Slashdot Mirror
IBM Plays SimCity With Portland, Oregon
Hugh Pickens writes "Portland, Oregon will be the first city to use IBM's new software called Systems Dynamics for Smarter Cities, containing 3,000 equations which collectively seek to model cities' emergent behavior and help them figure out how policy can affect the lives of their citizens. The program seeks to quantify the cause-and-effect relationships between seemingly uncorrelated urban phenomena. 'What's the connection, for example, between ... obesity rates and carbon emissions?' writes Greg Lindsay. 'To find out, simply round up experts to hash out the linkages, translate them into algorithms, and upload enough historical data to populate the model. Then turn the knobs to see what happens when you nudge the city in one direction.' One of the drivers of the 'Portland Plan' is the city's commitment to a 40 percent decrease in carbon emissions by 2030, which necessitates less driving and more walking and biking. After running the model, planners discovered a positive feedback loop: More walking and biking would lead to lower obesity rates for Portlanders. In turn, a fitter population would find walking and biking a more attractive option. But as the field of urban systems gathers steam, it's important to remember that IBM and its fellow technology companies aren't the first to offer a quantitative toolkit to cities. In the 1970s, RAND built models they thought could predict fire patterns in New York, and then used them to justify closing fire stations in NYC's poorest sections in the name of efficiency, a decision that would ultimately displace 600,000 people as their neighborhoods burned."
Tear up all the roads. Replace with rail.
Just make sure they disable disasters before they play. An alien monster destroying the power plant wouldn't be nice.
I can't help myself, "By using the software, Portland confirmed its plan to reduce carbon emissions 40 percent by 2030 would on the whole be a positive outcome." is IMO as news worthy as "By tasking a PwC opportunity analysis, Portland confirmed its plan to reduce carbon emissions 40 percent by 2030 would on the whole be a positive outcome."
What makes IBM's modelling so special?
a) build a slum with nearly zero taxes
b) Build acrologies with enough police stations around
Computer models don't tell you anything you didn't already know. If you are really lucky, they will tell you things you didn't kow you knew. The problem with this sort of model is that if one of your starting assumptions is wrong, all of your conclusions from the model will be wrong.
The truth is that all men having power ought to be mistrusted. James Madison
You can add pirates to the sim, and it would still look like the real thing (no, seriously, you can). No word on ninjas, though.
It's cool and all that IBM thought to do a sim of us out here in Stumptown, but I mean, we're not exactly going to be one of them thar model cities that will replicate easily to other towns.
I mean, hell, couldn't IBM choose something easier to do, like, oh, Des Moines or something?
Now to be fair to the fine folks in Iowa, they do have the Carp Festival, but seriously? IBM would have a *much* easier time there than here. Just a hint, fellas.
Quo usque tandem abutere, Nimbus, patientia nostra?
In the 1970s, RAND built models they thought could predict fire patterns in New York, and then used them to justify closing fire stations in NYC's poorest sections in the name of efficiency, a decision that would ultimately displace 600,000 people as their neighborhoods burned.
So the source is a wikipedia page, which cites this book, which is a dead end for now.
Are the authors talking about this study?
If anyone's got a source that actually backs up the notion that RAND explicitly recommended closing down fire stations in poor areas, or the actual claims that "they're just committing arson anyway", I'm very curious, as that's a pretty wild claim. I've emailed them for comment.
Now, maybe I'm just being paranoid here, but that sounds too good to be true. Nobody closes fire stations just because software says they can. But people do wrangle a study until it gives the results they like, and then use it as justification for the actions they were going to take anyway.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
A few tweaks here and there, throw out that historical data (it's obviously flawed), and tweak this historical data (it's flawed but we know how to "fix" it) and I can make the model "prove" whatever you want. Now you can justify your vote with "science".
I doubt Portland will do anything like have the models predict outcome of projects for the next 10 years, then if they show success use it for the following 10 years. Instead they'll start spending now because "science" says it's ok.
