Geoffrey West: The surprising math of cities and corporations
HTML-код
- Опубликовано: 25 июл 2011
- www.ted.com Physicist Geoffrey West has found that simple, mathematical laws govern the properties of cities -- that wealth, crime rate, walking speed and many other aspects of a city can be deduced from a single number: the city's population. In this mind-bending talk from TEDGlobal he shows how it works and how similar laws hold for organisms and corporations.
TEDTalks is a daily video podcast of the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes. Featured speakers have included Al Gore on climate change, Philippe Starck on design, Jill Bolte Taylor on observing her own stroke, Nicholas Negroponte on One Laptop per Child, Jane Goodall on chimpanzees, Bill Gates on malaria and mosquitoes, Pattie Maes on the "Sixth Sense" wearable tech, and "Lost" producer JJ Abrams on the allure of mystery. TED stands for Technology, Entertainment, Design, and TEDTalks cover these topics as well as science, business, development and the arts. Closed captions and translated subtitles in a variety of languages are now available on TED.com, at www.ted.com/translate. 
Наука
I think my favorite idea out of this talk was his observation that societies must accelerate their speed of innovation to avoid impending collapse. That is a scary truth but it is not do or die.
That is, it seems possible for societies to reach sufficient happiness, wealth, and population size before the crash, but only if they are able to innovate faster than the challenges they face. If they are not able, then they will either never reach optimality; rarely advancing beyond a point, or they may die away altogether.
This IS a scientific talk. I do NOT consider the underlying math to be simple. This is (advanced) statistical physics, and that is the context in which you should see this talk/video if you DO have a scientific background. There are several good scientific messages in this video, one of which being that VERY complex simulations of huge and interconnected systems CAN yield very simple properties. Kind of like how a random walk performed by N_A particles will always give you predictable diffusion.
One of the best talks on TED ever.
1:40， I saw Yang Lan among the audience!
Great TED talk! Fascinating stuff. I came hear after seeing West on Through the Wormhole.
before watching this vid, just from the title i feel that i'm going to love this talk :)
Excellent talk Mr. West.
this is excellent work!!!
Biologists have 'naturally' separated quite nicely into groups that deal with organism stuff and community / ecosystem stuff. Because these things do not function similarly. And though we've studied natural systems for longer than anthropocentric ones, we've found no real dependencies between these levels. So there's as little point in mentioning company growth rate when talking in the context of cities as there is in mentioning animal growth rate in the context of ecosystem level effects.
This is soooo awesome! I'm a recent graduate in physics and I sooooo want to study this topic :o
Karen Arias have you started studying it? I'm a recently graduate in physics and am also in love with this research area!
I would suggest also looking into theory of computation, information theory and economy. Grasping it would be of great help in pursuing research of complex systems.
this was true back then, but that was before the new liberals tried socialism
Geofrey west wrote a book called 'Scale'. This talk is based on that book.
sante fe institute, girl!
Some interresting points here, but he doesn't say much about how big variation there is in various things, just that the general concept is universal. I'd like to know what the standard deviation from this universality is, and if any examples violate it. Also if not, how big is the largest deviations from the predictions?
This reminds me of the charter city talk. If done well, that could perhaps change the statistics compared to traditional cities, especially if one sets limits for them.
Stats, manipulation of stats, and graphic representation of stats is a forté for me. One of the very few. This man is accurate enough and the presentation is easily understood.
13:50 those graphs are a bit misleading. One's x-axis increases at a rate of .5 while the other increases at a rate of .2. I think it would be better to have these scaled the same.
I'm still waiting for my company to start looking like a hockey stick.
I'll watch how things develop from my little shrew of a town, knowing I'm probably not very bright or I would have been drawn to the city to network with all the other superinventive people.
Either that or I'll take comfort knowing that the next innovation is right under our noses, this ability to network across all the lines via RUclips, etc until we realize our planet is just a bronchial tube in the greater network that is the universe.
The connections I've made with people in YT gives me hope
how find this math formul?
