the general thrust of things.  Best example is Exponential Growth, if you've only got x amount of something, and you can only use x-y% of it then using the Growth rate it is possible to determine, more-or-less accurately, when you ain't got no more x.  When you ain't got no more x either you - and everybody else - does without or find a substitute.  That's a general law applicable across the whole range of human activities.

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Sun Apr 26th, 2009 at 12:34:31 AM EST
[ Parent ]
What, the market knows best, now?
Well, we have had markets longer than we have had writing and we have only recently tried to analyze these markets with mathematical or computer models.  Our brains have been dealing with complex, emergent phenomena for a long time.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri Apr 24th, 2009 at 11:26:51 AM EST
[ Parent ]
It's an argument for "democratizing" Model building, I guess.  Book published in the US last year comparing the Experts' forecasting with quality of forecasting.  Turns out the greater the "qualification," the more eddymacation the expert had in the area, the worse the accuracy of predictions.  That, at least, is the rumor.  I need to grab a copy, sit down, and read the sucker.  

Skepticism is the first step on the road to truth. -- Denis Diderot
by ATinNM on Fri Apr 24th, 2009 at 10:32:36 AM EST
[ Parent ]
That's part of the process.

We have a strange idea, which seems apply almost exclusively to economics, politics and business, that experts are good at predictions.

In reality almost everyone is better at making predictions than the experts are.

There are two things happening here. One is that expertise isn't based on predictive ability, but on following the party line. It's a perfectly soviet system, where talking rubbish gets you immense rewards.

The other is the corollary that decisions are made by the people who are least competent to run it - because the goal isn't to make intelligent decisions, it's to maintain power differentials for as long as possible.  

If you had a modelling and prediction game you could find individuals who had a talent for intuitive modelling and insight and give them something useful to do. They wouldn't necessarily have to know how to build models out of differential equations, but they would have a better than usual batting average when it came to being right about the future.

These people would potentially be very, very valuable. Currently I suspect most of them are wasted in jobs which aren't ideally suited to their talents.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Fri Apr 24th, 2009 at 12:31:28 PM EST
[ Parent ]
Make this a diary. Throw in a discussion of The Club of Rome's The Limits of Growth and their World3.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Fri Apr 24th, 2009 at 04:11:23 AM EST
[ Parent ]
I read the Club of Rome report in 1975 and haven't had a chance to look at what they are up to recently; thus, my thoughts about their work are worthless.  Tell you what, why don't you add the CoR stuff making the diary a Mig/AT co-production?

Skepticism is the first step on the road to truth. -- Denis Diderot
by ATinNM on Fri Apr 24th, 2009 at 10:23:55 AM EST
[ Parent ]
Me like. I have a copy of the 2nd edition somewhere in my mother's flat and would like to get the 3rd edition and the details of the model.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Fri Apr 24th, 2009 at 10:37:56 AM EST
[ Parent ]
Great exposition of the difficulties of modelling reality.  Just when you think you have all the key variables nailed down, something else pops up from left field.  

If our Government models of the economy were that good, predicting the sub-prime or derivatives crises should have been child's play.  Hell, many people did it without the benefit of a computer model.

The problem is that the assumptions you feed in tend to be regurgitated in a new more "objective" quasi-scientific axiomatic form - and come to be seen as the "natural" or "real" world - when all they were is assumptions, deliberate over-simplifications, for the purpose of making the system modelable.  

And sometimes, of course, they are much more than just over-simplifications, but deliberate political choices to emphasise some factors and ignore others because it suits a particular political agenda.  See the rise of neo-liberal economics to obscure the insights of Henry George who worked from different assumptions to achieve different objectives.

Having said all that, I would be a fan of making computer models/simulations of the economy more generally accessible, if only to better inform political debate.  The key thing to be aware of, howeverare the assumptions which underlie the model - as these can bias the entire result sets towards some options and against others.

