[HN Gopher] A City Is Not a Computer (2017)
___________________________________________________________________
A City Is Not a Computer (2017)
Author : conanxin
Score : 22 points
Date : 2024-04-14 06:04 UTC (1 days ago)
(HTM) web link (placesjournal.org)
(TXT) w3m dump (placesjournal.org)
| 082349872349872 wrote:
| (pedantry: just about everything can be reduced to computation,
| just not necessarily deterministic computation)
|
| 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)
| Zambyte wrote:
| Software self-organizes as much as cities do.
| hermitdev wrote:
| Nah, waves of gentrification are just bubble sort in
| action... j/k
| goatlover wrote:
| Ant colonies self-organize, but they don't do any useful
| computation for us. What meaningful sense is there to calling
| things computers which aren't machines we've designed to
| compute for us?
| 082349872349872 wrote:
| Good point; what I had been trying to say is that software
| (and especially hardware) is architected, but architecting
| cities usually leads to problems.
| lo_zamoyski wrote:
| > just about everything can be reduced to computation,
|
| 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.
| brudgers wrote:
| Related, _A City is not a Tree_ :
|
| https://en.wikipedia.org/wiki/A_City_Is_Not_a_Tree
|
| 1965 essay:
|
| https://blogs.ischool.berkeley.edu/i103su12/files/2011/07/19...
| awaythrow999 wrote:
| Not just related. It __is__ the essay. This is like rewriting
| No Silver Bullet without giving credit to Brooks.
| ysofunny wrote:
| but how about a forest?
| tivert wrote:
| We are software engineers, and we proudly believe that every
| problem is a nail to our hammer, all meaningful knowledge is but
| a slight variation of the computer science concepts we've already
| mastered.
| lo_zamoyski wrote:
| This is what happens with the obliteration of true philosophy
| and the rise of intellectual philistinism. The sciences
| fragment into parochial principalities, and without recourse to
| genuine metaphysics, they elect some preferred field to take
| its place, making a metaphysics out of its method. In the
| physical sciences, physics is elected to that position. Mind
| you, neither physics as a science, nor any of the sciences as
| such, necessitate this, but scientists, like anybody else,
| cannot function without an _arche_ or orienting first
| principles, and in our culture, materialistic metaphysics is
| insinuated into the background intuition.
| idontknowtech wrote:
| Silicon valley in general seems to think everything should be a
| computer, or at least as efficient as one.
|
| 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.
| goatlover wrote:
| Biology is possibly even messier and less like a computer. For
| that matter, biology isn't even really machine-like, only
| pseudo-like in some ways. But some will rebut that the universe
| and everything is doing computation, as if it's meaningful to
| consider a rock or a dandelion to be a computer.
| tikhonj wrote:
| Only a silicon valley programmer... and virtually every
| modern(ist) government in the past 150+ years?
|
| 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.
| idontknowtech wrote:
| Yes definitely. But the fields of planning and architecture
| have learned lessons from those mistakes. These engineers,
| being fundamentally ignorant of their subject matter, are
| guaranteed to replicate the errors of the past. They're
| simply not qualified to do the job they're trying to do.
| hot_gril wrote:
| A SWE shouldn't even think of software as simple. "Layers upon
| layers, all interacting with each other, and all changing
| constantly" describes any software org.
| idontknowtech wrote:
| To take the metaphor further, it's like this except the
| software is hosted on GitHub and open to public commits with
| limited moderation. It grows and changes on its own without
| anybody really knowing what's happening in total. (Here your
| metaphor needs no expansion)
| somnic wrote:
| Hubris with regards to urban planning is hardly unique to
| programmers. Planning boards and transit authorities and city
| councils and everyone who's got opinions on those things is
| intervening, and sometimes messing things up. Cities are not in
| some romantic state-of-nature where everything is organic and
| bottom-up and intervening risks destroying the ecology.
|
| Between fatalism and high modernism, there's room to actually
| improve things on the margins, check to see how it's going, and
| iterate.
| idontknowtech wrote:
| Completely agree that hubris in this area isn't unique to
| programmers. I'm an architect, and plenty of architects have
| developed some truly idiotic ideas for ideal cities.
| Corbusier's plan for Paris was to level it and replace it
| with skyscrapers.
|
| 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.
| imzadi wrote:
| Similarly, a government is not a business.
___________________________________________________________________
(page generated 2024-04-15 23:00 UTC)