> ## Content Index
> Fetch the complete content index at: https://www.vitalcitynyc.org/llms.txt
> Use this file to discover other available public pages before exploring further.

# AI to Connect Neighbors
- URL: https://www.vitalcitynyc.org/ai-neighbor-connection-nyc-community/
- Published: 2026-11-18T16:49:00.000Z
- Updated: 2026-11-18T17:12:06.000Z
- Description: Most civic technology focuses on making cities more efficient, overlooking the relationships that hold them together.
- Author: Dominic Ramos-Ruiz
- Tags: #ai-and-the-city, Neighborhood Life, Artifical Intelligence, Better Prompts, Better Services

Two tenants spend hours at a public hearing testifying against a proposed rent increase. The City listens to what they have to say, and their testimony enters the public record. By some measures, their engagement counts as a success. But then they walk home without ever realizing they live in the same building.

This missed connection matters because these individuals share much more than a point of view. They share a neighborhood and a desire to stand up for something. Maybe there’s no reason for them to meet. But what becomes possible if they do?

While that question extends far beyond any one department in city government, Mayor Zohran Mamdani’s [Office of Mass Engagement (OME)](https://www.nyc.gov/mayors-office/news/2026/01/executive-order-07?ref=vitalcitynyc.org) offers a starting point. Signed on Mamdani’s second day in office, OME’s [founding executive order](https://www.nyc.gov/mayors-office/news/2026/01/mayor-mamdani-establishes-office-of-mass-engagement-to-transform?ref=vitalcitynyc.org) seeks to strengthen the relationship between residents and city government by bringing more people into its decision-making. As with many public-engagement efforts, however, less attention is paid to the relationships among residents themselves.

### The ties that bind

What we lose when neighbors remain strangers is much more consequential than checking another box. Amid a [citywide loneliness crisis](https://www.nyc.gov/site/doh/health/health-topics/loneliness.page?ref=vitalcitynyc.org), one connection can be the difference between simply living in a place and belonging to it.

Researchers have long tried to understand how these relationships shape communities. Nearly three decades ago, a study of [8,782 residents across 343 Chicago neighborhoods](https://doi.org/10.1126/science.277.5328.918?ref=vitalcitynyc.org) led to the concept of “collective efficacy.” The idea behind it was simple: neighbors who trust one another are more willing to step in when their community needs them. 

Cities should provide the ideal conditions for this kind of cohesion. Yet social media and evolving work patterns have made it easier to live among one another while remaining disconnected.

### Finding each other

While it feels counterintuitive to use technology to address the dilemma it helped create, artificial intelligence has expanded what’s possible. With adoption spreading through civil society, it’s worth asking whether we can design new tools that can help repair our social fabric.

The first step might involve the interactions already happening between government and residents. City Hall may hear people’s concerns individually, but AI could reveal connections between them. For example, someone advocating against a new subway entrance likely doesn’t know an older neighbor stopped taking the train because the stairs became too difficult. Two isolated concerns can now take on shared meaning, making empathy possible even when people disagree. If they’ve both opted in to being connected, a virtual assistant could prompt a conversation, opening the possibility of a solution neither might have reached on their own.

Not every relationship needs to begin this way. It could be as simple as a wayfinding app that points someone toward a nearby community event or mutual aid group that would otherwise be overlooked. A retired nurse might discover a tenant group around the corner is looking for volunteers to check on homebound residents, all because AI surfaced an Instagram post from an organization she had never heard of or thought to look for.

It doesn’t have to become another social network either; [Nextdoor](https://nextdoor.com/?ref=vitalcitynyc.org), for instance, has become as much a hotbed for complaining as a place for genuine interaction. That’s because, like most digital platforms, its business model depends on keeping people on their devices as much as possible. True public-interest technology doesn’t have to play by the same rules. Its success should be measured by how much it actually improves the quality of people’s lives and connections. 

> It turns out being a good neighbor doesn’t require much. Not even friendship. Sometimes familiarity is enough.

### What we leave behind

Some of this begins with governments doing a better job of listening. A few minutes at a public hearing or a survey response can only capture so much, while hours of conversation at a community board meeting can be reduced to meeting minutes. Too much of what people share gets lost along the way.

