Tracking Local Government with LocalLens

This is AI Inside Episode 25, recorded Wednesday, July 10th, 2024. Tracking local government with LocalLens. This episode of AI Inside is made possible by our wonderful patrons at patreon.com/aiinsideshow. If you like what you hear, head on over and support us directly, and thank you for making independent podcasting possible. What's going on, everybody? My name is Jason Howell, and this is another episode of AI Inside, the show where we take a look at the AI that's hiding inside everything. I always feel weird when I say hiding, because it makes it sound kind of devious, and I don't intend for that. But earlier today, there was a Samsung hardware event, and what do you know, AI was sprinkled throughout.

Maybe I should say sprinkled instead. See, I worked that out with you at this workshop. Really great to see you all.

Thank you for joining me on another episode of this show. Jeff Jarvis, obviously not here today. I could say he's on assignment, but that would make it sound like he's on assignment for this show. He's on assignment for his career outside of this show and his life. He's doing important things.

He'll be back next week, but he is one of the reasons why we are doing this particular show this week. I'm going to talk to you about why it's so important. That's coming up here in a moment. Real quick, though, I do want, before we get started, to thank those of you who support us directly each and every week, each and every month, our patrons at patreon.com/aiinsideshow.

You all are enabling this show that we do every week, AI Inside. We couldn't do it without your support, especially right now as we're kind of building, we're growing things. Matt Buoniconti, I hope I pronounced that right, is one of our patrons. We really appreciate patrons like you, Matt, for supporting us, so thank you for doing that. Yeah, you all are awesome. We appreciate you keeping this show going week after week. Also, just real quick before we dive into our interview, I just created a new section of the page. I'm kind of vamping because I was supposed to pull this up before we started the show, but if you go to aiinside.show/survey, then you can fill out a wonderful survey. I know that doesn't sound that exciting, but honestly, it's really helpful because Jeff and I are kind of in the process right now of trying to figure out how we grow the show, how we kind of create opportunities for people who might want to support us on a financial level. In order to do that, they often want to know like, well, who's your audience? Who's watching? Who's listening? So that they know that if they spend their advertising dollars for this podcast, that they're reaching the right people.

I know we've got an amazing audience. I know we've got people who are working in the AI sector, who are working just B2B, businesses that have a lot to benefit from AI. If you can, if you have a few minutes, honestly, it's not a very long survey at all, and you can answer as much or as little as you'd like. It's aiinside.show/survey.

Please go there. I'm going to be doing this for probably the next month in the hopes that we can kind of gather enough information about our listenership so that we can share it with people and start to make a little bit more of a business out of the podcast. So aiinside.show/survey. Appreciate it if you'll help us out and head on over there and give us some information about yourself. We want to know more about you.

And that's all there is to it. Okay. So with that out of the way, and thank you again for listening, let's talk with our guest this week. And I'm super bummed that Jeff can't be here because I know he would have so much to say about this topic. But we're going to talk a little bit about how AI and local government are finding some progress together because I'm sure AI and government as a whole, yes, there's things happening there. But I think more and more, I'm personally exposed to the power and the necessity around being more involved on a local scale when it comes to local municipalities and school districts and all these things that are happening on a local scale. A lot of times we have more ability to get involved and to shape our government on a local scale than we do on the wider nationwide scale.

And so that's why I think today's topic is actually really important. And I want to bring on Matt Post, who is the co-creator of a site called LocalLens. If you go to thelocallens.org, you can take a look at Matt's work.

And it's really, it's a service, and I'll give you a chance in a second to describe it in your words, but my understanding is it's a service that analyzes data around local government meetings of all different types, summarizes them, makes them searchable, all this stuff that could easily be lost in the shuffle because there's so much of this that happens kind of under the radar unless you know what to look for. So anyways, Matt, it's really great to have you here. Thank you for hopping on AI Insight today.

Thanks for having me. Yeah, totally. I'm excited to talk about this topic because like I said, I feel like it's a really important thing. Tell me a little bit about kind of your background and why LocalLens kind of appeared out of that to become something that you've been championing.

