For all the optimism, innovation and wealth that are produced here in San Francisco, the Bay Area can also feel like a place that doesn't work quite right.
The cost of housing has priced out teachers and line cooks. Income inequality is among the widest in the nation. The homeless crisis never seems to ebb. Traffic is a mess. On bad days, transit is, too. And local governments are locked in conflict.
Clearly, the region has not been optimised. "It could be so much better," said Ben Huh, who moved to San Francisco in 2016 after running the Cheezburger blog empire in Seattle. "There's so much wealth. There's so much opportunity."
In the maddening gap between how this place functions and how inventors and engineers here think it should, many have become enamoured with the same idea: what if the people who build circuits and social networks could build cities, too? Wholly new places, designed from scratch and freed from broken policies.
Huh leads a project begun by the startup accelerator Y Combinator to explore the creation of new cities. Hundreds applied to work on what looked like "the ultimate full-stack startup." Last October, Sidewalk Labs, an Alphabet company, announced it would team up with a government agency in Toronto to redevelop a stretch of the city "from the internet up."
For others in tech - intrigued by word of a proposed smart city in Arizona, a big Bitcoin land grab in Nevada, a special economic zone in Honduras - fantasising about newly built cities has become a side gig. They dream of utopias with driverless cars, radical property-ownership models, 3-D-printed houses and skyscrapers assembled in days.
While some urban planners roll their eyes, it is true that America's cities have always been built on someone's hubris, whether the characters who plotted New York's street grid or those who imagined the Golden Gate Bridge.
"Who were these guys who were thinking so big? Then the question is, where are those people now?" said Paul Romer, the former chief economist at the World Bank, whose ideas (and TED talks) on new "charter cities" have influenced some in tech. "Tech types - as much as people might talk about the parochial way they're approaching it - deserve credit for thinking bigger than anybody in government right now."
Their interest has an internal logic to it. The tech industry tries to produce better versions of familiar things - cheaper phones, smaller computers, faster chips. But cities like San Francisco don't seem to be evolving into more efficient versions of themselves.
And if you take literally the economist Ed Glaeser's assertion in "Triumph of the City" that cities are our greatest invention, it ought to be possible to reinvent them. The idea isn't such a stretch, the dreamers say, when Elon Musk is already shooting rockets into space and trying to bore tunnels for a transit "hyperloop."
"You now have a lot of people who have seen a lot of success thinking, 'Well, how can I one-up that? What's bigger than starting a multibillion-dollar company?' " said JD Ross, the 27-year-old co-founder of Opendoor, a homebuying company that has been valued by investors at more than $1 billion. "We have the home screen on our phone, we have the home button in every app. But it really comes down to people's actual homes - that's much more important."
To planners and architects, all of this sounds like the naivete of newcomers who are mistaking political problems for engineering puzzles.
Utopian city-building schemes have seldom succeeded. What we really need, they say, is to fix the cities we already have, not to set off in search of new ones. But it is hard to overstate the degree to which these tech entrepreneurs are looking at the world in ways that would be almost unrecognisable to anyone already working on urban problems.
The idealised city
The model cities Huh and others in tech describe are not so different from what many urbanists want. They aspire to tame NIMBYism and private cars. They want to create walkable neighbourhoods, albeit around hyperloop lines that would travel faster than any bullet train. They're focused on affordable housing, though the shortage of it looks to them less like a matter of policy than a problem that better construction technology can solve.
"We have not affected the fundamental building blocks of infrastructure and society," Huh said. "We've made this better," he added, gesturing to his laptop. "We've made the new things better. We haven't made the old things better."
In thinking about how to do that, people in tech prize "first principles," a concept that suggests that historical awareness and traditional expertise can get in the way of breakthrough ideas. The approach has worked before.
Uber wouldn't exist if Travis Kalanick had begun by researching how taxis were regulated around the world. Uber instead produced a service that violated those rules, and changed how millions of people travel.
With cities, this means stripping away the histories of other utopias, the building codes that shape San Francisco, the political dynamics that block change.
The tabula rasa is alluring not just for the lack of buildings, but also the absence of rules. Huh and others proudly say this leads them to odd-sounding questions: How much does a city cost? Why can't you construct a skyscraper in days? Could you fit a city's rulebook into a hundred pages?
Technologically optimised cities, said Nicholas de Monchaux, a designer and Berkeley professor, failed then for the same reason they would be unsuccessful now. Technology can help reduce traffic, or connect you faster to a ride home. "But a city is not at its fundamental level optimisable," he said. A city's dynamism derives from its inefficiencies, from people and ideas colliding unpredictably.
It's also unclear what you'd optimise an entire city for. Technologists describe noble aspirations like "human flourishing" or "quality of life." But noble goals come into conflict within cities. You could optimise for affordable housing, but then you may create a more crowded city than many residents want. You could design a city so that every home receives sunlight (an idea the Chinese tried). But that might mean the city isn't dense enough to support diverse restaurants and mass transit.
These trade-offs demand political choices. And so technologists hoping to avoid politics are bound to encounter them again.
Of the techno-urbanists, Alphabet's Sidewalk Labs seems to be closest to actually creating something. The company, run out of New York City by former deputy mayor Dan Doctoroff, concluded after a year of study that it needed a not-quite-blank slate to truly innovate.
With too many people or buildings already in place, it could never install an energy grid, or test what happens when you ban private cars. But a stand-alone city in the middle of nowhere wouldn't work, Doctoroff said, because people wouldn't want to move there.
"The smart city movement as a whole has been disappointing in part because it is hard to get stuff done in a traditional urban environment," Doctoroff said. "On the other hand, if you're completely disrespectful of the urbanist tradition, I don't think it's particularly replicable. And it's probably pretty naive."