This is an interesting talk, and there's a lot of valuable information in here, but I also fear some of the arguments are too simplistic, making the danger of increased externalization apparent. The indicators tested (such as income) are not necessarily primary goals. Also, the story seems to go counter to Joseph Tainter's more long-term and more comprehensive arguments about diminishing marginal returns, which govern systems such as cities as well and have profound effects in the long run.
At 3:05 I could think of only one thing: this guy would make an excelent action figure.
15% rule is 'universal'? Would have been interesting to compare cities behind the Iron Curtain and those in Western Europe. Once the Wall came down, Eastern European cities rapidly caught up with that universal. But kept in the aspic of central planning and state-controlled micro-management?
@Phelan666 How can the graphs be misleading if they provide all the information necessary to make that claim? We use graphs to find patterns. Graphs are useful because we can adjust the X/Y axis ratio to make patterns more observable. In this case, two examples were used to illustrate the inverse relationship of sub-linear biological network scaling and super-linear social network scaling. In controlling the linear variable, the X/Y were adjusted to make his point more observable.
@dingdongpingpong69 Yes and that ratio could get even get larger, note his chart is log scale. In the super exponential chart where as in such systems as you go down the X-axis not only do you get more Y, but the ratio of Y/X increases exponentially. Obviously as he pointed out in such systems, collapse is inevitable, you can only increase Y bounded by reality ( if Y was gas stations you can't have more gas stations than all the space available on earth etc..)
@fionn32 Nice to see that you've given this lecture the rigorous intellectual pondering it deserves.
@marcdaddy33 Observational studies are valid in scientific analysis; in many cases, it's the only way to go about it. In the case of evolution, you can watch the proportions of certain characteristics of a species change over time in response environmental pressures. One example that happens over a relatively short timescale (due to how they transfer genes) is the bacteria that are resistant to antibiotics. There are lots of classical examples too, so if you're interested, you can google it.
"The surprising lack of math in a video about cities and corporations" summarized:
In biology:
As size increases, metabolism decreases predictably.
This is not the case for cities. (Probably because this is a measure of an increasing collection of one species, and not a comparison of average size vs metabolism of differing species as in the biology example)
So, If we want to continue growth and avoid collapse,we must be more innovative.
And then a graph entitled: Revenue With Inflation Deflator
To answer his question about can we innovate faster and faster, YES! Predictive Innovation allows you to predict innovations for anything. So instead of reacting we can plan innovations. Using his math to predict when a big innovations are required can provide time lines for planning.
@balisticsquirel How is this not science talk ?? Please expand on that !
Explain New York and Baltimore's crime's rate ?
So happy, yet another great talk! Hate to complain but that intro BOOM should be on the way out... I'm not the first to say it...could still be the TED trademark or whatever at 60% volume, just saying...
@lukethegreat101 Which is, however, not that easy when it comes to highly empirical topics like this. As with statistics, you can find logical arguments for more or less any theory regarding sociology, economy and so on. But on the other hand, you've obviously got a point there.
@balisticsquirel The more i think about my post though, the more i'm haunted by thoughts from when i was learning the theory relevant to individuals and to ecosystems - that nobody ever has really tested for a dependency between levels. Imagine for instance that he were standing up saying "what we've found is a predictive relationship between the type of company making up a city and the type of city (characterised by the relationship between population and whatever)." THAT would be useful.
I wonder if he ever displayed every set of data he had in a bell curve. My friends statistics professor says that data for any characteristic of any thing in nature or economics is a normal bell curve
What if you slow down the growth?
Awesome.
@dingdongpingpong69 The growth he's talking about is the line (relationship) as it moves to the right of the chart. So if the X-axis is population size and Y-axis number of gas stations. When you have sub-linear growth and if that trends continues it means as you get bigger you don't proportionality increase in the number of gas stations.
Projecting forward it seems eventually you'll reach a population size where you will get almost no increase in gas stations even if u double the population..