I hope they didnt spend much money for a report that says if people walk and ride bike more the obesity rate will go down and subsequently people will enjoy going out to walk/ride bike more. I think anyone who calls themselves a personal trainer could have figured that out for a lot less money.
Rename the town "DebtCeilingAA+" to activate the infinite money cheat.
Surely they must have discovered something that's both relevant and doesn't cause people to groan at how obvious it is?
Anyway, these kinds of models suffer from a fundamental problem that can never be solved. While a lot of the steady-state behaviors of a city may be amenable to being reduced to evolutionary (in the PDE sense) equations - things like traffic patterns or crime distribution - the driving forces that change and disrupt things on human timescales aren't continuous.
Try all you want to reduce society's evolution to a set of coupled ODEs or PDEs, but you'd never predict the occurence or path of the Civil Rights Movement because it was initiated by one single person's decision. Sure you can have a decent shot at modelling the probability of a civil rights movement occurring in a segregationist society in a given timespan, in the same way weather models tell you the year-averaged expectation values for things like cloud cover, insolation and rainfall... but whether it actually rains on any given day? Whether one little black lady will refuse to give up her seat? Forget it.
When it comes to society, not only are we talking about chaotic systems (like the weather) but now the very events that seed the chaos occur discretely and randomly.
Provide high speed network to the entire city. More people will stay home because they can work and play from home. Then there will be fewer people on the roads, buses, trains, and in the parks and libraries. Slash the budget for those services like you plan to any way.
Having to work for a living is the root of all evil.
The City of Portland, Oregon will be extremely disappointed when they find out that holding down the shift key and typing "fund" doesn't work like it did in SimCity Classic.
Shameless plug for my photos on Flickr
Remember the hex edit? ...for $2,138,000,000.
At offset 39(hex 27), change: 00 02 42 20 to: 7F 02 FF FF...
Original simcity?
Yes, I always found the reduced maintenance cost of roads, outweighed the negative effects of traffic and pollution.
Rail only allowed greater land values, and therefore tax income.
Does anyone have a problem with governments spending money trying to model ways to "nudge" our personal behavior? I'm all for sound city planning, etc... But this seems to dive headlong in to limiting freedoms. I'm not comfortable with any government getting this up close and personal with me. I already have a wife that nags me about exercise...
So, about the RAND modelling mentioned in TFS... Is there any evidence that their model failed to produce the desired outcome(as opposed to merely being deeply callous and perhaps a bit tactless, two RAND traits that anybody familiar with their game-theory work during the Cold War should hardly be surprised by...)?
Apparently, the "planned shrinkage" policies were part of a broader 'Urban Renewal'/cost reduction planning strategy by the city of New York. "Stop providing police or fire service to the slums and let the poor bastards burn themselves out so that developers can have a crack at the area" certainly isn't a very pleasant notion, and I can understand why one would want a nice, shiny, 'objective', analysis justifying it in gentle and scientific terms; but it isn't as though those poor country bumpkins at the mayor's office were snookered into the idea by RAND's slick big-city gents...
If anything, that particular scenario is not so much a demonstration of how models can fuck up; but how models can end up being constructed to to provide the desired answer, something at which the more customer-service oriented modelers are very good.(The situation is arguably analogous to the absurd and self-serving 'blame those damn quants for the financial meltdown!' narrative that started circulating after the credit-default-swaps really started going south... Yeah, of course a tightly linked set of casinos where people play with other people's money fell apart because a few CS weenies made some math mistakes...)
and help them figure out how policy can affect the lives of their citizens.
You mean that until now, the people who are paid to take decisions for us have absolutely no idea of the potential outcomes of these decisions? That would explain a lot.
Seven puppies were harmed during the making of this post.
More walking and biking would lead to lower obesity rates for Portlanders. In turn, a fitter population would find walking and biking a more attractive option.
Would a fitter population find walking and biking a more attractive option, or would they find it a less unattractive option? These sound like the same thing, but think of it this way. Two scenarios where you are out looking for a sex partner.
