A City Is Not a Computer (2017)
placesjournal.org
placesjournal.org
I don't know if the authors of this article are substantially less guilty of abstractionism and impracticality than Sidewalk Labs or whoever. Knowing how pigeons understand cities might be interesting, but I'm not sure it's interesting enough to center urbanism around. The wisdom of local communities can be seen in angry letters to the council about how bike lanes are satanic, contemporary statuary seems to be about giving a well-connected artist a big chunk of cash to make something that the residents of a city find unpleasant to look at.
This is such a stupid dismissal and a provides a great example of what the article is trying to point out. This paternalistic ultramodernist and nearly authoritarian "we know better" mentality.
https://en.wikipedia.org/wiki/A_City_Is_Not_a_Tree
1965 essay:
https://blogs.ischool.berkeley.edu/i103su12/files/2011/07/19...
> It seems that even general structural configurations, such as graphs or digraphs with certain specified properties, will fail to characterize a city, whose complexity, at this stage, may well continue to be understood more readily through negative rather than positive descriptions.
A major difference between cities and computers is that cities self-organise. (the larger ones organise not only themselves but also their suburbs and rural hinterlands; for the US context see https://en.wikipedia.org/wiki/Combined_statistical_area but note that countries are ephemeral with regard to cities)
see also Jane Jacobs, The Economy of Cities (1969)
This is not a precise statement. What is meant by "reduced" and "computation?" "Computation" in often used in a hand wavy manner similar to how some people use "economics" to gobble up everything.
Computation as defined in computer science is, first and foremost, not an objective phenomenon, only a formal model (inspired or meant to formalize "effective methods") that can be simulated using various natural phenomena, but primarily using pen and paper. But there is nothing objectively computational going on in those natural phenomena. An abacus is not counting. A mechanical computer isn't adding. An electronic computer isn't really computing, because there is no objective phenomenon known as computation, except as a mental operation. We call something a computer for only one reason: it can be interpreted in a computational manner.
Another feature of computation models is that they are formalizations. Formalization, by definition, involves "desemantification", leaving us with a residue of gutted forms and some syntactic rules to work with. Formalization lends itself to mechanization.
I am reminded of an adjacent observation of Russell's on the subject of physics. Physics intentionally ignores everything about reality that doesn't yield to mathematical description. That's part of the method, and indeed, a good deal of the particular success of physics, and the physical sciences in general, is due to this approach. To declare that physics is all there is is shear selection bias. You confine your investigation to only those things that physics can handle, and then declare that, because physics has been successful handling such things, physics is all there is.
Yes, that is hand wavy, but it's what I had meant.
Nothing in this statement distinguishes computation from any other category of thought. The same case can and has been made for trees. These are empty words.
Government shouldn't be based on profitability, it should be based on services and outcomes. Adding all these stupid expectations dilutes the entire point of the endeavour.
For example, see people's obsession with Transit rates. Just make it free and extract the costs fairly across taxation brackets because even those who don't use it still benefit from the infrastructure it brings (for example, if you drive and someone transits, thats one less car on the road for you). But obsessing that public transit needs to run like a business is absurd because profit in this case doesn't actually provide any economic benefit as an incentive model.
Cities are incredibly complicated things that no central planner can ever fully understand. There are layers upon layers, all interacting with each other, and all changing constantly.
Only a silicon valley programmer, blinded by hubris, would assume they can "optimize" a city, as if their unique ability to program computers enables them to analyze and improve incredibly complex social structures. They will fail.
For practical purposes, everything can be modeled as a computer. But some things are impractically complex.
What's interesting is that modern computer software, and to some extent computer hardware (even after assuming perfect digital abstraction at the transistor level), is itself becoming impractically complex!
Attempting and failing to centrally control cities has been a recurring motif for ages. I'd love to say it's changing, and maybe it is, but not nearly as much as it should.
Between fatalism and high modernism, there's room to actually improve things on the margins, check to see how it's going, and iterate.
What bugs me is that the programmers of today seem to think they can optimize everything. Not only are they incapable, they're simply unqualified. At least professional urban planners have studied the topic in depth. Google throwing an engineering team at cities is just ridiculous. Might as well poll dentists on the best way to build software.
> Might as well poll dentists on the best way to build software.
A programmer with hubris would take VC money to "make better dental software" never having met a dentist or having done any dental work, perhaps only ever taken a scant look over at the monitor and deciding it's not fast enough or whatever. Then they'd build an MVP, try and sell it to some clinics, and upon being laughed out the door, would take more VC money to try and reinvent dentistry because their parents never told them "no"
But yeah, we can just barely even program a computer. Planning a city to a high degree of granularity is out of the question. Best we can do is apply constraints, observe the results, and iterate.
If the epistemology they're using isn't good, I don't see any real suggestion on what we should be using. What can we learn from the fields of Media Studies and Art History that can help us make better city planning decisions?
It also commits the sin of criticizing some nebulous other, an unclear amorphous cloud of tech-bros, capitalism, and other symbols for things people just don't like. Is the app that tells me if the bus is running late not an example or these connected cities? Traffic lights that adapt to traffic flows? Are we not building a smart city with every connected bus, every car-detecting traffic light, every pedestrian crossing with a button?
This article doesn't clearly point at a problem, or clearly impart any lessons that I can learn. It just vaguely points in the direction of companies and people the author doesn't like.