Model Metropolis

https://logicmag.io/06-model-metropolis/

by tekromancr • 7 years ago

112 60 7 years ago

33a• 7 years ago

I agree with the general sentiment of this sort of objection, but where it goes off the rails is the generalization to all sorts of modeling. If you think the model and it's assumptions are wrong why not go build a better model? That's the whole point of the scientific method after all.

selestify• 7 years ago

Because sometimes you don't know how to build a better model? You should be aware of even the best model's pitfalls.

logicchains• 7 years ago

If you think the model and it's assumptions are wrong why not go build a better model?

The better model may be intractable to fit (so complex that we lack enough data to train it). Or, it may be fittable but not easy to understand, e.g. a neural network model.

Fomite• 7 years ago

This. I work with models that are considerably simpler than "200 equations and 150 parameters" and we run into non-identifiability problems all the time.

yorak• 7 years ago

From how I see it the conflict is not technical but ideological. Or, in technical terms, what is the objective?: for Forrester and Collins this seemed to be increasing the economical growth of the city and managing KPI (crime, poverty etc.). Baker on the other hand considers also human dignity and our responsibility of keeping a fellow human fed, under a roof, and safe and does not ignore the externalities of the decisions.

shittyadmin9• 7 years ago

While I think I'm somewhat in agreement with this, this post seems to be making some claims without backing them up at all:

To Forrester, low-income housing was an especially egregious example of a “counterproductive” urban program. According to the model, these programs increased the local tax burden, attracted underemployed people into the city, and occupied land which might otherwise have been put to more economically healthy uses. Housing programs aimed at improving the condition of the underemployed, Forrester warned, “increased unemployment and reduced upward economic mobility” and would condemn the underemployed to lifelong poverty.

Where's the data to suggest this is false? In fact it seems to me like something that may hold true in data, even if deeper causes are at fault.

but we have decades of real-world evidence that demonstrates the disastrous costs of the “counterintuitive” anti-welfare agenda

... with no evidence provided?

Overall they actually seem like pretty reasonable theories at face value and calling them crap without evidence doesn't seem like a good solution if there's a problem with them.

ruytlm2• 7 years ago

Shouldn't the burden of proof should be on Forrester's model to prove that it is true?

As I understand it, this article is implying that Forrester's model incorporated those as base truths, without particularly rigourous evidence to support them - hence the comment about his lack of social science credentials. (Whether Forrester actually had evidence to support his model I do not know, as I have not dug deeper.)

shittyadmin9• 7 years ago

It's a bit like saying that math is wrong though - logically speaking, low-income housing has costs associated with it in more ways than one, not just housing but for law enforcement, education, etc, all of which will clearly have demand go up and income won't go up as much as if someone who had more money were to live in that city instead. Depressing, perhaps, but factual.

If that basic piece of math is wrong, it needs to be proven - how do low income households contribute more than their fair share of city revenue for example? Does homelessness cause law enforcement issues to increase to a greater amount than low-income housing? Do low-income households produce more future earners that add up better for the city?

We need some data here to show that the basic economics of this don't actually add up if that's the claim being made, because from a basic economics perspective the argument logically and mathematically follows.

supermw• 7 years ago

Demanding proof before continuing a discussion is a way to avoid discussions, not elevate people's understanding of a topic. Many people here do not have the same knowledge of the topic and may simply agree with the first theory that makes logical sense to them if no other data is presented.

Being that this is an internet forum, where people come to discuss things, there is no harm in preemptively presenting evidence to show why Forrester is wrong. There is nothing at stake here.

If he is wrong, it should be fairly trivial to find evidence showing why. I personally await this evidence because I find Forrester's model to be sound.

m_mueller2• 7 years ago

Well, he built a model that behaves this way. Whether that model is correct or not is another question, but the author should at least give insight as to what exactly in that model doesn't hold. The negative impact of social housing is an outcome of that model, as far as I understand.

Note: I'm pro social housing as I have anecdotal evidence for how well it works, however that's in a completely different system than the US, where urban planning puts high importance on mixing up low-, medium- and even high income areas to create opportunities for social mobility.

Lazare4• 7 years ago

Shouldn't the burden of proof should be on Forrester's model to prove that it is true?

That's not really how models work. A model is a prediction about how the world works; you can't prove them true. Even if my model of wheat harvests or global warming or city growth, or stock mark volatility works for the past 10 years in backtests, and worked this year, it doesn't mean it'll work next year.

It's like machine learning; you can't prove it's going to give you the right answers going forward; all you can do is see if it gave you the right answers so far, and hope for the best.

Kalium• 7 years ago

What I think you're pointing to is the rhetorical patterns employed by someone who believes they are preaching to the choir. There's no need to explain or justify why something is wrong. It just plainly is and your assertion will be taken on faith because it is aligned with what your audience already believes.

The final paragraph spells out, I think, what the subheading hinted at with "dangerously influential". This is ultimately a political polemic.

ruytlm2• 7 years ago

"Even though Forrester had no expertise in urban affairs—or in social science more generally—Collins agreed that a collaboration could prove fruitful."