Thus what is the impact of an increase in income tax on worker productivity is a valid question to ask.  But is it more or less important than: what is the economic impact of increased public expenditure on public health/education/social welfare?  And how do we measure the non-economic benefits of such expenditure?

notes from no w here

by Frank Schnittger (mail Frankschnittger at hot male dotty communists) on Fri Apr 24th, 2009 at 07:26:42 AM EST
[ Parent ]
The ability to model a local political-economy has been around for a long time: SimCity does as good a job as one can expect given their limitations and methodology. And since,

On January 10, 2008 the SimCity source code was released under the free software GPL 3 license under the name Micropolis.

that's where I'd start my research.  If I was going to do something ... which I'm not¹.  

SimCity, as is, does a good enough job that it is widely used by urban planners, city planners, & etc. as a tool.  Idly projecting here but it seems to me a neural network-like construct having lots of nodes - 200 or 300 - running SimCity autonomously tossing in cross-communication plus "macro-economic" influences together comprising a first-hack Fitness Landscape would prove interesting.  The easiest way to add complexity to the Fitness Landscape is to build some - say 10 to 20 - analog computers, cross-linkable with A/D and D/A conversion to/from the neural net to simulate continuous data streams which the neural net nodes have to 'chunk' - both for data processing and across time.

As an added benefit, you'd get the first major non-von Neumann computer built in a long, long time.  Which, by the way, is something we desperately need as a research tool.  

¹  Unless somebody wants to pay me to do it.  This gun for hire.  I can be had.  :-)

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Fri Apr 24th, 2009 at 11:04:30 AM EST
[ Parent ]
The ability to model a local political-economy has been around for a long time: SimCity does as good a job as one can expect given their limitations and methodology,..

Perhaps a model national economy, or series of national economies, could be constructed which consist significantly of the interactions of the city models, models of national social and economic policies, plus models of agriculture, mining and other resource extraction or harvesting activities.  Etc., etc.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri Apr 24th, 2009 at 11:48:18 AM EST
[ Parent ]
There is an interesting problem modeling resource extraction.  Resource extraction essentially it breaks into renewables and one-time consumables with the latter breaking down into recyclables, metals in particular, and use-once, e.g.,oil-as-energy.  

There's also two price points for the extractors:

  1.  Rape & Pillage, where the extractors do not have to pay for the ecological damage they create

  2.  Pay-for-It, where the extractors have to pay for the damage

The first is what we got.  The second is what we should be doing and, in some cases, is where we're headed.

So, how you program/process resource extraction depends on if the Model is what-we-got XOR "The Economy."

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Sat Apr 25th, 2009 at 10:14:07 PM EST
[ Parent ]
There's also two price points for the extractors

Creating a user choice or a continuum of choices between those two poles as a user option would rather quickly get the idea across to users in a city-state-with -hinterland basic Sim type game, especially with quarterly or yearly turn times.  Some versions of Railroad Tycoon modeled timber, coal, petroleum and uranium, along with associated industries, as available resource-industrial exploitation opportunities for the railroads.  Something like that could be incorporated, along with water and road transport.  Experienced game developers could probably produce addictive type games that would drive home the implications of some of the choices.  Call the game "Can you save the world?"

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sat Apr 25th, 2009 at 11:45:22 PM EST
[ Parent ]
Model ≠ Game

An accurate Economic Model would require hardware components that are not generally available.  Get around this by having the system internet 'savvy.'  Problem: this implies the Model will spend its time running -- let me put it -- "naive" simulations at the expense of more "interesting" ones.  For any given value of "naive" and "interesting."

It would also require running and managing a software system beyond the sophistication of the average computer user.  There's no getting around this one unless a 'brain damaged' version is released to the general public.

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Sun Apr 26th, 2009 at 12:04:31 AM EST
[ Parent ]
Were the system to incorporate a group of analog computers it would inherently require specialized hardware, anyway, and would be used for "serious" work by business, government and academia.  Game type applications would need to be rather restricted in scope, all digital but could be sold cheaply.  The idea there would be to get certain ideas across to the user with a sledge hammer.  Naturally that usually mangles something.

