Editor’s Note
Mike Szpot

Cities are for people (and bots)

In our lifetimes, it’s hard to think of a topic that has been on as many people’s minds as artificial intelligence these last few years. 

Since ChatGPT began dazzling users with the seeming authority and ersatz humanity of its responses back in late 2022 — and with every new model released by OpenAI, Anthropic, Google and their competitors capable of deeper, faster thinking and agentic action — everyone’s been trying to make sense of what our new robot helpers (or overlords) mean for our economy, culture and democracy. You hear it at kitchen tables and in classrooms, at dive bars and at fancy parties. This is the rare case where the general public may be out in front of elite actors in politics and universities.

The doomer version of the questions swirling around go like this: Will AI replace workers en masse in a range of industries? Devastate human relationships already scrambled by the ubiquity of social media? Hollow out the creative process? Make it utterly impossible to tell disinformation from fact, and supercharge fraudsters? Lead to widespread cognitive offloading or downright cheating in our schools, emptying education of its point? (Oh, and will the advent of superintelligence end civilization as we know it?)

The techno-optimist version sounds different: Will AI take over the drudgery that eats up so much of our days? Give people living in isolation a source of intelligent help at any hour? Put new creative tools in the hands of those who never had the means to make films, music or software? Help journalists and citizens check claims faster than liars can spread them? Give every student, no matter family income, a tutor that adjusts to how they learn? (And oh, will AI soon cure cancer?)

We are as interested in all this as the next Homo sapiens, although, perhaps inevitably, our answers tend to come out in the unsatisfying middle. Instead of wrestling with either the dark portents of a Skynet future or the utopian hype of tech hucksters, as an urban policy journal, we want to ask a basic set of questions: What can these new technologies do for cities? What capabilities will they unlock? And, just as important, what risks will they pose?

Cities in general and New York City in particular differ from suburbs and rural areas in important ways. They are densely packed centers of economic activity, with more than their share of knowledge workers on the one hand and people in poverty on the other. They have high tax bills with expansive (if hole-riddled) safety nets, large police forces and elaborate transit networks. 

All this is fertile ground for intelligent technology-driven reinvention, or potentially for disruption that breaks things first and only bothers to pick up the pieces later.

As many of our readers will recall, New York has already learned some of this the hard way. Early this year, the City under Mayor Zohran Mamdani pulled the plug on its MyCity chatbot, launched by Mayor Eric Adams, after reporters caught it advising business owners to break the law. The City’s own required inventory of algorithmic tools now lists 86 of them being used for everything from handling 311 requests to flagging child welfare cases for review.

About half the authors in this issue take up the most practical question at hand for New York: how chatbots, AI agents, computer vision and old-fashioned data crunching can make government more efficient and effective, while avoiding the pitfalls. Many of the essays are reasonably optimistic about how AI can be used for good; that’s by design, because here at Vital City we are genuinely hopeful that innovations can be harnessed to make cities work better.

The essays also often argue with one another, as they should. One author says Washington, not City Hall or Albany, should be the one regulating AI; another says the job falls to cities precisely because Washington won’t do it. A former deputy mayor proposes a bargain in which the City shrinks its workforce through attrition as AI takes on more of the work, while a government tech veteran warns that a state law already on the books may forbid much of that. 

Beyond the focused look at transforming the way government works, we also wanted to step back and wrestle with some of the broader ways AI might reshape cities for better and worse. There’s little doubt that over time, it will change the mix of industries that thrive here, but how quickly and how? Amara’s Law holds that we overestimate a technology’s effect in the short run and underestimate it in the long run, but it’s been broken plenty of times — the iPhone and mRNA vaccines come to mind for remaking daily life far faster than that rule would predict — and there’s no guarantee AI will follow the slower path. And will it remake the character of large urban centers differently than it does the rest of the country?

We know we’re only touching the tail of the elephant here. Once you start looking, you see new questions emerge everywhere you turn. That’s especially true given the breakneck pace at which AI is evolving. Yesterday, many AI models struggled to do basic math; this fall, OpenAI said a model it built had cracked one of the world’s hardest math problems, though mathematicians are still arguing over what it proved.

A word on how this issue was made. Though some of our contributors used AI along the way, authors have assured us that every piece we’re publishing is their own work, and we checked each one with Pangram, a tool that detects AI-generated text. We used AI ourselves, for research and, in some cases, to test our reasoning. We even fed a draft of this note through Claude and asked what we were missing. And built a subway-style idea map of the themes in the issue.

Pick a line or a stop to explore. Open the full-page map ↗

AI can do a lot, both good and bad. How the balance works out for cities, and New York especially, is something not even superintelligence can wisely predict.


Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Vital City.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.