This is one of the most interesting sentences to me. I have long been convinced that the efforts to reduce behaviour and models down to pure numbers, and disregard the human element - whether willfully or through ignorance or naivety), are more harmful than we think.

I think we are seeing similar things with how the attempt to 'personalise' services like news by crunching user behaviour into 'predictive' algorithms has led to the rise of echo chambers and fake news.

Or, I suppose you could consider the attempts to reduce bundles of derivatives to simple algorithms, that largely led to the 2008 financial crisis.

It seems a particularly human thing to assume we know far more about what's going on than we actually do.

Lazare4• 7 years ago

The article was interesting; I'd never heard of Forrester before.

What I found interesting though, is the fact that the article never seemed to get into the question of whether he was right. Is that because the author considered it too obvious to warrant discussion, or because he didn't think it was relevant?

The tone of the piece feels like it's intended as a takedown, but the only actual criticism I noticed was here:

we have decades of real-world evidence that demonstrates the disastrous costs of the “counterintuitive” anti-welfare agenda

But this came immediately after a section tracing this line of thinking back to some of the 20th centuries most respected thinkers, and after stating that this was now the default mode of analysis for policy wonks.

It reads like the author thinks they're administering the final blow after building up a damning case, and yet...

Jtsummers2• 7 years ago

I agree. I think the real conclusions ought to be (non-exhaustive):

Keep the model's assumptions in mind when making decisions based on it.

Always remember that it's a model, and consequently always wrong (as measured against reality).

All models have to be fine-tuned over time, especially dynamic models (where the system being modeled is itself subject to change in structure, consequently invalidating the model).

=====

If you emphasize that the model is just a model, and its assumptions (both for excluded factors and included ones), then you can make reasonable decisions guided by the model. If you fail to do that, you'll start to believe the reductionist model and make flawed decisions.

Lazare4• 7 years ago

Yes. Models are basically programs, and GIGO absolutely applies: Garbage In = Garbage Out. :)

hoaw2• 7 years ago

Which is essentially the whole argument of the article. For example as summarized in the last paragraph:

"Expert knowledge, of course, has an important place in democratic deliberation, but it can also cut people out of the policy process, dampen the urgency of moral claims, and program a sense of powerlessness into our public discourse."

danbolt• 7 years ago

My friends and coworkers often speak a lot of praises about the survival/industrial game Factorio. In the game, pollution from the player’s generators attract hostile creatures that one has to manage. The player can use renewable energy an option, but it’s not as easily productive or scalable as other sources of power.

I haven’t played the game, but I always thought of the pollution/conflict dynamic and renewables being more safe (but harder to get right) as a statement on environmental policy.

aero1422• 7 years ago

I think the actual game is much harsher than this. You play as the main character and the game is really just about expanding the factory for the sake of expanding the factory. It serves no purpose other than making the factory bigger. You might be able to produce a little less pollution with solar but at the end, the game is about expanding the factory. You will cut down the forests, and pollute the environment, and pave over the beautiful landscape and the remove the homes of the natives that stand in your way. Why? Because it's what you do. It's factorio. It's slightly subtle, but not really. I think everyone realizes this is the backdrop of the game but you just keep building, and paving and polluting, because it's Factorio. It feels like it should be some moral lesson, I'm not sure what it is though. You just keep doing stuff and having fun.

joshu• 7 years ago

er, there is an endgame. the rocket.

gizmo686• 7 years ago

I'm a big fan of Factorio, but the pollution mechanic isn't well done. Its mostly an early game mechanic to encourage you to go out an massacre the natives that are inside the polluted areas. With fairly hard difficulty settings, you might start to scale back production in the early game to limit your pollution, but that's about it.

The real benefit of solar in the game is frame rate.

Also, solar is the easiest energy source to get right. Just add solar panels if you brown out during the day, and accumulators if you black out at night (or lookup/compute the solar/accumulator ratio; and lets face it, if you are playing Factorio, you probably already did). Burning fuel for power has all of the logistics problems of keeping your fuel supply saturated (plus boiler/generator/pump ratios). And nuclear can get as complicated as you care to make it.

m_mueller2• 7 years ago

I did a build where I really saw the positive effect of solar. I teched up to electric furnaces and replaced all of the coal driven ones with electric, including efficiency modules, powered by solar. The steam engines became used only for brown-out protection. As soon as I did that the biter attacks became much more manageable and I could easily wall in with hardly any resistence. This was on death world.

So I wouldn't say the mechanic doesn't work, for me it made a lot of sense.

jonathankoren• 7 years ago

Adam Curtis talked about Forester and the study of “cybernetic systems” (to use the contemporary term) in general, in episode 2 of “All Watched Over By the Machines Of Loving Grace” [0].

