A Conversation on Civic Tech, AI, and the Future of Cities with Andrew Rasiej

The founder of Civic Hall gives me a hall of an earful.

Sam Sklar

We’re recording now. I’d love for you to start by introducing yourself and your background—just so my readers can get an understanding of who you are and why we’re talking today. You’re not my typical interview partner in a lot of ways. I mostly talk to transportation planning folks, but my blog is called Exasperated Infrastructures, and I think that’s a lot longer and broader a topic than simply transportation.

Andrew Rasiej

My name is Andrew Rasiej, and I’ve been at the intersection of technology and the public good for thirty-some years. My first foray into this arena was an organization I started back in 1995 called MOUSE.org, which wired public schools to the internet. At the time, almost all schools in New York City had no internet connection and very few computers. That program still exists.

From that experience, I spent a lot of time talking to elected officials about the need to wire public schools and got deeply interested in what I’d call civic tech—the idea that technology could be used to solve problems, create more transparency in government, and help deliver services.

In 2003, I started a conference at the intersection of technology and politics called the Personal Democracy Forum. I worked as a senior advisor to the Sunlight Foundation, which built an API allowing people to track where money in politics was going.

In 2015, I co-founded Civic Hall with Micah Sifry—a collaborative workspace on Fifth Avenue and 20th Street in New York that brought together people from the technology community with policymakers, nonprofits, government officials, foundations, philanthropy, and corporations. Within a year, we had over a thousand members. Within two years, we had done over two thousand events—brown bag lunches, book talks, panel discussions, demo nights, hackathons. We launched something like fifty or sixty new nonprofit initiatives out of that experience.

In 2017, I responded to a city RFP to build an office building in Union Square, and I proposed a new and larger Civic Hall. Seven years later, Civic Hall at Union Square opened—90,000 square feet. It’s the largest digital skills training center of its kind in the country. I’m a big follower of Eric Klinenberg, who wrote Palaces for the People about the need for civic infrastructure, and I’m very happy that Civic Hall got built—thirty-five million dollars of fundraising later, which was really hard to do during Covid.

Along the way, I’ve spent a lot of time advocating for New York City to catch up with the private sector in how it thinks about technology—how it could save money, deliver services better, create more equity and accountability. I was shocked from the beginning at how disconnected not just the school system, but the political and governmental systems were from the internet.

A quick side story: I ran for Public Advocate in 2005 on a platform to make New York City a wireless city, arguing that broadband was a public utility. When I went to the New York Times editorial board for my endorsement interview, they asked me what Wi-Fi was. This was 2005. People just didn’t understand.

During the twelve years of the Bloomberg administration, even though they created the office of data analytics, very little was done to renovate government digitally. The de Blasio administration appointed a CTO for the city, but with a very small budget and little opportunity to lead digital transformation. The Adams administration also did almost nothing. And now we have a new mayor who, even though he’s digitally native, has not shown real interest in this topic. With AI marching through our society largely unfettered, I’m very worried about the future of the city and, to some degree, the future of the country.

MOUSE, the Digital Divide, and AI

Sam Sklar

That’s a fascinating story. I think a lot of my readers are going to be very interested in MOUSE, particularly because you identified a very obvious challenge. I’m curious—how long do you think it would have taken for the city to wire all the schools without your vision?

Andrew Rasiej

Public schools are wired now, but they’re only open about fifteen percent of the time in the year. And now we have a cell phone ban, which I don’t completely disagree with, per se. But to me, the failure to make the internet available in public schools twenty-five years ago, and now the banning of cell phones, both reflect a failure of public policy—a failure to understand these technologies. Technology is a tool. It can be used for good, and it can be used for bad. If policymakers don’t understand it, they can’t develop policies to take advantage of it.

The same goes for AI. In my opinion, the only way to fight bad AI is to build lots of good AI. And the only way to build good AI is to empower the people responsible for good in the world with the same tools and skills that the bad actors have.

Unfortunately, most foundations are still just developing strategies around AI. Most nonprofits don’t understand how to leverage it—they’re still debating governance and ethics, which are important, but they’re even afraid of using ChatGPT to write grant proposals. There’s a real digital divide. The original divide was getting schools connected to the internet. The second was getting people to understand how to use it for education. And now the divide is understanding how to leverage AI.

The policymakers are either oblivious to how the technology works, or they’re reacting by sticking their heads in the sand and abdicating their responsibilities. It’s a very precarious time.

Reactive vs. Proactive Government

Sam Sklar

Governments—especially local governments—seem set up to be reactive. In my career, I’m trying to figure out the best use of the public sector versus the private sector in building physical, digital, and social infrastructure. You’ve said ‘reactive’ twice now. How do we make our policymakers less reactive to what practitioners can see coming and more proactive in building the groundwork so new technology has a place rather than a wall?

