AI and Urbanism

 [Note: I've added a postscript concerning generalists' role at the bottom.] 

The Congress for the New Urbanism is meeting in Seattle: one of the cities where software is made. How can we use the current rage – artificial intelligence – to build communities people love?


Some people reward themselves by cocooning themselves from the rest of humanity. Those of us who tend to like our fellow men and women will invent pretexts enjoy their company. Do coffee houses always have better coffee that we can make at home? No. We can drink at home. We can get take-out. We can shop at home almost perfectly. We want the freedom to be social. Let's choose our technologies, then, according to how we want to live.

Self-driving cars

Are auto-autos going to help us walk, bike, and enjoy neighborhood life? They won't if they swamp our crowded streets with even more traffic. We need them to take us the "last mile" to transit (pdf), but also let us arrange to "just happen" to enjoy social life there. Already, some of our most desirable urban neighborhoods help people enjoy serendipity on the sidewalks, A social transportation system could create real value.

Autonomous vehicles should take people the last mile to transit-served nodes of the polynodal city. Those autonomous vehicles should be built as comfortable, rolling public space.

We should be able to choose to walk or take an auto-jitney from our home or workplace to a bustling town center. The auto-jitneys should be comfortable – with space for strollers and so on. Most of all, though, the whole experience should build our community life.

Robotic Legs

Robots are starting to walk. Many of us probably know the DARPA Robotics Challenge from its comical failures. Robots keeled over for no apparent reason. Now, though, Agility Robotics's Cassie seems to manage walking and balancing. Boston Dynamics's formidable Handle robot does too. Years ago, it looked like the iRobot could revolutionize mobility. It didn't happen. Now, though, Toyota is working on it. SuitX's Phoenix robotic exoskeleton promises to enable people with mobility impairments to walk. All of these use battery power, though, so they need power public places.

It's all impressive technology, but technology doesn't make our cities and neighborhoods livable by itself. We have legacy buildings, porches, front doors, and impressive stairs up to public buildings that ghettoize people who can't walk easily. Yet, those features are important. They give churches dignity or porches privacy – and make the public realm culturally supple. Meanwhile, some of us like to take the stairs – or want to remind our fellows to take them. The 150-year-old technology tends to force a trade-off between equal dignity and fluid urbanism.

Health insurance should cover robotic chairs and exoskeletons. Those devices should help their users to walk with everyone else. Public and private buildings should offer charging points for mobility devices and prosthetics.

People who use wheelchairs should enjoy the social mix with everyone else. As the technology improves, we should use it to help everyone walk with us. 

Augmented Reality and Surveillance

Some software companies, notably Facebook, are trying to make augmented reality happen. Not to be confused with virtual reality, which is a 3D environment often experienced with opaque goggles, augmented reality overlays virtual elements on top of real reality. Think of it as Pokemon Go crossbred with Google Glass.

This augmented view could enrich neighborhood life. It could help transit riders find buses, and it could tell tourists about where they are. It pipe information like Foursquare's by placing information in lenses – limited only by our science-fiction nightmares. Unfortunately our brains are too good at throwing away unhelpful information. If we try to pay attention to both real reality and augmented reality, we might walk into traffic. We need keep the augmentation from becoming a mental scrim between us and the people around us.

Then there is the information itself. Walmart has experimented with facial recognition that helps catch thieves. Software tools can detect certain behaviors in video. Amazon is piloting Amazon Go in Seattle. It lets shoppers avoid going through a checkout, just by watching what they do. (Employees only for now.) It can be scary. Yet, there is promise too. Perhaps friends or people who have given us their cards would let us pull up their information when we see their faces. Police departments could show us Amber Alert details. While technology like Amazon Go's might put some employees out of work, we also go out for human contact. Even if we get our (Seattle-branded) coffee from a dispenser instead of a barista, the barista might become a sort of coffee geisha paid to host customers. For all the science-fiction scares, there are hopeful explorations of emergent public life.

Augmented Reality should help people engage the real world, not lose track of it. We should explore those new uses for public and private information that augment our public life without chilling it.

Part of the magic of public space is engaging with our fellows; part of the magic of cities is urban anonymity. We should welcome technologies that let us enjoy this magic, and we should reject technologies that don't.

Putting it Together

An auto-auto / robotic / surveilled-and-augmented future can be as exciting as it is scary. The exciting part is easy to imagine: sociable neighborhoods where everyone can walk (or bike on April 1st), and where we can be both present in the moment and well-informed. This vision is easy to imagine because it just enhances the way our best places work now. The scary vision is familiar from cautionary fiction. Our employer might buy our location history to see if we crossed paths with competitors. We might be overwhelmed with restaurant reviews while strolling hand-in-hand. We might not be able to cross streets easily for all the robotic cars. The military might monopolize robotic prosthetics and mobility devices so we won't be able to use them when we come to need them.

All of that can happen if we keep silent about our humane agenda for fear of sounding like geeks. We can't afford to take a wait-and-see attitude. Car manufacturers, giant retailers, armies, and everyone else has an agenda. Just because ours is benign doesn't mean it can't compete.

We should adapt technologies that use artificial intelligence to build places people love. We should not prostrate ourselves to technologists' agendas.

Postscript:  I received some comments and questions off the Pro-urb Listserv. I answered them, but will elaborate here. 