Yeah, I think your sort of analysis on the potential of involvement with local government is exactly correct. These are entities, there's about, in our estimation, about 150,000 of them across the country, and they shape everything about our daily lives. The quality of our schools, the quality of our streets, the quality and composition and appearance of our neighborhoods and communities. And because there are so many, and they are so small, and they so rarely hear from people, you can have an enormous impact on your community by getting involved and stepping up.

But it's actually really bizarrely difficult to follow what these local entities are doing. You know, I served, I come to this because I served on a school board here in Montgomery County, Maryland. And, you know, decisions we would make, several billion dollar budget, you know, literally changing, you know, the circumstances of the conditions where kids are learning and that being the future of the community. And yet, you know, the actions that we took, because local news is shrinking, it's really hard to understand what those local government entities are doing. And so it was behind the as a school board member that I got to see, you know, the way that this lack of transparency and public participation really ultimately hurts democratic decision making in a community. Only the people who are, who have the time to sit through a four hour long meeting in the middle of the workday are getting a say over local government. And then I also experienced this sort of on the other side as an advocate working for a big national advocacy group that had interests on issues that local governments cast votes on. And it's hard to follow what your local government is doing in your community.

It's even harder to make searches for what local governments are doing nationally. You know, if there's a particular issue that you're interested in and you are curious, you know, how are school boards around the country dealing with this? How are city councils dealing with this? How are library boards dealing with this? There is currently no way other than LocalLens to conduct those kinds of searches.

So, okay, so LocalLens right now is really a way, a system of collecting the information that's out there somewhere, that data exists somewhere, if you know how to get to it, if you know how to find it. And essentially, like, what is that data? Is that data, like, in the beginning, is it a video recording? Is it an audio recording? Is it a transcription, notes, that sort of stuff? Just like all of the above?

Yes. And I should just say what LocalLens is. Sorry, I maybe should have started with that. But LocalLens is an AI powered database of local government meeting summaries, searchable by all sorts of metadata. And the intention is to make it much, much easier for organizations, researchers and journalists to keep tabs on what local government is doing, both on a local scale and on a national scale. And, you know, every state and Washington, DC requires local governments to publish some recording of what happened at their meeting.

And for a long time, this was mostly just meeting minutes, which, you know, are notes, they vary widely in quality, they're usually published after the meeting. Wow, the website does not look good with dark mode.

Oh, I'm sorry. My apologies. It's a production technique that I do to make the camera look better. So I'll turn it off here. Apologies.

This is the normal look of the site. That's a good flag that we need to add that to the things we need to work on.

I completely forgot that I put on the dark mode. That's totally my bad. No, it does look a lot better like this. It's very colorful. It looks nice.

All my friends make fun of me because I still use Twitter on light mode, which is a hill I will die on.

Well, you and Jeff have a lot in common then. Jeff absolutely detests dark mode.

But so, you know, government entities during COVID made this switch, this pivot, a lot of them to live streaming their meetings because they couldn't actually meet together all at once. And I think that experience showed a lot of local entities. Well, live streaming is actually pretty cheap. It's pretty easy now. And it's a good way of getting potentially more people involved in the community. And so the simultaneous advancement and inexpensiveness of AI and large language models at the same time that suddenly a lot more government entities were publishing recordings of their meetings opened up the door for LocalLens. LocalLens really could only have existed when we started it because of the explosion in this unstructured data that is interesting to some people and the ability to structure it with AI.

Yeah, no kidding, right? That's been such a moment that we've seen in the last three to four years that's really ushered along this competency and comfort and for a while, that total necessity that if we're going to do business, it's got to happen online. And that opens the door for the possibility of recording all these things for sharing at a much wider scale, again, if you know that it's there.

And so, yeah, that's just really powerful. Now, the site itself, so you've got these recordings, let's say, from Zoom, or you've got minutes, it sounds like a multimodal system to a certain degree. It's got to do a lot of things. How did you all approach the technology that's required to handle all of these things? Were you interested in AI prior to this? And did that inform your thinking around the directions that this went? Or was it, I think, that there's a possible ability for AI to streamline, to enable these things to be possible? How do we find the person or the people who can help us build this with the technologies that exist today?