Andrew Rasiej

I want to start by defining what I believe civic tech is: any technology for the public good, full stop. Some people refer to GovTech as any technology the government uses to deliver services—but depending on who’s in power, that may or may not be civic. For example, if Palantir builds a massive database for ICE to track undocumented people, I wouldn’t call that civic tech, even if they have a government contract. The real question is: do these technologies shift power from the few to the many, or from the many to the few?

If government doesn’t understand how a technology works, it’s very hard for them to leverage it. City agencies often have data scientists on staff, but most government agencies don’t have a deep bench of technologically capable people. Digitally native young people coming into government are often shocked by what they find—I use the analogy of getting into a 1998 Pontiac with bald tires, a leaky gas tank, bad brakes, and a steering wheel that comes off the column. It would be great to get them to maybe a 2015 Toyota Prius. But they have a long way to go.

And the reasons for that gap aren’t really technological—they’re structural. Things like procurement rules, civil service regulations, hiring practices, and data silos. You can’t directly hire a state-of-the-art data scientist because civil service rules get in the way. So agencies do workarounds—they hire consultants like Accenture, Deloitte, KPMG, or IBM, who charge a fortune. And even when those consultants deliver, they don’t change the culture inside the agencies. They’re like band-aids.

So the big question is: how do you create a culture of innovation inside city government that recognizes that the technologies everywhere in our private lives could be put to use solving problems inside government?

I’ll give you a really great example of civic tech that saved lives, saved money, and transformed the way we deal with mental health. It’s called Crisis Text Line. It was started about ten years ago by a social entrepreneur named Nancy Lublin, who received a text message while president of Do Something from a sixteen-year-old girl being repeatedly raped by her father. Nancy realized that teenagers don’t call 800 numbers—they text. Using a platform called Twilio, she created Crisis Text Line with three major innovations.

First: all the guidance counselors receiving and responding to texts worked distributed, from their homes or offices. No call center needed. Second: when a text came in, the counselor could see the full history of that person’s previous texts in real time, so they could understand the full picture. Third—and most amazingly—after receiving millions of texts, they analyzed the data and could tell policymakers things that were never visible before. For example, eating disorder problems spike for kids returning to school on Mondays after dysfunctional weekends at home. Or that the second-largest cohort of people claiming suicidal ideation is not teenagers, but unemployed white men in the Rust Belt. Or that there’s more sex trafficking of teenagers in Montana per capita than in Florida.

That kind of information was never visible because every city and state had its own isolated outreach system, and they weren’t talking to each other. Crisis Text Line is now a global platform, and cities, states, and countries have shut down their own systems and use it instead. That wasn’t a government service—it came from outside.

New York City, in particular, has a very empathetic muscle of technologists ready, willing, and able to help build solutions to long-standing problems. What has to happen is that the people responsible for those problems have to recognize they have an asset right under their noses. Engaging the civic tech community in government is a really smart thing to do.

I wrote an op-ed back in October, in advance of the mayoral election, advising all the candidates to pay attention to this topic and pointing out that the issues are not really technological—they’re cultural and structural, related to the way the city approaches innovation. The success of any future administration now relies on that perspective.

Autonomous Vehicles and Regulatory Readiness

Sam Sklar

Let’s talk a bit about autonomous vehicles and the future of movement and infrastructure in cities. I went to a great conference at Hunter College last week about the future of AVs. There was a fascinating panel about the regulatory environment around autonomous vehicle introduction—including a former FDNY commissioner who said exactly what you were saying: she doesn’t have technologists on staff who can analyze new technology that intersects with public safety and fire life safety operations. It’s not about the technology itself—it’s about the apparatus that allows smart public servants who want to do good to actually do their jobs, rather than being held back by ossified rules.

So my question is: what are two or three things that, say, a Mike Flynn at the Department of Transportation could do right now to help enable digital transformation at DOT?

Andrew Rasiej

I’m not a transportation expert, and I’d actually recommend you talk to Cordell Schachter, who was CIO of the Department of Transportation under Polly Trottenberg during the de Blasio administration and then went with her to the US Department of Transportation when she became Deputy Secretary. He may have a better specific answer.

But from my perspective: first, collect, understand, and index all of the data at your disposal—what you already collect, what publicly available data could be useful, and how they intersect. Develop a roadmap for how to leverage that data, particularly around AI, to surface solutions that are right in front of you or identify efficiencies you wouldn’t otherwise see.

