Greetings and welcome back to ZacCast your official podcast for local government nerdery. I'm Chad, that's Pat, and today we're back to the legislative session that's going on right now. Pat's got some updates. This is not something that I can keep on like up with, so I rely on Patrick to like feed me and he's going to feed you as well. And then we might talk a little bit about some some AI machine learning stuff that we're working on. Just. Let me nerd out for a little bit if we have some time so Patrick. Let's start by our normal method of starting, which is where I just ask you how you're doing.

I'm good man, I'm good. How are you doing?

Pretty good. You've been traveling a bit lately, huh?

been on the road. I was in Scottsdale last week, right? And for, we've got the team in St. George, Utah for the Utah management association and like a mid-year get together, I think of like the municipal league there. We were at the Alliance for Innovation Conference. If you don't know anything about AFI, pretty cool organization. I would say it's like a case study conference where

We got the team in St. George right now too, don't we?

different cities come in and kind of talk about some of the cool things and cutting edge things that they're doing. So got to see quite a few of our clients present, which was cool, presenting some really cool things. A couple of them using data from our system to do that. I would say taking, using our system to clean data so that they can use the data in a different way, which was kind of cool to see. So I really liked that.

Give you any ideas for things we can incorporate or or no

Yeah, I mean, you know, like, know, city of Fort Worth, did a presentation specifically on how they are analyzing their future bond for 2026. and. You know, it was more like a one-off project, but it was pretty cool. Like, you know, they traditionally you and I are ex Fort Worth employees. So we know that traditionally the bond process in Fort Worth, is, basically each department kind of makes their list of projects. And then you have like this. very heavy political process inside of City Hall to try to get your projects to the top of the list so that you can get to whatever bond amount is going to go to the voters. I think in this case, was like 800 million. can't remember the exact number, but I think they were looking at $800 million in bond projects, a GO that's going to go to the voters. we traditionally would work with departments and so like

you

Harks would come in here and be like, well, these are our top five priorities. And they would try to make their priorities more important than maybe transportation's priorities, so forth and so on. It was kind of cool to see how that process has changed using data and data analytics and where those projects are going because of where data is driving it. So there was this really cool chart that they showed in the presentation where they basically like use the old process.

you

and then use the new process. And they showed the difference between the rankings, between the old process and the new process. And they've got a lot of department buy-in, but at the same time, their priorities in the analytics process were significantly different than maybe the political priorities of each individual department. So that was fun to see. So this is kind of the departmental process and they're going to put these priorities forward and they still have to go through the city council process where council is going to look at those priorities and say, okay, yeah, we think there needs to be some adjustment here. Um, you know, for lack of a better term, it's all kind of an algorithm of how they put that together. And they've kind of gotten all the parties to agree to what that algorithm should look like. And, uh, so they have like a starting point and now they're going to go through their council process. So I thought that was pretty cool. That was a, that was a really cool project. I saw, um, the use of camera systems on trash trucks for code enforcement, not necessarily to automate the code enforcement, but to maybe, uh, better direct code enforcement officers on where they should go.

you

I thought that was a very interesting project. think the libertarian in you or maybe me a little bit was a little, I don't know, the use of cameras driving down a street on a city vehicle and that being public record to me was just, was a little scary and how they were doing that. think our rules, it was a city outside of the state of Texas. I think our rules in Texas are a little different on how we have to hold onto that material. But that was kind of cool. the use of AI to identify high grass or graffiti for a graffiti abatement and, you know, different types of things there. So that's kind of the cool thing of AFI though. It's like all these cities are doing these things. And it's a good spot for you to kind of go for two to three days and just listen to all of these different case studies and see what's happening in other areas and what innovation is happening. So very different than any other conference somebody from a city goes to, which I thought was cool. So. Yeah, that's kind of my roadshow lately.

So they're using like computer vision to check for, are they using a language model? Like just sending pictures to like chat GPT and saying, is this grass high? Or are they actually using like a custom computer vision model that can pick that stuff out?

Ahem. Yes. So they're using a company that has built a custom model for this purpose specifically for code enforcement, like crime prevention, things like that. You know, and they mentioned it, but like Axon, you know, the Taser company, the body cam company, Axon, yeah. They have also developed in their most recent updated body cam system, they have automated police reporting.