Yeah. So, I mean, I'm coming to this from someone who experienced this problem of lack of local government transparency and was really frustrated by it. And then when LLMs were all of a sudden getting really good and really cheap, immediately saw the connection. But not coming to it as an AI guy looking to use AI to solve a problem, but coming to it as someone who deeply believes that more people understanding what local government is doing is important and needed. And I wanted to build a tool that I would have paid a lot of money for when I was working at this national advocacy group. And I think we have it now, which is really exciting to sort of have that light bulb idea and then to spend a year working towards it. I mean, the central technical challenge, I'll say, is this would be really easy to do with C-SPAN recordings.

Someone probably already has done that because you have one source of government recordings. When you have the 150,000 local government entities that exist in the United States, and they all do stuff differently, they all publish their meetings in different ways with different naming conventions on different platforms.

Different quality, even, I imagine. Yes. You know, even when we're talking about video streaming and stuff, there's probably the municipalities that have a single webcam in a corner. And then there's the ones that have a much broader setup. Yes. How is that interpreted? It's a lot of different angles.

Yeah. And I mean, that's something that also wouldn't have been possible, really, without AI, because at a certain point, you cannot plan for every variation of those characteristics, especially because it's also not static. I mean, local governments are changing the websites that they use every day because probably someone had the domain and then let it expire, and it was on someone else's credit card, and they had to change domains. I mean, local governments can be like 200 citizens. You know, that's who they're governing. So there are just a ton of these bodies, and the central challenge of LocalLens is how do you create a pipeline that works for all of them?

Yeah, no kidding. Yeah. So I mentioned earlier that I have a little bit of a connection to this, and I realize it's not my direct connection, but it's through my wife.

She here in Petaluma for the last three years, actually since I think COVID maybe is when she started, but she started to get really interested in local government and what's going on here in Petaluma, and kind of the realization that, like I was saying earlier, that I think for a lot of people, we see government and we want to make some sort of impact, but sometimes it seems so much bigger. We're like a tiny little drop in the bucket. How much influence can we possibly have other than voting or that sort of thing? But she made the realization not too long ago that, oh, well, actually I can have a major impact if I focus locally. And so now she is part of the Parks and Rec Board and the Bike and Pedestrian Board here. And so as a result, that's connected her to having purpose in local government and being really involved.

And I'm just so impressed by her work and how she's really progressed through it. But also just a really greater understanding of this place that we are so intrinsically connected to. We have kids in schools. And so it's good to have this direct line into the vein of what makes that system work, not just from a, I send my kids to school and all these decisions have already been made. But now she's very aware of these decisions as they're kind of bubbling up and as they're working their way through the system.

And I see a site like this as a really effective way for people to be more aware of that. Because I've seen how many meetings she's gone to and all of the different topics that come up. And I guarantee you, so many people who live here have no idea of the machinations.

They might have a feeling about kind of the outcomes, but the way you get there, they actually have influence if they just knew to be involved, they knew that they could. Yes.

And I think something LocalLens does is just makes it much cheaper to be involved. Sort of our lofty mission is to democratize local democracy. Because before, if you wanted to know what the Parks and Rec board was doing, you had to sit through all of those meetings. And as I'm sure you've experienced now, there are so many meetings of all these tiny boards. And sometimes they're pretty long.

Yes, hours long. And so I think the real utility that LocalLens unlocks is as a metal detector for these meetings. You don't have time to sit through the whole thing, but you can glance over a summary and see, well, was there something relevant to what I care about discussed at this meeting? A policy that they are considering, a broad issue area that is important to me as a citizen or as a statewide advocacy group, as a local business, as a national trade group.

And that function of being able to know, I think is going to unlock a lot more involvement and people pulling these levers of policy that they may not know exist.

Yeah. Yeah, indeed. Indeed. Now, so essentially this has been created within the last year. You said you started this about a year ago? Yeah, about a year ago now. About a year ago. How has it developed from day one? I'm assuming, correct me if I'm wrong, but I would imagine if you're developing this from day one, did you focus on your own local government first as a trial balloon to see how it works? And then you've got the major challenge of scaling it out, of which the United States is a very large place with lots and lots of local governments throughout.

How has it gone for the past year? What is the approach on how you achieve that scale? Because I imagine with all those data sources, that's a challenge in and of itself, being able to get access to all of that and know that you're being comprehensive, I suppose.