The core argument (which is alluded to in the Logic Magazine article) is that in the 1960s when computers were beginning to be integrated into companies and governments, there was a strong desire to have the computer solve all the world’s problems. Of course, the problems were not well understood (and still often aren’t), while at the same time the computers were woefully underdeveloped. Therefore simplifications were made so that the computer could calculate something that looked somewhat reasonable. Of course the models were symplistic, and the data encoding symplistic and biased as well. But hey, the computer “solved” the problem, and as we all know, “Computers don’t make mistakes.” Eventually (and Forrester encouraged) the simplistic world was confused for an actual complete description of the real one. Most damningly, industrial leaders and politicians, started trying to make people and society fit the simplistic model, instead of improving the models to better fit society. Why? Well, there was a computer model made by Very Smart Men(tm), and it gave cover for what was already decided, so it must be right.

It’s as if the proverbial dairy farmer that received the report that begins, “Suppose you have a spherical cow...”, started selective breeding and cutting his cows to optimize for sphericality, as opposed to milk production, because the report says “spherical cows.”

It’s a similar pattern we’re seeing play out with ML, where the assumption is that the classifier is correct regardless of its actual performance.

[0] https://archive.org/details/AllWatchedOverByMachinesOfLoving...

pasta• 7 years ago

At the moment I am playing 'Cities: skylines' which it one of the best city cims at the moment imho.

But the real question is: does it simulate what we know or what works best?

Ofcourse I think it simulates what we know which might not be the best for people.

When you read the works of Christopher Alexander you will get the feeling that we need to redesign our areas. And this is not what a sim simulates..

EricE• 7 years ago

The vast majority of social programs when played out to their natural conclusions do make problems far worse.

Just look at the mayhem easy to obtain government backed student loans have caused in "higher" education - skyrocketing tuition and a bachelors degree being worth about the same as a high school diploma was in the 80's. But now with crippling debt for new graduates.

Yeah! But we are improving things!

so33• 7 years ago

Just look at the mayhem easy to obtain government backed student loans have caused in "higher" education - skyrocketing tuition and a bachelors degree being worth about the same as a high school diploma was in the 80's. But now with crippling debt for new graduates.

This is a problem that, as far as I can tell, exists mostly in America. I doubt this is a problem in European countries where tuition is mostly free or extremely low cost to begin with.

Maybe the key for effective social programs is not to apply a bandaid but to fix the root cause.

james_s_tayler• 7 years ago

I'm sure there were some great benefits too.

The problem with anything is when you choose something you choose it's benefits and it's detriments. The other problem is that everything has benefits and detriments. So, no matter what you set the knobs, dials, switches and sliders to you always make things simultaneously better and worse. Sadly there isn't a solution to this problem, instead we just politic about things in and endless tug of war so that we are occasionally not on the losing end.

int_19h2• 7 years ago

As one moved down the class ladder in the urban dynamics model, classist assumptions about the urban poor piled up: birth rates were higher

Why is this controversial, or, for that matter, an assumption? There's plenty of studies demonstrating correlation between income and fertility... or are they implying that the correlation is only meaningful nation-wide, not within any particular city?

Lazare4• 7 years ago

It was puzzling. I suppose the argument isn't that the assumption was factually wrong, but morally wrong?

hoaw2• 7 years ago

No, it is arguing that it is factually wrong based on prejudice. Most obviously that the groups negative factors aren't inherent and their contributions aren't valued.

colinb• 7 years ago

Does anyone know if there's a strong link between Forrester and Donella Meadows, of Thinking in Systems? I'm reading that book after seeing it praised more than once on HN.

Meadows provides lots of food for thought, and I think I could make my workplace better if I figured out how to apply some of the ideas she expresses. I'm not convinced that all of her reasoning is very rigorous, and I suppose that matches the feelings of the author of this article, not that I think he does a great job of rigor either. Once the fine article extended its ideas from city planning to no platforming, without apparent irony, or data, I got the yawn.

Jtsummers2• 7 years ago

She worked with Forrester’s group in the late 60s/early 70s. So there’s a direct connection.

thanatropism• 7 years ago

Literally yes. They're coauthors and appear to have collaborated extensively.

acoye• 7 years ago

LGR just released a video retrospective on the subject, 30 years later. You can find it here : https://www.youtube.com/watch?v=TrScy1icWjI

AbrahamParangi• 7 years ago

“Using this rhetorical tactic, the conservative speaker can claim that they share your social goal, but simultaneously argue that the means you are using to achieve it will only make matters worse.”

I think this line in particular really demonstrates the resistance to ideological debate on the left (to say nothing of resistance on the right which I think has a different character). It’s like, these ideas that aren’t already part of your ideology are discredited because you assume they come from a place of bad-faith.

If you refuse to engage with someone’s ideas (who supposedly wants to solve the same problems you do!) because you’ve already assumed they’re evil- that seems like a pretty dangerous mindset to me.

mwkaufma• 7 years ago

For more on Jay Forrester's philosophy and influence on politics, internationally, check out the All Watched Over by Machines of Loving Grace documentary series by Adam Curtis.

Model Metropolis

Related Stories