Scenario One: You find yourself in a position where you have a choice of one of two partners. It is pretty clear that either one will go home with you. Do you choose the one that is more attractive or the one that is less attractive?
Scenario Two: You find yourself in a position where you have a choice of one of two partners. It is pretty clear that either one will go home with you. Do you choose the one that is less repulsive or the one that is more repulsive?
In Scenario One, you will probably choose the one that is more attractive. In Scenario Two, a large percentage of the population will choose to go home alone.
The truth is that all men having power ought to be mistrusted. James Madison
Overall I think this is a positive way of trying to improve a complex reality. There was an experiment in 1990 of an imaginary town called Tanaland, and most people failed miserably in improving the long term life conditions for its inhabitants. From http://tersesystems.com/2011/06/10/the-logic-of-failure:
When you are sure of something, you probably are wrong (search for "Unskilled and Unaware of It").
The ability to walk to work requires a job within walking distance, which almost never exists. In this, reality seems to follow the SimCity algorithm of job creation - jobs in another city are always better than the ones in yours. I would usually build a rail connection between them all and whenever you switch cities you'd always see an increase in rail traffic coming into the city you're playing. SimCity then tries to give all those commuters jobs in your city. Since there are none, it looks for jobs in the neighboring areas and routes the rail traffic there. After playing all four adjacent cities it is easy to create a couple of hundred thousand commuters just riding around all four areas in a giant loop. They don't get onto the train in any city, they don't get off the train in any city. They just live on it, riding round and round and round.
Meanwhile, IBM continues to locate all of its sites exactly as far away from public transportation as possible. The Toronto site, 10 years ago, went so far as to move *further* away from the downtown area.
In turn, a fitter population would find start heckling the fatties, who in turn would commit suicide or die a lonely death. The surviving fit people are statistically relevant, and deaths do not factor into our scenario.
A planned economy requires massive quantities of mathematics to be performed.
And IBM was there to help. The Soviet Union was one of the biggest customers of IBM equipment, all through the 20s, the 30s, the Holodomor, the purges, etc.
So what does this "nudge" involve? There are significant sections of the population without the ability to walk or bike for long distances (think the elderly, many medical conditions, late term pregnancy, young children and so on) who would all now presumably be second class citizens?
Speaking as someone who was fit, active and a regular gym goer right up until a year ago, I would have applauded the plan and scoffed at anyone who criticised it.
But then a year ago I suffered from a neck problem that has left me with ongoing nerve and muscle related symtoms that make it impossible to even stand for long periods of time. It's been tough, and has completely devastated every aspect of my life. I've lost a lot of
In the words of Click & Clack: BooooOOOOooo-GUS!
I love how proponents of anti-vehicular living conveniently forget that there are a huge number of tasks that can't be performed on a bicycle or on foot. In addition, studies like this fail to point out just how far people are willing to walk or that inclement weather isn't conducive to walking or biking. Ultimately, what these sort of studies point to is that huge masses of people need to live in giant apartment complexes built within a fairly small radius. I'd say 1-2 miles max.
In addition, people are people and people want what they want. Given a choice I'd say most people would want a plot of land and a house versus being packed into an apartment lick cattle. Sure, lots of people live in big cities but what percentage of them actually enjoy it versus wanting to get the hell out but not having the money to be able to do so. And then there are lots of people who believe they are special and are entitled to spacious living and driving around everywhere. They refuse to associate with "the great unwashed". How many elites do you see riding the bus?
It took a computer model capable of modeling 3,000 equations to tell Portland that more walking and biking would reduce obesity? How is this not common sense? Are there people who don't think getting more exercise will reduce obesity? Is this a symptom of our society that we need to wait for a scientist to tell us with mathematical formulas and experimental studies that heat burns things and light makes it possible to see things?
The thing is, people like popular social 'zones' with say 60 eateries in one street and a dozen pubs in the same street.
The planners don't like this at all, nor do landlords or developers who want to share and disperse the 'hot zone'. The clumping of like businesses leads to public transport efficiencies.