 I don't think that we can wait for the technology to shake out because while science is value-neutral, technology is laden with its developers' values. Its development needs to pay for itself, which means that the system will have to include a turnstile somewhere. We should watch the European Union. Their Citymobil2 pilot and their general concern with privacy address some of the issues above. However, we need to understand their values too. They like to protect incumbents, as we know from Leistungsschutzrecht. Citymobil2 shows how much people like the atmosphere of the vehicles and public realm, but since traditional urbanism is old hat to them, they would probably be surprised by the added poignancy they might have in the United States. As for prostheses, we have a good attitude that comes from the tragedy of so many veterans needing devices. We should watch the Veterans Administration. We should also watch Japan. They have been using robotic exoskeletons because they have to care for an aging population. Neither of these is necessarily optimized for dignity in public, although the VA comes closest.  

 I appreciate that I am not a technician. I have refrained from commenting on the technology itself. However, technicians are not always aware of their own biases. 

Auto-autos, Serendipity Apps, and Social Hubs

We often hear that autonomically-controlled vehicles will eliminate parking lots, reduce congestion, make streets safer, and make streets into places we love. The National Association of City Transportation Officials (NACTO) hopes so too. Yet, this vision also requires new ways of choosing itineraries and new social hub.

Traffic will get worse rather than better if auto-autos roam around empty or if riders just want to nap in solitude. We must change two more things besides transportation: the people we travel with and meet and the physical places we pass through.

TL:DR: If we want self-driving vehicles to work as advertised, we need social hubs and apps to encourage us to share the ride.

Read More

Architecture for Urbanism

Prompted by a 2015 meeting in Charleston South Carolina, I have worked with a few people to craft a set of "Canons for Humane Architecture," to complement the Canons of sustainable Architecture and Urbanism adopted by the CNU. This is a very preliminary draft, yet it probably comes fairly close to what most New Urbanists who prefer traditional architecture would agree with.

I will be proposing this in an "Open Source Congress" meeting at CNU 24, and welcome comments and suggestions.

De-structured space

What I call de-structured space here (pdf) is the kind of work we find in naturalistic Japanese and Chinese art and gardens, and in picturesque Western gardens and art.  

"Sargent Juniper, 1905-2007" by Ragesoss - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons

"Sargent Juniper, 1905-2007" by Ragesoss - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons

As natural as such works look, they require a great deal of care in design in order to sidestep the typical human desire to create simple rhythms and symmetries. The bonsai above looks natural in its way, but in reality it's quite difficult to create. The bonsai artist has to balance the asymmetry, make the trunk look like it's striving against mighty forces, and incidentally weave in driftwood.

Look at the wall behind, though. If the bonsai were at the side of a driveway next to a coiled-up garden hose, it would not even be apparent that it was tended. The blank wall behind it is just as important as the bonsai itself.

1st Principle: Break Order

The first principle of de-structuring is to break the easy symmetry of a doodle. 

"Doodle" by Dmn - Own work. Licensed under Public Domain via Wikimedia Commons

"Doodle" by Dmn - Own work. Licensed under Public Domain via Wikimedia Commons

This collection of doodles is decidedly untutored. However, several of the doodles have repeating, alternating patterns, symmetries, and rhythms. A couple of them grow from small to large. This basic instinct to pattern, which comes out even in a sketch, has to be carefully broken in order to create something like the bonsai. 

2nd Principle: Provide a foil

The bonsai would be difficult to appreciate against a random backdrop, or any strong  backdrop. It sings when against a blank wall. Together, the blank wall supports the bonsai. Even a very large building can attempt to erase itself, so to speak, by making itself a foil.

"John Hancock Tower - Boston, MA - DSC08138" by Daderot - Own work. Licensed under CC0 via Wikimedia Commons

"John Hancock Tower - Boston, MA - DSC08138" by Daderot - Own work. Licensed under CC0 via Wikimedia Commons

The John Hancock Tower in Boston was supposed to melt into the sky, so that its grid worked as a foil for nature: the clouds and reflections of the city. Of course it doesn't always work, but it certainly can. In the previous post, I mentioned that a pergola can act as a neutral backdrop. Certainly the best pictures of modernist architecture have trees and free-form sculpture in the foreground.

3rd Principle: A Subjective Stance

The third principle is to engender a subjective stance. Artists can usually depend on people to try to appreciate their art. This subjective stance is easy to come by in fine art, but difficult in architecture -- particularly urban architecture. When we come on architecture in everyday life, we're usually preoccupied by mundane concerns. Of course, we can often enjoy modern architecture when we have a chance to prepare.

Structured and De-structured Space

The following is an excerpt from Structured and De-structured Space (pdf). That paper introduces two kinds of compositional space I call structured and unstructured. Those are explained here. A third kind of compositional space, de-structured, is described in a subsequent post. (This post has been edited since it was initially posted.)

Read More

Evolutionary Architecture: 2

The style wars between modernism and traditional design usually get stuck on the false choice between continuity and change, as if tradition just means copying the past, and breaking tradition equals modernity. Neither really captures the algorithm we need to follow in order to build better places: evolution. Evolution begins with the past, copying it. Then it goes much further. 

Read More

Evolutionary Architecture: 1

There used to be a big debate on the beautiful versus the sublime. Forest edges are usually beautiful. They do not usually include things like sheer cliffs or water that goes to the horizon. Those have an "awesome beauty" or “sublime” quality that puts one at least slightly on guard. Our ancestors usually sought out beautiful places and were appropriately on guard around the sublime or awe-inspiring.

Read More

Markets on Platforms as a Way to Solve Problems

Sometimes problems are big and regulations are onerous. They are too much for a small business, small builder, or everyday homeowner. One approach, taken in Lean Urbanism, is to reduce regulations, or to make compliance easier. Another approach is to both cooperate and compete: cooperate to create a platform supporting each other, and then compete in a market.

Read More