Your concerns regarding patents agree with my own experience.  And after you get one, it is only a license to sue.  Copyright may be more useful.  Putting patentable ideas firmly in the public domain is probably the best approach there.  Aside from possible profits I find the whole idea fascinating.  I only whish I were better able to contribute on the programming side.  But that is a hopeless prospect at this point in my life.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sun Apr 26th, 2009 at 12:29:39 AM EST
[ Parent ]
This thing is a Blue Sky pipe-dream.

My informed WAG is a project of this magnitude would require serious funding: at least in the $25 to $50 million range.  (Hardware engineering is incredibly expensive.)  And with a small chance of re-couping the $ injected.  Certainly no private funding would be forthcoming and we're not USDA Stamp-of-Approval Serious People® so government funding is out as well.

It might re-coup the money in spin-offs: games, books, seminars, & etc., but "might" cuts no checks.  


Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Sun Apr 26th, 2009 at 12:48:56 AM EST
[ Parent ]
I am not sure what you are saying (in general) about modeling an 'economy': that it is impossible, too expensive, or something else?

Economies are modeled - they have to be modeled in some way, for governance. That is why there are national budgets, and offices of budget and management. Now, these models may be imprecise, contentious or too narrow, but they are used to predict the effect of changes in the rules of different types of transactions of citizens, companies, organizations, government and 'external transactions' (i.e. how these rules might interact with the rules of other budget systems in other nations).

However 'inaccurate' such national budgets may be, it is correct to say that building a parallel one would an expensive project. However, the Dutch government is already considering the use of XBRL to 'code' their national budget in such a way as to allow any citizen to play 'what if' with it. The mandated use of XBRL would also bring transparency to corporate accounting, because the tax authorities could also play 'what if' with those accounts. Stress testing, I suppose you'd call it ;-)

An explanation of XBRL is here. There's an XBRL conference in Paris 23-25 June.

So there may be a 'game' to be played in the future that does not require the building of the model.

TBG's 'game', as I understand it, is not about producing a perfect, but coarse, real 'what if' modeled budget, but a simulation of the budget decision processes so that citizens can better understand the processes and thus vote in a more informed manner. Lemonade Stand 3.0.

You can't be me, I'm taken

by Sven Triloqvist on Sun Apr 26th, 2009 at 03:17:07 AM EST
[ Parent ]
For a brief while in a prior incarnation I "managed" (more like conceived of and rode herd over) a project to develop a digital editor for the 3M Digital Audio Mastering System, a 32 track digital audio tape recorder.  The overall system design for the system had been developed by the BBC to allow digital transmission of 16 bit, 50 KHz program streams over the air to remote transmitters.  

It employed a Hamming Code to protect against lightning induced drop outs.  This involved taking two words a chosen interval apart, adding them into a check-sum and recording that check-sum down stream by the same interval.  As a consequence one could punch holes in the tape about the size of a paper punch and not loose any information.  The problem was that a tape splice edit produced a giant pop.

I spotted a TRW 16 bit multiply-accumulator that could operate with a 110 nano-second period.  Back-of-the-envelope calculations showed that one such device was sufficient to perform more than 32 channels worth of such operations in the 20 micr-second period of one 16 bit sample period.  "We" got a contract to develop such an editor from 3M and delivered a working system in about one year.

I hired one hardware and three software engineers on a part time consulting basis.  The programmers all worked for a local microprocessor based manufacturer of telephone transmission test equipment.  My hardware guy was a MIT EE who worked for aerospace.  This was an evenings and week-ends project and all kept their day jobs.  This was, IIRCC, 1978, and we paid our consultants $35/hr.  It was exhausting for them but they more than doubled their income for that year. Me, not so much.  I should have demanded their deal!