Yes, the central challenge. I think we have learned so much over the year that we've been doing this. In terms of where we started, me and my co-founder flipped a coin and we started where he was a school board member in Allendale, New Jersey, and then expanded to a county in New Jersey, and then expanded to all of New Jersey. And now we're in three states, Massachusetts, Florida, and New Jersey, and looking to continue expanding across the country until we've got every single local government entity in the pipeline.

Yeah, yeah. And starting off in the single area that you did, were there any, I don't know, aha moments or moments where you're like, well, wow, I didn't see that coming. It must have been super, super interesting, I have to imagine, when you're creating something and launching it and that's really your trial balloon to see, is this going to work? How was that first experience? What was the major kind of hurdle or anything that you were up against there that helped inform how you could do this better on a wider scale?

Well, I'll say one thing. When we first launched, maybe a week or two in, one of the local government meetings that LocalLens, one, flagged as noteworthy and two, summarized was the school board meeting where, it was just after a graduation ceremony where the valedictorian of the high school spoke about her experiences with racism in the school district. And at the following school board meeting, the school board members were talking about how to censor future graduation speeches. And LocalLens summarized that meeting, noted it as noteworthy. And a reporter who was visiting our site saw this, saw that no journalist had been at the school board meeting. There are so many school boards in New Jersey, it'd be impossible to cover them all.

And so this reporter used LocalLens to investigate and write a story and the backlash at the school board caused them to back down from this censorship. And that was the moment where I realized, I had the sense that this would be useful for me. I would have loved to use this, but it was the sense that this is important and this is driving real government transparency that has not been possible up to now. So that was really the moment where I was all in on this idea.

Oh, that's wonderful. Yeah. And that's so interesting. So essentially what that illustrated is, like you said, I guess rewording what you were saying is that there's so many of these meetings that happen. There's almost certainly something of worth, something worthing of being noted from a journalistic perspective. The real challenge though, is that you couldn't possibly have journalists at every single one of these because there are so many. That doesn't mean that they're any less notable. That doesn't mean that the information that comes out at these meetings that is discussed is any less important that people wouldn't want to hear about it. Of course they would, but a system like this really enables that first line of, not defense, but that first line of awareness that maybe would be similar to if there was, let's say, a journalist sitting in all of these meetings. It's a way to call it out and say, hey, you know what, in this one, these are the important things that could lead to greater awareness, not just for people who are on the site using it, but as a source to pull from for actual journalists who, say, want to write about some of the bigger things that are happening. That's really powerful.

Yes, I think that's exactly right. We're a metal detector for noteworthy developments. I think journalists and local media, because they are shrinking and do not have the resources to be at every meeting, it's exactly as you said. We want to be a resource for them so they know which meetings are actually worth investigating and covering and doing more relevant journalism.

Yeah, that's awesome. All right, I'm going to take a quick break, and then when we come back, we've got a few other questions for you because I love this topic.

That's coming up here in a sec. All right, so we've got more to talk about as far as LocalLens is concerned. I'm not concerned as in what's going on here, but concerned with, in general, AI is not 100% accurate. Its accuracy level is definitely not 100%. There is no such thing as perfect AI.

We've seen it time and time again. It has to be asked around a system like this that's pulling all this information, and not just light information either. When it comes to information, you want to get as much information as possible correct, but you definitely want to get things that involve official acts of the government and that sort of stuff. You want to get that accuracy as high as possible.

How does LocalLens deal with that? Have you seen situations where it has been less than accurate, and how do you work through that to ensure that the system doesn't do that as much?

Yeah, this is the number one question we get, and I think one of the most important. Most common inaccuracy we see in the summaries are misspelled names. It's really hard to transcribe names, and so you can imagine some of those misspelled names make it into the final summary. We started automatically scraping the names of local officials from the low government entity websites, and then correcting the slightly misspelled names, which has helped for regular officials, but people still reach out to us sometimes.