They pretty much had this 15 years ago with SimCity 2000, when I was in skill. There was a competition to build the the best city with various variables taking into account. There were very strict rules on what you could build, such as none of the futuristic crap like fusion power plants. After you had your city you have a build a model layout of a section of it. This was at least 15 years ago, and using a child's computer game. Why has it taken so long to develop something that obviously wouldn't have taken that much thought since it pretty much existed?
Just because you are wrong and I called you out on it doesn't mean I am a Troll.
Oh dear gawd, listen up everyone, I'm sure I'll get weird moderations for this, but here goes.
This is the successor to Facebook if they do it right!
It combines everyone's favorite Watcher mentality with everyone's favorite 90's game!
Everyone loads themselves in, some people are "helped" etc. You can watch the entire town buzzing merrily along! Click on people! Their phone sends them a hello text! Click on stores! See what they have in stock. Click on the DMV. Check the lines.
The possibilities are both endless and endlessly terrifying! Whee!
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
You can't cut back on funding! You will live to regret this!!
A forklift can tow water pumps, tanks, and other equipment without any trouble.
Ever drove one? I have, for work, and for years (when I was working full-time while getting my BS degree in CS.) They are great for lifting things up and down, but you have to drive them in reverse for safety (unless you drive the small types), they are designed for carrying things on their forks, not for hauling (very important distinction, just look at their wheels for Christ's sake), and I believe (I could be wrong in this) they are not designed for efficient traveling over the same distances you would expect a firefighter truck to do (not to mention they cannot sustain the same type of speed a truck would.) Granted, my experience with forklifts and similar equipment has been limited to their typical usage in warehouses (lifting heavy shit and move it from a warehouse/container to/from a transport/another container.) But based on my work experience alone, I do not believe what you are proposing is viable.
So this model doesn't take into account the Mule (most models don't). You can edit for different circumstances, but it's definitely a "if this were to happen, then this might happen" thing rather than a "if we do this, then this might happen". You can't actual control obesity levels any more than you can control how many roads are in a city.
Scientist: If we replace roads with walking areas and light rail, we can reduce pollution in the city by 80%, make travel more efficient, and have 30% more green space.
Politicians: We will NOT be the first city to remove cars. We get way too many kickbacks. Find something else to do. Can we have light rail that runs on gasoline? Would that use more gas than all the cars? If so, we'll put that on the table.
Instead of paying IBM for software to play god, Portland could fix all their roads so cars aren't sitting idling in multi-hour traffic jams. That would easily cut 40% of emissions. Walking/cycling are also great ideas, but you have to construct dedicated biking paths and get them out of the roadways - it's the only way that can work properly. Take up that task at the same time you completely revamp the roads, and then you'll actually have something to crow about.
Do you think they'll actually do any of this? When pigs sprout wings and begin to fly. I know it's a pipe dream. In reality, they won't re-engineer all the roads and their SimCity experiment won't work either. By 2030 there will be an increase in emissions of 40%, and people will be sitting in traffic even longer.
There are many benefits to walking and biking but losing weight is not one of them. This is covered in good detail, citing latest medical research, in "Good Calories, Bad Calories" by Gary Taubes. The reason is straight forward while fatty acids move from your fat tissue to your muscles when you exercise, those same fatty acids return to your fat tissue when you rest. If you add that to the effect of an increased appetite that you are compelled to act upon after you exercise (aka "working up an appetite") the net is that you don't lose weight by exercise. Changing your diet, on the other hand, is very effective. There is obviously a lot more to this, but this is the rationale in a nutshell.
Ever been to Portland? I am a traveling consultant, so see a lot of places. Portland already blows away every other city in the country in terms of number of people who walk and ride bicycles. I have never seen another windmill on top of a skyscrapers anywhere else either, and there are electric car recharging stations scattered around town too. Seems to me like they have it pretty well figured out on their own without any simulations needed.
"People don't like to be meddled with. We tell them what to do, what to think, don't run, don't walk. We're in their homes and in their heads and we haven't the right. We're meddlesome."