We were using Motorola 6800 microprocessors and I had partitioned the system into four separate, interacting processor systems.  One system consisted of the edit hardware, which executed digital cross-fades using stored PROM coefficient tables, MSI logic and the TRW chip, another was the tape machine controller, another was a SMPTE time code reader and sync machine that could synchronize two machines for assembly edits and the fourth was the user interface machine.

The actual cost of the whole project was around $500,000, including hardware.  3M eventually supplied us with recorders on which to test our system.  I don't know what "we" charged 3M.  I was "just" the project manager.  I understood the application and conceived the over all approach and then got burned for my efforts.  If "we" charged 3M $1,000,000 at that time it would not have been too outrageous.

In today's environment there are probably lots of guys and gals who would be quite happy to make $80,000 for a year's work and who have the knowledge but not the day job.  Make that $100,000, including paid medical of the caliber of Kaiser and one would have choices.  One or two hardware guys, depending on whether one engineer could adequately handle both the digital hardware and the analog function modules, two or three programmers and overhead and we would be under $2,000,000/year for development of the full blown "professional" system.  A secondary focus on a consumer type game could produce revenue within a year or so.

I have always liked blue skys.

 

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sun Apr 26th, 2009 at 01:59:05 PM EST
[ Parent ]
There may be some mileage in developing an Open Business model at the same time.

You'd need some start up cash - probably $500,000 or so - but you certainly wouldn't need $50m.

Second Life charges outrageous land rents - a lot of people are paying >$200/mth for a not very powerful virtual server.

I don't think cash is the issue. Nor are processor cycles - clouding should give you all the cycles you want. It might not give them instantly, but some lag wouldn't necessarily be a huge problem.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Tue Apr 28th, 2009 at 06:57:01 AM EST
[ Parent ]
¹  Unless somebody wants to pay me to do it.  This gun for hire.  I can be had.  :-)
Another reason to copyright and patent as possible.  Prior to release as open source why not provide fair compensation to the programmers?  

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sat Apr 25th, 2009 at 08:50:05 PM EST
[ Parent ]
Can't patent software in the EU.

Patents and copyrights are essentially unenforceable against Chinese companies in China.

Before you get a patent in the US the USPTO publishes the patent on the internet for "comments and objections."  DeepPockets, Inc. regularly scans the site files objections using their patent staff, and forces the applier into a complex, expensive, legal battle which small companies can't afford.  A filed patent that is rejected by the USPTO is automatically in the public domain, meaning DeepPockets, Inc. gets to use it for the cost of keeping their patent staff around, which they have to do anyway.

In short: if you have a good idea, Shut Up.

Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Sat Apr 25th, 2009 at 11:45:34 PM EST
[ Parent ]
Um, why not set up one of Chris Cook's LLPs? If the thing makes money you get compensated, if it doesn't nobody loses anything but their time...

Chris, can we have the story of that film you made with an LLP, again?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Sun Apr 26th, 2009 at 04:05:46 AM EST
[ Parent ]
Well, you can find it in here

Social Investment Business

The "Common Source" approach I advocate is outlined there, and that is to put IP into the hands of a Custodian, and then to encapsulate all the rights of use and usufruct/use value within a Master Partnership agreement or protocol.

The outcome is new form of co-ownership or Common Source property right of indefinite duration. Essentially the entire Creative Commons property relationship is then encapsulated within a Corporate personality.

The result is an enterprise model which is both Open and Closed. It is Closed because only LLP members can use it, and  Open because anyone who consents to the agreement may be a member.

It's worth pointing out that if all the stakeholders are members, then limitation of liability is unnecessary. That is why I wax lyrical about the LLP being the first example of an Open Corporate. (Cost)-Free limitation of liability is IMHO a very dubious add-on in terms of the public interest.