Large language models can make sense of far more dialogue by recognizing similarities across thousands of accounts even when each person’s contribution is unique. 

Google’s Jigsaw and The Napolitan Institute recently tested this idea with [We the People](https://wethepeople-250.org/?ref=vitalcitynyc.org), an experiment designed to bring the depth of a focus group to a national conversation involving over 1.6 million words. More than 2,400 Americans from every congressional district shared their personal views on freedom and equality, with AI interpreting the themes that emerged. [Nearly 80%](https://jigsaw.google/our-work/scaling-traditional-focus-groups/?ref=vitalcitynyc.org) said the experience helped them better understand differing points of view.

Similar capabilities are now being explored at the neighborhood level in Boston. Together with Dorchester residents, Boston University developed a tool called [On the Porch](https://sites.bu.edu/misi/2026/07/03/on-the-porch/?ref=vitalcitynyc.org) that draws on 311 and 911 data, policy documents, meeting transcripts and community newsletters to answer questions about what’s going on nearby. In an [early demonstration](https://www.linkedin.com/posts/bsurya27%5Faiagent-llmagent-communitytech-activity-7403830446490656768-IpT9?utm%5Fsource=share&utm%5Fmedium=member%5Fios&rcm=ACoAAAsmQY4BfJ7Sj%5FZNzlEFTxTv1Ve6DRji560), the team used it to find streetlight outages and upcoming events, but the larger goal of this project is to empower more people to take action. Whether they ultimately do remains an open question.

Greater potential exists where residents already have a say in public decisions. New York City’s [participatory budgeting process](https://www.nyc.gov/site/civicengagement/our-programs/how-people-money-work.page?ref=vitalcitynyc.org) lets residents propose ideas and vote on how public funds are spent through [Decidim](https://decidim.org/?ref=vitalcitynyc.org), an open-source participatory democracy platform. This year, the City took another step through a [Data Design Sprint](https://www.nyc.gov/site/civicengagement/news/nyc-cec-and-service-design-studio-host-civic-design-forum-20260612.page?ref=vitalcitynyc.org) that brought together 24 agencies to explore how these ideas might inform new programs and services. 

What this still doesn’t capture is the experience, expertise and knowledge each person brings with them. A parent who spends every afternoon at the park may notice something a planner misses. A retired architect may see a solution others don’t. Imagine what the two could accomplish together if they only knew each other. Large language models could surface that knowledge, giving people more ways to collaborate and contribute. Now, these examples all have great implications for government agencies, but they still put people in service of an institution rather than people in service of people. The latter requires a level of understanding that might only be possible through human relationships. 

Much of what allows us to be there for one another never enters an app or formal process at all. The librarian who guided someone to a shelter, the NYCHA leader who knows which resident can translate for the family upstairs or the dry cleaner holding the spare keys for the customer around the corner. That kind of knowledge is earned through trust, not collected as information.

New York’s [Talk to Tenants](https://www.nyc.gov/mayors-office/news/2026/07/mayor-mamdani-launches--talk-to-tenants--campaign-to-build-tenan?ref=vitalcitynyc.org) program acknowledges this. Volunteers knock on doors and talk with tenants about conditions in their buildings, their right to organize and resources that might support them. 

Digital tools could make more of those conversations possible, but what comes next still depends on the neighbors themselves, because neither government nor technology can create the sense of belonging that comes from feeling recognized, seeing familiar faces and becoming someone whose absence is noticed.

### What trust requires

For any tech-enabled solution to work, people have to trust the technology as much as the intentions of those in control of it. That is no easy feat.

A resident who speaks about a rezoning may be comfortable having their comments heard at a public meeting, but that doesn’t mean they want those comments following them. People have a right to determine where their information goes and who gets to see it, and to expect faithful representation of their views. Since large language models can overlook less common perspectives when summarizing input, human judgment is essential to ensuring all viewpoints receive equal consideration.