They went and they testified at the meeting, and they're upset that their name was misspelled in the summary, and that's sort of just a very difficult problem to try to solve. And then you get into sort of the more substantial errors, and we have not experienced the kind of wholesale hallucinations that a lot of people think of when they think of AI inaccuracy and misinformation. I think this is probably mostly because we are just using large language models to summarize large, unstructured amounts of information, and so what the AI can draw from is sort of confined, and it doesn't really have to make anything up because all the information is there. I tend to be pretty skeptical of AI uses, honestly, that go beyond summarization and structuring of data. I think sort of the dream of AI as a search engine or an answer machine is not... How do I put this tactfully? I think it's overly optimistic on where AI is right now, and is almost an irresponsible use of the technology. It's certainly not what it's best at. We think it's best at the summarization stuff that we do, or making very small judgment calls at discrete points in a technical pipeline.

So we have not really experienced hallucinations. There are rarely inaccuracies in terms of who said what. Transcripts can be messy if a meeting is rambunctious, which happens sometimes, certainly has happened in the past few months around ceasefire resolutions with Israel-Palestine, and people are talking out of order, people are interrupting.

Most times what the summaries end up being is really vague, which is ideal because the AI doesn't know who's speaking, but sometimes it'll say something the public said, shouted out, was really something that a council member said, or vice versa. Those errors are really bad. I think they threaten the reliability of a service like LocalLens, and so we are always trying to figure out new ways to prevent them. On every page of LocalLens, we link to the recording, to the transcript, and prominently say, one, this is a starting point for research, and before doing anything with anything you read here, you should make sure and double-check it against the transcript you're recording. But also, if you see an inaccuracy, let us know, and we'll immediately correct it.

I think our average time to do a correction is under an hour, so we really think that the information is accurate. But broadly, I think of LocalLens as, like I said, a starting point for research, and it's not human-vetted news, and so I think it still serves a really important function, but like you said, it suffers from all the problems that AI suffers from, ultimately.

Yeah, well, I mean, at the end of the day, and we definitely talk about it on this show quite a bit, is it depends on whether a user's belief on what the AI-generated content is good for really determines the responsibility that they take. If everyone that was presented with information that was the results summarized from a number of different places by an AI, if everyone kind of took that information and by default thought the way that you're talking, which is, here's some information, it's meant as a starting point, this is not meant to be the single source of truth that you 100% go with for whatever you're using this information for, but it gets you started, or brings you to the right place so that you can dig in deeper and go further, that would be amazing. And fortunately, some people just kind of go with what they see, and so I think that it sounds like the approach that you're taking is the right approach.

Here's the information, but do know that this is not the be-all end-all of the information, and it is your responsibility as a visitor to go deeper on this and just realize that this is what brought you through the door, but it's still your responsibility to go deeper on that. I think that's the right approach.

Yeah, we try to be responsible. We don't want to be contributing to the thoughts of misinformation out there, or as Jeff has written about, just sort of the sludge, you know, just content for the sake of content. Well, yeah, absolutely.

And I should also point out that is a real challenge when we're talking about AI-generated content online. And I don't know, how have you managed that? Because I've read through some of the content on the site, and I feel like it's done really well. Sometimes I see AI-generated content, and even though it's grammatically correct and all these things, you just get that sense. You're like, I don't know, there's something going on here. I didn't really get that spidey sense from what I was seeing, LocalLens anyways.

That's such a great compliment. Thank you. Because I personally sort of hate the way that AI writes, oftentimes. Maybe it's from just having read too much of it. I mean, we've done so much testing now. I can, you know, all the major LLMs, I can tell you all the quirks, like a taste test. You could present content to me, and I could tell you which model generated it. So we've done a lot of fine-tuning, both on the model end and that sort of sanitization on the backend. Because there are a few things that models suffer from, like use of cliches, use of sort of like the same weird words, use of exaggeratory language, use of subjective judgments, a bias towards positivity that I think is the result of like some of the alignment work.

And we put a lot of thought into how can we, you know, use fine-tuning and sanitization to reduce that stuff that is fluff and not useful for the user in keeping tabs on local government. So it's cool that you, also having looked at a lot of AI stuff, see that. That's great.

Yeah. Yeah. No, I was impressed by that. And, you know, I'm happy to see that because I think it's really important work that you guys are doing. What kind of approach, well, let me rephrase this. How could the approach that you're taking with LocalLens impact other facets of government, you know, maybe from a wider scale? Is there, I mean, you know, granted, you're obviously, you know, focused on growing LocalLens as the local kind of, you know, the local solution. But how could what you're doing, whether you are the ones to do it or someone else, like, how could you see the lessons that you're learning with LocalLens be applicable maybe at a wider scale for government?