Blew my mind! 🤯
@marcdaddy33 Then why are you bothering to watch TED videos?
What European cities have people walking backwards? I think I misunderstood that graph.
@CHUCKYCHUCKYBOBUCY I understand what you say and of course, you have a point. Let me expand on the original comment. Money in itself is of course very important, but it's a sub-indicator. It's rarely the end goal in itself, but always a vessel to a spectrum of other aspects ( such as status, comfort, satisfaction, independence, etc.) which can also be achieved in other ways. Therefore, money is not a prime goal in and of itself. Evaluating using a sub-indicator can lead to sub-optimizations.
All the comments below keep 'fanboying/fangirling' over this talk, while conveniently ignoring one of the most salient and worrying points, that is the need for ever faster cycles of innovation, so that we dun reach our carrying capacity. So can we start discussing this pressing problem now ??? How on earth do we cope with ever increasing amounts of waste (domestic etc.) and demands for resources as cities get bigger ?!
Wei Shen You're asking the RUclips comments?
@N3rdyDav3
most of google's revenue is from internet advertising. considering the current "recovery" of the economy, it could be that google won't have many clients and will shrink significantly.
@guyboy625 You are right to question. I made an assertion and the onus is on me to back it up. Science is a method before it's anything else, and this is (basically) the method. But the speaker doesn't do this. He offers only implied things to support an assertion. Actually, he never even really makes an assertion (except again by implication), so we can't even get to the stage of testing it. He uses the tools of a scientist (data and graphs), but not the core scientific method.
@eagleeye1975
I LIKE your comment!
Don't you think though that the biggest 'issue' with the talk is that he doesn't actually offer any conclusions. (possibly because he can't because as you point out he has made no basis from which to conclude anything). And those that he sneakily infers, let's call them .. 'speculative', are presented to the public as 'science' cause he's a physicist and cause he's got graphs right.
I live in the city. It's pretty awesome.
@dingdongpingpong69 Several things, as noted in the talk corporations scale sub-linearly so they are fundamentally limited in some fashion and their growth will taper off not matter what but cities scale super-linearly. Now why is that? I suspect its due to the free social order vs fixed hierarchical ordering of the two.
What this means is, larger cities produce more wealth and less waste. So as populations grows, more big cities, more wealth and less waste.. So Malthus was wrong.
Yes, the dangers of externalization need to be more apparent, and in my opinion the ideas proposed in this video are too simplistic to do that.
As far as Tainter's argument goes, I believe he has the whole of human history to support his hypothesis, and I don't think there exists any fundamental criticism on it, so I'm not sure why, as a human being on this planet hoping to survive, one would want to disagree or even divert from Tainter as a base component in arguing societal dynamics.
Government officials and policy makers worldwide need to apply this knowledge to innovate a different way. So far, everything they have done, in the name of expediency, accelerates us towards undesirable outcomes, and ensure the demise of the human race. Is a huge systematic shock / reality check / wake-up call what is needed? Are we able to overcome that eventuality? Given a free choice, I would pick the path to "accelerated empowerment" over "wealth destruction" any day of the week.
Has anyone seen that graph; if we were to put every person alive on the planet today into one city, and set the scale to the average amount of population per area relative to other cities, then we would need a city of a certain size. Well that graph says that if we did that relative to the PpA of Paris we would only need 3 USstates worth of space and the rest of the entire world coudl be used for creatign energy (solar power, wind power ect). It is pretty far fetched but an interesting idea.
BTC Power Law model creator is how i discovered this video.
The only problem is cost of electricity goes up in the city. There should be a savings in scale.
No, he just said that cities scale superlinearly meaning that as they get bigger you get more. Costs go up in cities, not down. Poverty goes up in cities not down. The number of wealthy in cities goes up not down. The price of electricity tends to go up in cities not down.
