AI could cover Lexington’s public meetings someday soon. Should it?

In my two years as LexObserver’s full-time reporter, I spent an almost embarrassing amount of my time attending public meetings.

Monday night, Select Board. Tuesday night, School Committee. Wednesday night, Planning Board. Town Meeting, usually Monday and Wednesday nights, when in session (but certainly not during my first year, when there was that epic five-session Special Town Meeting on gas-powered leaf blowers). And those were just the meetings I attended most often!

Of course, some of those meetings are bi-weekly (while Town Meeting is limited to specific spring and fall intervals), and I did not attend every meeting for every committee. Also, the pandemic made Zoom a default format in most cases, so I attended many through a screen. But the meetings I attended either remotely or in-person added up to dozens, and possibly hundreds of hours, mostly late at night.

So you can imagine my intrigue when, in mid-May, I received a Twitter DM from a reporter who had talked to the founder of a local news site who was working on developing an artificial intelligence model designed specifically to cover public meetings. 

I interviewed that tech-savvy publisher, Mark Talkington, about his work developing this model with his friend and former colleague at Microsoft, Peter Loforte, for a Nieman Lab story. Talkington and Loforte even showed me some examples of what the model could produce. They left me thinking: This technology isn’t perfect, but it is going to be able to do this work. And we need to talk about whether it should.

I don’t have an answer to that question — I think it’s generally my job as a reporter to find and give people accurate information that speaks to a question or issue, so they, and you, can make up your own minds. But thanks to my too-many hours spent in public meetings as LexObserver’s reporter, I have some personal experience that is relevant to this question, and perhaps, specifically, to Lexington.

One of my favorite tweets about being a journalist contrasts some people’s impression of the job (writing, holding power to account, discovering things — seems “cool and glamorous”) with the reality (“crawling on the floor in a dress trying to get a good photo of someone at a 6 pm public listening session in the department of revenue”). I became a reporter because I do love to write, and I love to try to understand the world through writing and listening. But I think that being a good reporter requires a lot of decidedly unglamorous grunt work, like wading through pages and pages of reports and records — and showing up to dozens, or hundreds, of hours of public meetings. 

This might be especially true of local reporting. Attending all of those School Committee, Planning Board and Select Board meetings over many months led me to a lot of the most important (and interesting!) stories I covered. 

Over time, I feel that all of those hours of showing up had another effect — they helped me develop a bigger-picture understanding of how a community functions. The budget affects the teacher’s union; athletic field renovation is connected to the new high school project and environmental concerns; you can’t talk about economic development without talking about housing, and vice versa. Nothing about a town exists in isolation, and understanding that made me a better reporter.

Another thing about public meetings: You don’t always know when a mundane-sounding agenda item is going to be a big deal. For instance, you wouldn’t necessarily expect an item labeled “discuss committee appointments” to draw a comparatively large audience and fraught debate about the volunteerism and civic participation that define Lexington. Similarly, “discussion of additional funding requests under American Rescue Plan Act” sounds pretty dense — if you don’t attend the meeting or read the meeting packet, you wouldn’t realize that item means the police radio communications system is having serious issues. And you might think an annual public budgeting process would be a pro forma affair, not a surprising series of negotiations that reflect both school and municipal challenges marked by last-minute twists just when you think you’re done covering that story. Sometimes, you just have to show up for two hours of boring moments to catch the 10 minutes that tell you the important thing the community needs to know.

It’s also true that when you attend and cover so many meetings, there are tradeoffs. In my two years at LexObserver, I never had enough hours in the day to cover everything I wanted to; had I attended fewer meetings, perhaps I could have completed more of the enterprise stories that languished as half-finished collections of notes and an ever-growing list of bullet points in a google doc (and in a guilty corner of my tired brain).

And there’s also the question of communities without any local news organization at all — an AI bot covering public meetings might be better than nothing. 

But here’s the irony: What really fascinates me about public meetings is they are so…human! There are funny moments, telling looks exchanged, emotional public comments. These aren’t necessarily the most newsworthy aspects of the meetings — but sometimes they’re important, and part of the story. And I think we live in a time when we need to see the human moments within our political conversations more, not less, including on the local level. I suspect that this is not the direction things will go if AI takes over covering public meetings. 

A local reporter should earn trust, and I think, as with pretty much any role, you earn trust in large part by showing up. Some of my favorite parts of my job involved spontaneously running into people at protests and ribbon cuttings and groundbreakings and election results parties, at coffee shops, and yes, at public meetings — especially later into my time at LexObserver, as in-person became more of a norm. And still, in the recurring theme of my tenure, there was never enough time — I could not say yes to every event. But I think the times I showed up, including at long public meetings, perhaps helped build up my credibility and usefulness as a reporter, and certainly led to me receiving follow-up tips and having conversations I wouldn’t have otherwise. Again, I dare say the human reporter still has the edge over AI on this front.

There are different ways to look at what a reporter is supposed to do. Holding people in power accountable, telling your audience what you need to know today, bearing witness, creating a public record that will last. Lately I’ve been thinking about the work as a little bit like the role of Jonas in The Giver — he’s the “Receiver of Memory,” and becomes the sole repository of pain, joy and memory for his community. My immediate instinct in making this comparison is to water it down, to apologize for it and say it’s too grandiose — but I think that’s the wrong instinct when we need to talk more about why reporting, and human reporting, matters, not less. Reporting is emotional work, and at LexObserver, I did feel like it was my job to listen to, and share, frustration and anger and pain and joy and hope of people with sometimes very different views on issues. I wasn’t the “sole repository” of these emotions and perspectives by a long shot, but I was too often the only one writing about many of the stories, and public meetings, I covered. I was not a perfect reporter, but I felt that responsibility acutely, and it always informed my work.

AI will probably creep into public meeting coverage, and reporting writ large, more and more over the next few years. There may be some benefits to that, especially in places that have no reporters at all right now. But it will never feel that responsibility or empathy. And that will be a loss.