Yeah. I mean, I think ultimately what LocalLens is trying to do is something that has already been done at the federal level, which is, you know, make everything that's happening searchable, easy to find, easy to understand. And, you know, at the national level, there's a lot of money in it. And so that work has been done. But now that we have these AI tools that are really good at structuring unstructured data, which is all of these like sort of meetings and behind the scenes government work is, you know, there are a lot of other aspects of government that are opaque and that ultimately hurt local participation in decision making. And, you know, like RFP processes, contracts, bidding is one. Someone may already be on that because there is probably money in that. But, you know, wherever there is a lot of information and it's decentralized and unstructured, I think those are great AI opportunities for democratization of access to that information.

Yeah, yeah, indeed, indeed. Now, I know you mentioned a little bit earlier that kind of the summarization aspect of AI is what you see, and correct me if I'm getting your words wrong, as kind of the real kind of benefit of AI's involvement with the data that you're talking about. But AI, you know, can do a lot of other things, arguably some better than others. Are there other strengths of AI that you could see tapping into in the future?

Anything kind of on the, you know, within the lens of LocalLens, no pun intended, that might actually be useful? Other ways that AI is used and other facets that could be applied here?

Yeah, I think something that I'm really interested in is as the multimodal aspects of AI get cheaper, the ability to process a video by both looking at frames and by looking at the transcript simultaneously is really interesting to me. Because it would solve a lot of the problems of, you know, potential inaccuracies in transcripts or potential ambiguities.

If the AI could look, okay, this person is talking, I know from a knowledge bank who this person is, and I'm going to use the transcript to understand what they said or, you know, transcribe it.

That's going to tighten up the accuracy even more, yeah.

Yes, so I think those opportunities are great. And I also think, you know, in addition to summarization, like I said, you know, when you have vast amounts of data that exists in totally different forms, having, and you don't have like the human bandwidth to review every instance of this, you know, having AI look at it and say, this is A, B, C, or D is, you know, not really prone to inaccuracy and hugely useful on a fundamental level. So those kind of like discrete decisions that used to have to be done by a human, I think are going to unlock a lot of potential.

Yeah, it's interesting. Just earlier today, you know, there was a Samsung event and the word that came up a handful of times was multimodal. And it just kind of made me realize, like, you know, AI is still the buzzword. AI is still the thing that everyone is realizing, recognizing, wanting to figure out some way, shape, or form to have it as part of what they're doing. And in many cases, like LocalLens, it's the right tool for the job. Last year, the like sub buzzword was LLM.

This year, it's definitely multimodal. You know what I mean? That's interesting. Now, suddenly, everybody is looking at multimodal and being like, well, but if we could just use that, that's going to enable a lot more. That's kind of the cool thing about this technology and how it's evolving right now. It keeps opening up. Yes.

Yes. I mean, something that I'm sure you have experienced this more than anybody, but it's the most frustrating part of doing this work, honestly, is like, I go and I talk to other people who are in AI, or I, you know, I go to these conferences or presentations, and I feel like at least 50% of the people talking are just saying a lot of bullshit or bullshitting or just using buzzwords, which is really frustrating, because I think it detracts from the people and projects and initiatives that are using this new technology in, you know, responsible and extremely cool ways. So, I don't know, I say jokingly, I think I'm an AI skeptic, because, you know, you just, especially when people who have sort of been obsessed with like other fads just immediately pivot to AI. Oh, yes. There's a lot of that going on right now. It hurts us who are doing real work with this tech.

Well, indeed, you are doing real work, and I really do appreciate the work that you're doing with LocalLens. I think it's a really cool solution. I think the bummer is that it's not at a wider scale, and I know that's a real challenge, I imagine, for your team. Just to kind of round things out, like, what are your plans for increasing that scale as you go into the next couple of years, I suppose?