@balisticsquirel Ok thanks, that's what I thought. At high school we learn how to work scientifically by doing physics experiments and writing papers on them.
uh. does linearity of the line really matter? wouldn't the gradient have to be 10 for it to actually be a linear change wrt mass? His conclusion is right but even though I don't know if my math is right I just know his math is wrong.
we make less "good" use of resources as income grows. just depends which simplistic variables you want to plot to support a theory. a lot of biological life and systems scale according to golden mean, with beautiful spirals of fractal like growth patterns. there are no such symmetries (yet) in cities or companies. but then if u try u can statistically compare resource devouring waste producing cities based on artificial economies to a resourceful zero waste beautiful forest.
@marcdaddy33 Scientific process is about establishing degrees of confidence in assertions, and there are multiple ways to logically do this.
For instance, while our understanding of evolutionary history is predicated on our forensic analysis of the physical record (which is itself predicated on our scientific models of physical processes), evolutionary processes themselves are experimentally observed routinely. Our understanding of biological evolution is by no means unique in this matter.
Malthus was wrong. He was wrong because he did not (and could not) account for innovation.
Natural systems are a lot like free markets. Corps, like species, grow fast if they fit their market niches and adapt to survive changes.
Systems of consumers and corporations are like ecosystems. "Bionomics."
at min 7:50: it is 68%
@truevoice08 It seems to me that the economically literate and those very familiar with "human nature" have gotten us to this point. I prefer a different path for humanity.
He sounds like Christopher Lee
So we need to innovate a way to continue exponentially expanding cities into new growth cycles once we reach the end of the currient industrial / resource economy's curve? Anyone got any ideas on how to build a city without oil, steel, infrastructure etc? And no, the one in The Matrix doesn't count.
@MrKohlenstoff true enough. I think West could have done a better job overall. After watching this I was left thinking...and the point of all this was
In biology size is an impediment to efficiency. The larger something is, the farther stuff has to move. Think of the waste a cell must get rid of. A large cell must work harder to move waste farther from its center to the outside (radius) than a smaller cell. However, to make this analogy to cities, cities are not "bigger". They are more efficient as the distance between people is much smaller allowing them to work together and interact far better than when separated by space.
Cities are meant to grow faster, taller and wider. But efficient and sustainable growth can only be achieved with the commitment of people, governments and policymakers altogether. Duh yes
They used the word "mathematics" to lure me in. I don't feel the talk lived up to the promise of the title though.
:(
You can buy the book "scaling" by him it gives a lot more details!
boring title, surprisingly interesting. I guess I should have trusted the title (surprising)
Saruman says against the power of Mordor there can be no victory!
@except10n - Income != a primary goal? Money is an abstraction of raw energy.Every single thing, until we exit the monetary paradigm we've created, which may never happen, revolves around money. One need look no further than every major, non natural event in the world to see that money (income) is not *a*, but *the* primary goal of human existence, through transitive relation. (See E=MC2 for more information).
@sdrawkcabgnipytmi nope
A radiated tortoise named Tu Malila from Madagascar is considered as the longest-lived animal according to authenticated records. Tu' Malila was hatched in 1777 and died on May 19, 1965 with an age of 188 years. Verification of the age of another tortoise, Adwaita, is still pending; that animal was said to have been born around 1750 and died in 2006 at the possible age of 256
Um, i would think that levels of consumption increase by more than 15% per capital with the doubling of the population in a city, thus more environmental damage and degradation of resources. We all know where this is going to lead in terms of climate change. It is going to rise exponentially and eventually overwhelm us as a species. Pity all the innovation in the world will still not safe us from our own greed.
@eagleeye1975
sweet summary
Something doesn't quite add up. Women are smaller than men but women move more slowly and live longer. By the way, urban planners have known these parametrics for a long time.
the same pattern of growth in plants is capable of producing a small or gigantic tree. the complexity and wasteful resource usage for ever taller (but only partially and part time occupied) skyscrapers is counter intuitive. but artificial economy makes it so. it's not engineering and humanity prowess to build ever bigger skyscrapers, bridges, etc; it's madness.