Second, I would immediately—and I mean immediately—start an upskilling program for everyone who works at the agency, to understand how AI works, how it can be used effectively and safely inside government, and how it applies to their own lives and jobs.

Third, be the poster child for procurement reform in New York City. Procurement is a massively challenging topic—it’s very hard to address holistically. But one agency with a digital vision could build a new model that makes it easier for small and innovative companies to actually sell their solutions to the city. Right now, many don’t even try because the process is so convoluted. Or they have to wait until their third round of funding to hire lobbyists just to get into a meeting. That’s a broken system.

One example of technology the city could be using very effectively right now: LIDAR. It’s not surveillance like cameras—it’s situational awareness technology that tracks movement without identifying individuals. That’s just one example of many that are sitting right there, waiting.

Culture, Communication, and the Story We’re Not Telling

Sam Sklar

One of the big challenges, in my mind, is communication around civic technology. We have a cultural problem in this country—we don’t tell the right stories about the benefits and costs of including technology in civic life, what the private sector does versus what the public sector does. There’s a lot of nuance, a lot of finger-pointing, and nobody wants to be holding the football when things eventually go wrong.

Andrew Rasiej

Government is risk-averse—and that’s to be expected. Government is ultimately a byproduct of a political process, which is highly competitive, driven by money, and as you push further, our democracy has real structural problems. We don’t actually elect our elected officials in a meaningful sense. We vote on Tuesdays, but there’s far too much money in politics, and so on.

The United States was a successful industrial-age organization. We leveraged the industrial age in ways other countries couldn’t. The digital revolution was born here—out of Bell Labs, early IBM, then Silicon Valley. But when you think about something like Eisenhower’s interstate highway system, the political will to build infrastructure was all oriented around an industrial model. We’ve lived off that legacy without reinvesting in maintaining or modernizing it.

It’s infuriating to drive down the highway in New York or New Jersey and see debris at the side of the road—often pebbles and broken cement coming off crumbling bridges and highway dividers—because it’s not being replaced. It’s almost a picture of our broken government: not investing in industrial-age infrastructure. So how do you expect us to invest in digital infrastructure when most people in government don’t understand it, or don’t understand how it can be used to make the world work better?

On autonomous vehicles: even though it’s proven that autonomous cars are safer and kill fewer people than human-driven cars—and if you look at recent estimates, something like 5,000 people a year are killed by truck drivers alone—it would be a safer world if fewer humans were driving vehicles at sixty or seventy miles an hour. But those people have jobs, and we don’t have solutions yet for how to make those people feel useful in society and able to make a living.

The basic contract with the American worker—arrived at in the early 1950s after World War II—was that for forty hours a week of work, you could get a house, a college education for your kids, and a pension. That contract has been broken. And as automation and AI arrive, there’s a real reaction that these technologies are going to take away jobs and leave people with even less than they currently have. I don’t completely disagree with them, because we don’t have the economic, political, social, cultural, or educational institutions capable of dealing with this rate of change.

Where to Start: Public Education and AI Literacy

Sam Sklar

You’ve seen the growth of technology and the civic ability—or inability—to maintain and regulate it. We have private business as essential to economic growth, government with no technological runway to keep up, and a public that doesn’t have the foundation to engage with these questions. What’s step one?

Andrew Rasiej

The United States is an industrial success story, but not necessarily a digital one. There are other countries way ahead of us—I would postulate that China is one of them, with all its complexities and human rights issues acknowledged.

You recently wrote about the Center for Urban Mechanics in Boston, which I think is the best-in-class for thinking about how to apply technology in an urban setting. There are a lot of people talking about smart cities—or as I like to say, “less dumb” cities.

But to answer your question about step one, I’m going to raise the bar. I believe there needs to be a public education campaign about AI. In 2017–2018, the government of Finland, working with the University of Helsinki, published a program called The Elements of AI, with the goal of educating one percent of the Finnish population in the basics of AI. It was so successful that they re-released it in thirty-five languages, and it’s now been distributed to over one hundred and seventy countries.

Research shows that sixty percent of the public is fearful that AI is going to take away their jobs or further erode democracy. AI is viewed very negatively by the majority of the public—and I think rightfully so, because government has largely failed them on this. A handful of states have passed laws to regulate AI. The federal government, particularly under Trump, has abdicated all responsibility. And the public is fearful. But fear is not a substitute for comprehension.

You’re never going to create public policy to leverage technology for the public good, or to limit its damage, unless you engage the public in understanding it. Think of it this way: if there were another pandemic like COVID, the government would issue public education around masking, social distancing, and vaccination. Well, I would say unfettered AI is a digital pandemic. It’s a public safety issue. The government needs to step in and educate the public—not just about data centers eating up electricity or AI taking jobs, but about things like: that TikTok video you’re watching of someone saying Donald Trump is the greatest thing since sliced bread? That was an AI-generated bot, not a human being.