Yes, well, I'll put this pitch here. I think, you know, LocalLens is a project, a small project. If there are any organizations out there who care about low government transparency or understand sort of the upsides to empowering businesses and advocacy organizations to following low government, we would love to be part of a bigger organization and to find a home for LocalLens, any organization that is sort of aligned with our mission and vision of universal government transparency. So, I mean, I think the resource constraint is the main barrier right now, not that it's very high. We've just been bootstrapping it the whole time. So, we've been thinking about trying to find a home for LocalLens to continue growing.

Well, that's important. So, yeah. So, if people want to reach out to you on that, or just in general, if people have any questions or thoughts or anything like that, how do you prefer people get in touch with you? Would you follow you on socials or how would you like that?

Yeah, just shoot us an email, hello at thelocallens.org. We also have a contact forum on our website. And if you live in Massachusetts, New Jersey, or Florida, we have issue alerts as well.

If there's a particular issue that you're interested in, and you're curious if any local governments are talking about, you can sign up for email alerts to know when those issues are discussed, or just get a digest of all the local government meetings in your town or state or county. So, stay in touch. And love any feedback that anybody has, too. We're always trying to improve.

Yeah. Well, I love what you're doing. I'm so happy that Jeff put LocalLens on my radar, because I think it's solving a real problem. And it came at just the right time in my life, because like I said, my wife has been so involved on a local government scale. So, it's like I have a new kind of appreciation for the challenges that something like LocalLens is aiming to help solve, or at least alleviate a little bit.

So, I love what you're doing. Thank your wife for her service. Oh, absolutely. Those positions are, I can tell you, haven't been on a school board, are thankless, but so important.

Super important. And you don't make any money. You're there for the love of the town. And that's certainly the case with Stacey, for her love of the town of Petaluma, but also just interest in what's going on and wanting to stay connected to the information. As we've talked about, it's just really important. It's really important if you want to make a change in your government. Start local. You've got a lot of possibilities there, as far as that's concerned. And if you are in any of the municipalities that we've talked about that LocalLens services, at least right now, then you should definitely check it out.

TheLocalLens.org. Great stuff. Thank you again, Matt, for coming on today to talk a little bit about it.

Matt Post, TheLocalLens.org. Appreciate your time. Thank you for being with us. For sure.

Thanks for having me. Yeah, absolutely. And we'll check in on you again down the line, see how things are going for you. Thank you, Matt. All right. And that is the end of this episode of AI Inside. Thanks once again to Matt Post for sharing his thoughts on the development behind The LocalLens website, or TheLocalLens.org.

So again, go check it out. I know Jeff wished he could be here, so we'll have to have Matt back again so that they can connect, because this is very near and dear to Jeff's heart, as anyone who watches and listens to this show already knows. AI Inside records live, normally every Wednesday at 11 a.m. Pacific, 2 p.m. Eastern. This week, we did do it one hour later, so sometimes the recording time does shuffle around. But if you just follow the Techsploder YouTube channel, youtube.com slash at Techsploder, that's T-E-C-H-S-P-L-O-D-E-R, by the way. We were talking earlier about AI transcriptions getting things wrong. It never gets Techsploder right.

It thinks I'm saying like Techsploder or something like that. So anyways, at Techsploder on YouTube. And you can actually, if you're subscribed there, then you'll be notified when we have live recordings like this, regardless of the time. We do publish this show to the podcast feed later in the afternoon regardless. So if you're subscribed, you're going to get the podcast delivered to you like the magic of podcasting that it is.

Like, rate, review, subscribe, wherever you happen to listen. It really does help this show out, I'm telling you. Thank you so much for doing that. And of course, support us directly at our Patreon. That's patreon.com/aiinsideshow. We offer ad-free shows, early access to videos, discord community, regular hangouts with me and Jeff and the rest of the community. We also offer the ability for you to be an executive producer of the show. And this week's executive producers, DrDew, Jeffrey Marriccini, WPVM 103.7 in Asheville, North Carolina.

That's right. A radio station. I've been chatting with them.

It's pretty cool. Anyways, thank you all for your support. Couldn't do the show without you. And everything that you need to know beyond all of that can be found at the website, AIinside.show. Just one last reminder, aiinside.show/survey.

Fill out the survey. Appreciate you all so much. We'll see you next time on another episode of AI Inside. Bye, everybody.