If cities operate super-linearly, then why does doubling the size only increase the individual components by 15%? Shouldn't it be at least 100%, because the components have to increase by MORE than double?
Late to answer your question but the gradient of that graph is 1.15 so he means there is an ADDED 15%
@BJ219
Not, true. I've researched it. The ones who accept it , many of them, are accepting it for political reasons so they'd be able to say something against big companies and corporations
no matter how you look at it oil has been the catalyst for the development of cities and when the oil runs out (and one day it will, it has to) the collapse will come
so,,.,... we should just build a huge city to minimize waste
so the more wealth I have the better I am for the economy.
Scalability... is it webscale?
@cristoretornebiblia The world had 35 million people back then: en. wikipedia. org/wiki/World_population
So Mohenjo-daro's 35,000 represented one in a thousand worldwide.
.
The world has 6.9 billion now, and NYC has 8.2 million -- which is more than one in a thousand.
I am always skeptical to someone who uses statistics to prove their point. Statistics can be easily manipulated to suit whatever argument you may want. Instead i prefer an argument based in logic.
42
@dreapster I did
@cristoretornebiblia "ancient cities like Mohenjo daro [...] were bigger than new york"
en. wikipedia. org/wiki/Mohenjo-daro "At its peak of development, Mohenjo-daro could have housed around 35,000 residents."
You think NYC has less than 35,000 people?
Am I stupid or is it just a normal occurence that the numbers of people in a city go in exponential tandem with everything related to what the people do? I don't really understand the big revelation here
NickFisher
It's the bigger picture that this represents. I suggest his book, "Scale." Basically, the revelation is that there appears to be a physical law that governs different types of systems that work on similar principal.
Now i just have to find the city with the proper 15%
Get this man some water
This is interesting, but kind of depressing
Entropy ;_;
somebody doesn't believe in aliens @ 6:40
It seems to me that maybe the reason we don't see animals bigger than the blue whale is that this scaling property has its limits. That would suggest that cities stop scaling sub linearly beyond a certain size.
+Brian Newman their size is limited by their networks, so the size of cities is limited to the time it takes to get places by car, wheras the size of an organism is limited by how long it takes for blood to cycle through the whole body.... think of how much smaller cities used to be before the invention of the car and highways
@mazdaplz he's a Math-a-migician!!
Poor companies :(
@StopFear Someone missed the entire point of the lecture. Fail.
@marcdaddy33 - however :) I did up-vote your remark, due to your distrust of those who seek to empirically prod everything, instead of using common sense at times, where it make sense. (ie, when you can't possible have infinite data about every constant and variable in an experiment, empirical research is nothing more than a big pat on the back).
What about Buffalo?
The same basic content in a six page article. www.pnas.org/content/104/17/7301.full.pdf
+Alex Mercado Thanks for posting this.. My teacher wants me to write a few pages on this video lecture... but now I just could read the article instead....
@YawnGod sure, why not.
interesting man.
Argument 1) I want these correlations between biology and sociology to be true, therefore they are.
Argument 2) Correlation equals causation... no this is not fallacious.
Argument 3) Look at these nifty statistical scatter plots, specially manipulated to make them all appear to correlate.
i don't think all companies are destined to die, but most likely have or will. i feel that many companies die due more to coincidence than a natural fate. very often a new president/CEO could be hired and makes dumb decisions. global economy could hinder a company. a meteor could hit the headquarters. there could be a handful of stupid employees that give the company a bad name. a company could last indefinitely, no matter how wealthy, popular, or amount of employees it may have.
@Depotmaster Exactly my thoughts... Yeah, I dont really feel like ive wasted my time, but I do in some way. Heh.
since all living things originate from the sun, it shouldn't be too surprising we're all similar in some perspectives. let's get to the source.
notwithstanding some simplistic insight into city growth; comparison of some graphs on city stats to company stats to an ecology, is nonsensical. the only reason seemed to be to prove some mathematical (apparently) theories. or was this the theory that you can draw similarities between chalk and cheese.