"The future is already here -- it's just not very evenly distributed" William Gibson

by ChrisCook (cojockathotmaildotcom) on Sun Apr 26th, 2009 at 06:28:10 AM EST
[ Parent ]
The easiest way to add complexity to the Fitness Landscape is to build some - say 10 to 20 - analog computers, cross-linkable with A/D and D/A conversion to/from the neural net to simulate continuous data streams which the neural net nodes have to 'chunk' - both for data processing and across time.

Analog computers with >90dB dynamic range working with 16 bit A/Ds and D/As could give really fine levels of resolution and speed.  And your point about non-Van Neumann modeling systems is important.  The ability of human controllers to make inputs to either the analog or digital models based on intuition or just trying to crash the system would also be valuable. This might be interesting enough for someone to write a grant for funding the initial phases.  What about Bob? :-)

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sat Apr 25th, 2009 at 11:57:49 PM EST
[ Parent ]
The neat thing, for us, about analog computers is they are "noisy."  Even the A/D and D/A conversions interject noise.  Great!  We're dealing with a thundering horde of lousy data and there's little point in pretending any different.

Skepticism is the first step on the road to truth. -- Denis Diderot
by ATinNM on Sun Apr 26th, 2009 at 12:10:32 AM EST
[ Parent ]
I know someone doing research into analog computers. They're much, much faster than anything that would be needed here. And - using his approach - very cheap too.
by ThatBritGuy (thatbritguy (at) googlemail.com) on Tue Apr 28th, 2009 at 06:58:41 AM EST
[ Parent ]
For a long time I've toyed with the idea of buiding an analog computer...

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Tue Apr 28th, 2009 at 08:03:58 AM EST
[ Parent ]
The ability to model a local political-economy has been around for a long time: SimCity does as good a job as one can expect given their limitations and methodology. And since,
On January 10, 2008 the SimCity source code was released under the free software GPL 3 license under the name Micropolis.

that's where I'd start my research.  If I was going to do something ... which I'm not¹.  
How clean is the SimCity codebase?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Sun Apr 26th, 2009 at 04:02:50 AM EST
[ Parent ]
How clean is the SimCity codebase?
And how well annotated is it?

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sun Apr 26th, 2009 at 04:51:05 PM EST
[ Parent ]
According to the wiki article the Micropolis codebase is a reworked C++ version of the original C code of the X11 port of SimCity... So, there is hope...

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Mon Apr 27th, 2009 at 04:03:19 AM EST
[ Parent ]
predicted both the stock market slide and the financial debacle some months ahead of the events. You can read Williams' predictions without subscription in the archives on his web-site,  www.shadowstats.com  

Many of us here at ET and elsewhere recognized the 'signs' ahead of events, as well. I switched my 401k to a foreign-stock fund in time to enjoy a little bounce there, then into a gov't bonds/cash fund in time to save my financial butt - mostly because of discussions here and other similarly oriented blogs.

I'm not saying that I don't think that there are Black Swans, but I find the world a whole lot more predictable than not. Marx and Henry George may have had different perspectives, but I can still find salience in both.

In any case the economic 'game' can be 'won' in a variety of ways, but I want to found the system on both moral and practical bedrock: Solidarity and Sustainability. Don't we all?

I think that, if we start with that, we'll do well enough, whether the result is Chris' system, socialism, entrepreneurial capitalism, or some combination (which is my position). And with respect to this conversation I think that choosing some systemic approach allows us to 'cheat' the myriad curves into a path/math that is linear enough to both create a plan and to measure progress.

As to the current economic situations around the world - and the overall interplay of such - it's not that tricky. If we add the analyses and critiques of Krugman, Williams, Jerome, and most of the rest of us here - which share many elements - we cover the picture well enough.

paul spencer

by paul spencer (spencerinthegorge AT yahoo DOT com) on Fri Apr 24th, 2009 at 11:04:31 AM EST
[ Parent ]
I would be a fan of making computer models/simulations of the economy more generally accessible, if only to better inform political debate.  The key thing to be aware of, howeverare the assumptions which underlie the model - as these can bias the entire result sets towards some options and against others.
The key to this would be having the code open source AND having an extensively and collaboratively annotated code.  The point of the endeavor would be to examine, test and refine the assumptions and how they are implemented in code.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri Apr 24th, 2009 at 11:08:12 AM EST
[ Parent ]
Something in the LISP family... (before you roll your eyes at this, consider it includes LOGO and R...)