Sam Sklar

Do you think the City could actually do this?

Andrew Rasiej

I do. Besides the op-ed I wrote about the city’s infrastructure last October, I recently penned a piece in the New York Daily News—unfortunately behind a paywall, but I’m happy to send a Google Doc version—basically asking the governor and the mayor to launch a public education campaign around AI. The mayor announced the Office of Mass Engagement. I would say one of their responsibilities should be to educate the public about these technologies. Think of LinkNYC kiosks, buses, and subways with QR codes delivering AI education at a ninth-grade reading level.

In fact, I took The Elements of AI—the open-source curriculum from the Finnish government—and put it into Claude. I asked it to update the content for a US context at a ninth-grade reading level, then fed it a dozen research papers, pro-AI and anti-AI, and asked it to update the curriculum further. It built a seven-chapter course in about ten minutes. Pretty simple language—just enough to help people understand what an LLM is, what an agent does, and why it matters.

By the way, New York City right now does not have a Chief Cybersecurity Officer.

Empathy, Media Literacy, and Listening

Sam Sklar

Before any public education campaign about AI, I think we need an immediate media literacy campaign first. People can’t tell the difference between propaganda, editorial, and reporting. This goes down to eighth-grade civics.

Andrew Rasiej

I couldn’t agree more. Media literacy is job one. But I would argue that before you even get to media literacy, we don’t teach empathy in public schools.

Sam Sklar

You’re in my brain. If I ever have kids, the only thing I care about them learning in the first eight years of their life is how to be empathetic, how to ask a question, how to be in a room with people who are different from them, and how to listen and learn from different viewpoints. The learning follows. There are studies showing that you can learn math a year later and catch up just fine—as long as you have an inquisitive mind shaped by empathy and openness to difference.

But this raises a deeper problem: the “who” matters. The mouthpiece matters. If the messenger is the New York City government, will anyone believe it? What’s the angle?

Andrew Rasiej

When I ran for public advocate, nobody knew what I was talking about with broadband. Now there are people like Alex Borras running for Congress who actually get this stuff. Over time, a younger generation will be in positions of power and may be able to change the dynamics.

But it’s going to take a lot of pain and suffering to get there, because we’ve underinvested in public education for decades—and on purpose. Think about the industrial model: you can have a school on one side of the street and an old-age home on the other side of the street, and the human beings in those two buildings never interact. Because we’ve siloed and factory-modeled every activity of our lives based on a hundred years of industrialization—not just of infrastructure, but of thinking.

Sam Sklar

The young person in me wants to blame whatever zombified version of capitalism we’re in right now. But I think that’s only part of the problem. Markets could work for some of this. But I’m not sure anyone believes the capitalist model being sold to them is working anymore.

Andrew Rasiej

Correct. And the oligarchs building bunkers in Montana know the pitchforks are coming. Here’s what I’d love for people to take away from this conversation: this is a battle between good and evil. The question is whether these technologies shift power from the few to the many, or the many to the few. Right now, what we’re watching happen in real time is the latter.

I don’t believe you can stop this technology. But I do think you can mitigate it or match it—and the only way to do that is to arm the people responsible for good in the world with the same skills and tools that the bad actors have. I still believe that’s possible. I’m still an optimist that we can reboot our society around a more data-driven, responsive government and democracy.

It may not look like the democracy we’ve had for the last hundred years. Maybe it involves new decision-making platforms like Polis or participatory budgeting as a different way of deciding where community resources go. But civic engagement isn’t just about voting or jury duty or joining a community board. What civic engagement should really mean is being present in your community, taking responsibility for yourself and your neighbors. If everyone had that mindset, we wouldn’t have the polarization or the problems we have today.

And the last thing I’ll say: we don’t spend enough time actually listening to each other. I believe that people’s desire to be heard is greater than their desire to be right. If we spend time actually listening—even when we don’t agree, even validating another person’s perspective as crazy as it may be—we create the conditions for finding shared values, shared decisions, shared solutions. But we can’t get there unless we at least acknowledge the other party’s point of view.

Closing: Tech Needs to Be More Civic

Sam Sklar

And that’s the basis for civic tech, right? Listening to our neighbors, understanding what problems actually exist, and building toward solutions together.

Andrew Rasiej

You know, the term ‘tech’ is maybe a little bit of a misnomer. We don’t actually need more civic tech. What we need is for tech to be more civic. That’s not even my quote—it comes from an old friend, Nick Grossman, who used to work for Union Square Ventures and came to my conference many years ago. It’s a great line.