-- River Tam
Go ahead, turn some knobs. See how it affects the lives of complete strangers.
~Dalcius
Rome wasn't burnt in a day.
From TFA: "Westâ(TM)s conclusions are only as good as the data and the models (patents equal innovation) he has to work with."
/.ers, move along, nothing to see here. Thank your $DIETY you don't live in Portland.
"patents equal innovation" == epic failure of model.
Ok,
Only a moron believes human behavior can be modeled with 3,000 equations. Or 30,000 equations. Or 300,000 equations. Or, well, you get the picture.
take a bunch of experts
turn what they say into rules
guess the coefficients/interaction levels
multiply it all together
the result is no more than a bunch of wild guesses with enormous error margins.
It's hard to model systems even when the basic mechanisms are fairly well understood (e.g. weather, planetary warming). A giant social system where the mechanisms are not really understood at all isn't going to yield to IBM's supercomputer.
VLC Remote for iPhone and Android
'which necessitates less driving and more walking and biking'....
bad news IBM, 30% of americans are too fat to put on their shoes without breaking a sweat. Unless you plan to radically redesign the bacon-double-cheeseburger in the process i contest that 40% is a rather ambitious goal.
Good people go to bed earlier.
> 'To find out, simply ROUND UP THE EXPERTS to hash out the linkages, translate them into algorithms, and upload enough historical data to populate the model. Then turn the knobs to see what happens when you nudge the city in one direction.'
(emphasis mine)
Here's your problem!
- For the complete works of Shakespeare: cat
You keep the taxes at 0% until the month before they are paid then crank it up to 20%. After they are paid, drop them back down to 0. It would appear that Sims have the long term memory of your average teabagger.
The basic description of this software sounds remarkably like the indie game 'Democracy', a government-sim that is well worth trying. It's on PC Mac and I know some people on the devs forums have said it runs fine under WINE, but I'm on PC so cannot vouch for that :D
Anyway, it's worth a look if you find the idea of this stuff interesting:
http://www.positech.co.uk/democracy2
> In the 1970s, RAND built models they thought could predict fire patterns in New York, and then used them to justify closing fire stations in NYC's poorest sections in the name of efficiency, a decision that would ultimately displace 600,000 people as their neighborhoods burned."
And so... it worked.
Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
"After running the model, planners discovered a positive feedback loop: More walking and biking would lead to lower obesity rates for Portlanders" .. o rly. No wonder their so fat if this is news to them.
... ...
After running the model, planners discovered a positive feedback loop: More walking and biking would lead to lower obesity rates for Portlanders.
They need a computer simulation to figure this out? Fire the assholes, hire people with decent IQ, save big money and also be fit.
I live in Portland and I'm an avid walker and biker. I mean, I was, until I moved to Portland. Did one of these 3000 equations consider that it rains here 8 months out of the year?
It's actually quite simple: You need local infrastructure for almost everything people do. Live, work, shop, schools, everything. In old villages that grew before modern transportation showed up, you sometimes still have this. You need to drive long distances much more rarely then (even if you can, if you have to) and walking or cycling suddenly becomes not only an option but a natural choice.
The trick is not making transport faster or more efficient, but to make it unneccessary as often as possible. Of course this happens not by itself and is hard to plan for.
One important thing is that many people have learned to love driving their cars for thousands of reasons. It's often their only basically "free" time with nothing else to do but driving, listening to music or just dreaming. It's also a way to get far away from somewhere you don't like to be (often to somewhere else you don't like to be for different reasons, but, hey). You'd need to supply many things in the local environment (including the work environment) to make up for that.
3000 equations is a LOT, so this is surely unnecessarily complicated, and thus very unlikely to work.
If just one expert gives a wrong part, then the whole model can become faulty.
And there appear to be no testing of this model.
This is surely going to give even more wrong results than the climate models and nutrition models.
Great. Hopefully this will be as successful as when IBM provided the Germans with the computers they used to plan and manage the Holocaust.