5 ms·

The proliferation of 'smart-' things is the equivalent of deferring everything to runtime, and throwing out design and planning and replacing it with a reactive

by hyperion2010 7y ago

The proliferation of 'smart-' things is the equivalent of deferring everything to runtime, and throwing out design and planning and replacing it with a reactive (or perhaps more accurately reactionary) approach. Interpreted languages are nice, but compiled languages have enormous advantages, especially when problems are well understood. City planning, the design of public spaces, thoughtful zoning, etc. is like the compiler. If you want an efficient city, you probably don't want it to only run as an interpreter. Somehow this evokes the depths of American anti-intellectualism for me. "We don't need to understand anything, we can just measure it and then understand it!" without so much as a pause to wonder whether an understanding of the problem might already exist in such a way that it can be incorporated into the design, and further, that if you don't already understand 'it' then there is a good chance that you will end up building a device that measures the wrong things and that you can't change easily.

gkolli 7y ago
great comment and analogy
bobthepanda 7y ago
Cities are chaotic, complex creatures. Anyone who claims to understand them is more or less just working from the gut; and anyone who claims to "solve" them via technology, planning, or otherwise is fooling themselves. The history of city planning is littered with failed examples. In America, freeing everyone from the vapors by separating them in the free, open countryside mostly led to people being trapped in their cars in traffic because they had to drive to get anywhere. Demolishing slums and shoving everyone into spacious housing projects mostly led to isolation and ghettoisation. And even the mere act of requiring parking has separated everything with inhospitable parking lots, making it a requirement to drive to do anything. Arguably the most successful large cities did very well with low amounts of planning; Manhattan's famous grid was literally imaginary roads plotted onto a map, without a care for what would fill in the gaps. Planned cities are like the manicured lawn; they look pretty but they don't really support any actual life. The wilderness is chaotic and messy but it supports incredible biodiversity.
- hyperion2010 7y ago
  I agree completely. Planned communities (top down?) have failed repeatedly (for a variety of reasons). The kind of planning and design I am talking about is the design of local rules (bottom up?), such as the order in which you put cars, parked cars, bycicles (and friends), and peds. Another example would be about the zoning around university campuses and how they interact, or don't with public transportation. The difference between US vs EU for these two cases and the results of the differences is quite revealing.
  - em-bee 7y ago
    in other words, designing a smart city is actually just as much top-down as any planned community. and what we really need is for each community to decide for themselves which technology they pick to improve their lives
    - notfromhere 7y ago
      desire paths, but for urban planning. Plus, the whole development model since post-ww2 has basically been a veiled ponzi scheme
em-bee 7y ago
that's an assumption, but we don't know if it is true. we have yet to actually see a smart city, and until we do we can't know how it will work out. we only know the effect of some of the pieces. we know about the effects of surveillance. of insecure technology, data leaks, etc. in other words, what we do know is that the technology is not ready yet for a smart city to be safe. but apart from that we don't know how it will work out beyond that. for that we need to experiment. someone needs to participate in that experiment so we can learn from it. is it time already to experiment now? i don't know. we could wait for the technology to be safer. but do we know what the actual threats are? even that we need to test out.
- bobthepanda 7y ago
  We have an attempt at a smart city already, and it hasn't gone well: https://www.citylab.com/life/2018/06/sleepy-in-songdo-koreas-smartest-city/561374/ https://www.citylab.com/life/2018/06/sleepy-in-songdo-koreas...
  - em-bee 7y ago
    it doesn't seem that the smart aspect of it is the problem there though, but rather bad planning and the expectation that smart alone would attract people to move there. the discussion continues here: https://news.ycombinator.com/item?id=20475258 https://news.ycombinator.com/item?id=20475258 turns out there is quite some criticism about the article

gkolli 7y ago

great comment and analogy

bobthepanda 7y ago

Cities are chaotic, complex creatures. Anyone who claims to understand them is more or less just working from the gut; and anyone who claims to "solve" them via technology, planning, or otherwise is fooling themselves. The history of city planning is littered with failed examples. In America, freeing everyone from the vapors by separating them in the free, open countryside mostly led to people being trapped in their cars in traffic because they had to drive to get anywhere. Demolishing slums and shoving everyone into spacious housing projects mostly led to isolation and ghettoisation. And even the mere act of requiring parking has separated everything with inhospitable parking lots, making it a requirement to drive to do anything. Arguably the most successful large cities did very well with low amounts of planning; Manhattan's famous grid was literally imaginary roads plotted onto a map, without a care for what would fill in the gaps. Planned cities are like the manicured lawn; they look pretty but they don't really support any actual life. The wilderness is chaotic and messy but it supports incredible biodiversity.

hyperion2010 7y ago
I agree completely. Planned communities (top down?) have failed repeatedly (for a variety of reasons). The kind of planning and design I am talking about is the design of local rules (bottom up?), such as the order in which you put cars, parked cars, bycicles (and friends), and peds. Another example would be about the zoning around university campuses and how they interact, or don't with public transportation. The difference between US vs EU for these two cases and the results of the differences is quite revealing.
- em-bee 7y ago
  in other words, designing a smart city is actually just as much top-down as any planned community. and what we really need is for each community to decide for themselves which technology they pick to improve their lives
  - notfromhere 7y ago
    desire paths, but for urban planning. Plus, the whole development model since post-ww2 has basically been a veiled ponzi scheme

em-bee 7y ago

that's an assumption, but we don't know if it is true. we have yet to actually see a smart city, and until we do we can't know how it will work out. we only know the effect of some of the pieces. we know about the effects of surveillance. of insecure technology, data leaks, etc. in other words, what we do know is that the technology is not ready yet for a smart city to be safe. but apart from that we don't know how it will work out beyond that. for that we need to experiment. someone needs to participate in that experiment so we can learn from it. is it time already to experiment now? i don't know. we could wait for the technology to be safer. but do we know what the actual threats are? even that we need to test out.

bobthepanda 7y ago
We have an attempt at a smart city already, and it hasn't gone well: https://www.citylab.com/life/2018/06/sleepy-in-songdo-koreas-smartest-city/561374/ https://www.citylab.com/life/2018/06/sleepy-in-songdo-koreas...
- em-bee 7y ago
  it doesn't seem that the smart aspect of it is the problem there though, but rather bad planning and the expectation that smart alone would attract people to move there. the discussion continues here: https://news.ycombinator.com/item?id=20475258 https://news.ycombinator.com/item?id=20475258 turns out there is quite some criticism about the article