There are also risks when the technology works as intended. Making it easier to connect with others who share similar concerns can inadvertently create problems of its own, deepening echo chambers and widening polarization. And more people organizing for a perceived common good does not necessarily make a city better at moving forward. For example, the same tool that helps neighbors mobilize for safer streets could also fuel NIMBYism. This makes the choices behind these tools just as important as what the tools themselves can do.

[Nextdoor’s Kindness Reminder](https://about.nextdoor.com/press-releases/nextdoors-transparency-report-highlights-impact-of-ai-led-innovation-maintaining-very-low-levels-of-harmful-content-reported-on-the-platform?ref=vitalcitynyc.org) prompts people to reconsider a potentially harmful comment before posting it, while a newer generative-AI version can suggest a more constructive rewrite. In 2023, [36% of people](https://about.nextdoor.com/press-releases/nextdoors-transparency-report-highlights-impact-of-ai-led-innovation-maintaining-very-low-levels-of-harmful-content-reported-on-the-platform?ref=vitalcitynyc.org) who encountered either version changed or withheld their original message, with the creation of violating content dropping by 15%. A kinder comment has value, but trust still asks more of us. 

After years of research, [New\_Public](https://newpublic.org/?ref=vitalcitynyc.org) came to a similar conclusion with [Roundabout](https://newpublic.org/local?ref=vitalcitynyc.org), a social platform it developed as a public-interest alternative to Nextdoor and neighborhood Facebook groups that allows people to exchange resources and organize around local issues. From the beginning, it saw [paid local stewards](https://www.niemanlab.org/2026/08/can-public-media-build-its-own-social-media-chicago-public-media-teams-up-with-new%5Fpublic-to-try/?ref=vitalcitynyc.org) as essential to creating the healthier digital spaces it had envisioned. Layered on top of the tech, the stewards guide conversations and step in when needed, even as AI helps moderate them. Following pilots in Southeast Portland, Oregon, and Richmond, Virginia, New\_Public recently announced a partnership with [Chicago Public Media](https://www.wbez.org/pressroom/2026/08/13/chicago-public-media-and-roundabout-partner-on-ambitious-network-of-digital-communities-on-chicago-com?ref=vitalcitynyc.org) that is bringing the platform to five local communities. 

Trust, in practice, is built through repeated encounters that let people know who they can count on.

> AI cannot manufacture human connection, but it can give us more reasons to recognize each other.

### Who shows up for whom

Perhaps the real test lies in the moments between these encounters. Did people cross paths off-screen or outside of meetings? Did someone from that event become someone they recognized and supported?

When a blind man had more tennis balls for his guide dog than he could use, he posted an offer on a neighborhood platform called [Hoplr](https://help.hoplr.com/hc/en-us/articles/360027913071-User-reviews-and-experiences-with-Hoplr?ref=vitalcitynyc.org). A nearby teacher took him up on it for her students, and they’ve since stayed in touch. She occasionally runs errands for him, and he’s even agreed to share his story with her classroom. A recent pilot in Austria suggests their experience is far from unique, with [more than a third](https://services.hoplr.com/case-studies/connect-tulln/?ref=vitalcitynyc.org) of survey respondents saying the platform helped them connect with neighbors they hadn’t known before. Hoplr [now uses AI](https://services.hoplr.com/solutions/social-services-and-volunteers/?utm%5Fsource=chatgpt.com) to suggest more matches based on proximity, needs and availability. 

It turns out being a good neighbor doesn’t require much. Not even friendship. Sometimes familiarity is enough. 

A [2024 study by University of Queensland researchers](https://www.tandfonline.com/doi/abs/10.1080/07352166.2024.2407355?ref=vitalcitynyc.org) found that over half of the 1,000 people surveyed would be more likely to help a familiar face in trouble, even if they had never met. 

So the distance between a familiar face and a neighbor could be much smaller than we think. Maybe you come back to another meeting and someone finally remembers your name. A few months later, when something happens on your block, you know exactly who to call. 

There’s something very New York about this. After all, we spend our lives sharing hallways, sidewalks and subway cars with strangers we might benefit from knowing.

AI cannot manufacture human connection, but it can give us more reasons to recognize each other. And over time, maybe we’ll all become a little less anonymous.