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Sun Apr 26th, 2009 at 04:01:00 AM EST
[ Parent ]
Ah, memories.  All your GNUS R belong to us.

Now where are we going and what's with the handbasket?
by budr on Sun Apr 26th, 2009 at 01:10:12 PM EST
[ Parent ]
I have no doubt that the problems are non-trivial.  An earlier vision of the same type of approach was Kenneth Arrow's vision of modeling societies, sometimes referred to as "Peasants under Glass" back in the '90s.  Were it easy it would have been done by now.

I don't think that a complete solution is necessary.  The one thing I remember from college is that complex equations are most easily solved at boundaries.  In the case of Raleigh's wave equation for acoustics, which involves the interrelations of pressure variations and the variations of air movement or "volume velocity," this is typically done at a rigid boundary or wall at which volume velocity is zero.  An amazing number of practical measurements and useful techniques come out of this simplifying assumption.

A similar approach might prove fruitful in economic modeling.  What I had in mind was a system of interacting agents in which the A.I. agents automated certain aspects of their response with other agents but in which the controller for each agent could intervene on whatever basis seemed appropriate to that controller.  The goal would be to define sub-sets of assumptions that would tend towards stability under a wide range of perturbations, including the desire of individual agent controllers, or players, to win.

In this fashion it may be possible to empirically move from very crude and simple systems to increasingly complex systems while maintaining stability.  In such a game I would at best be a mediocre player.  

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri Apr 24th, 2009 at 11:03:41 AM EST
[ Parent ]
What I had in mind was a system of interacting agents in which the A.I. agents automated certain aspects of their response with other agents but in which the controller for each agent could intervene on whatever basis seemed appropriate to that controller.  The goal would be to define sub-sets of assumptions that would tend towards stability under a wide range of perturbations, including the desire of individual agent controllers, or players, to win.

Well I want a sound reproduction system with 100% accuracy.  So ... GET TO WORK, dammit!

(LOL)

Your requirement are easy to 'spec' but - alas - a wee tad harder to accomplish.

;-)


Skepticism is the first step on the road to truth. -- Denis Diderot

by ATinNM on Fri Apr 24th, 2009 at 11:15:07 AM EST
[ Parent ]
Well I want a sound reproduction system with 100% accuracy.  So ... GET TO WORK, dammit!
Well, at least the initial phases of the "research", if carried out with a number of subjects, would be enjoyable, but the long term expenses would well exceed my budget, and my wife would object, to say the least.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri Apr 24th, 2009 at 11:35:00 AM EST
[ Parent ]
And of course for the audio reproduction system, in analog systems at least, been there, done that to a significant degree.

As the Dutch said while fighting the Spanish: "It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sun Apr 26th, 2009 at 12:02:22 AM EST
[ Parent ]
Okay, guys, I'm hijacking this subthread from this point downwards and sending it here.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Sun Apr 26th, 2009 at 03:50:25 AM EST
[ Parent ]
It might be important to brand, patent and copyright such software and commercialize it so as to make it more difficult to suppress or control by stasis seeking forces, or perhaps I am overly paranoid.

You make the source code open source so the codebase cannot be hijacked. But what you want to prevent is for someone to use the code to set up their own alternative game and through more successful branding, drown out yours and subvert the political purpose of creating the game in the first place.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Sun Apr 26th, 2009 at 03:56:13 AM EST
[ Parent ]

Display:

Top Diaries

Occasional Series

24 July 2014
by dvx - Jul 23
59 comments