Axon.

And they're kind of like Zoom has automated Zoom meetings, right? So they kind of discuss that. It's a very similar software to what they're using to do that. Obviously there's a lot of AI in the background kind of doing some things there, but yeah, there was a department that kind of spoke on their police department using that and how much time that's saving police officers and their report writing. There's still a lot to be seen on that. Like what is the accuracy of that system versus what actually occurred in the video, so forth and so on, you know. problem with language models is that they can't guess at what's next. Sometimes they guess wrong, right? So I would worry about that. We, you know, there were some conversations there, but, yeah, I mean, it was, it, that was kind of a cool look, but yeah, they're doing the same thing in the code enforcement side that they're doing to the body cam police body cam footage side.

feel like if you put cameras on the trash trucks, which are driving through the city at least once a week, if you also put those like LIDAR or radar systems that did the pavement conditions, then you could have like, you could like fully automate so much of your, you know, like the inputs that go into some of those departments. Like, there's a new pothole here.

Ahem.

at this particular place. Go fix it.

Yeah, I think when I drive through a city sometimes, I judge a city based on its cover, like graffiti, trash, how neatly kept medians and right-of-ways are, things like that. there's some people in the career field that are very gut driven for that. We need to focus in this area because we're seeing the certain, it's the broken glass theory, right? or shattered windows, broken windows. It's one of those things. Yeah. Broken windows. There we go. One of those combinations that we learned in grad school. This obviously is automating those inputs, right? So it's not like a trained city official who's driving around seeing things that need to change because there's something that's going to happen or decline because of it. So I, yeah, I found it. I found it really interesting. conference in general, but it's kind of like that. It's just a lot of like thought.

Broken windows, yeah. One of those combinations of words.

inducing thought full just, you know, kind of wow moments like, okay, that, that makes sense. I did, I did find myself on multiple occasions though, kind of going back to Chad circa Hudson Oaks with the chart on the desk about like, what is efficient and what is not. Like it's, you know, some of these cities are spending a lot of time to automate things that maybe are quicker, not automated. So I did question some of that. don't want to call those presentations out, but I, I thought that, I thought those were pretty interesting. but you know, it is cool to see that culturally things are moving the way that we would love to see them move just from a standpoint of if, if you can have high level support and you can have creativity.

you

you can get a lot of efficiency just out of the creative side. And to see governments, specifically local governments and city manager form of governments move away from because this is the way we've always done it and do it very willingly is pretty cool to see. that's the coolest part about the AFI conference to me is it's like a room full of people who don't really care about how we did it in the past. That doesn't, that's not a confining variable.

you Mm-hmm.

like it is in most communities. Right. So, we're at a inflection point of change. And I think some of these cities stepping up and understanding that that change can be good for them as, as, yeah, it's, it's cool to see. So,

Cool, man. Hopefully I'll get to go to one of those one day.

Yeah, I mean, I think I don't know where AFI is next year. They also do some, they technically call it the TLG conference. If you're like looking up the, the AFI, but that one's the TLG, which is their bigger conference. They, they do do some like smaller conferences as well. And then ironically, we did the UNT MPA conference on Friday of last week. And the director of AFI, Troy Riggs was the, was the keynote speaker. So, so literally Troy and I were in.

Kudos.

Arizona together and flew back and we're right back into the UNT conference, which I just have to say, so good to see everybody at the UNT conference. Such a cool environment. We brought an outside vendor with us this time, somebody that we've gotten to know in the California market, just to kind of get a feel for, okay, this is what we do. And this is what we financially support for folks that don't know. Most of our clients know, cause we talk about it, but for folks that don't know, like Zach in general, financially

you

supports two things, the growth of their career, like MPA programs specifically, and we love assistants' organizations. And so if you're one of those reach out to us, it doesn't have to be UNT. It doesn't have to be our local assistants' organization. We try to support all of those folks. That's where we feel like our money should be given so that we can grow.

you

the actual career itself because for two reasons, one, we love the career. And the second one is just straight up selfish. If we don't have really great smart people, they can't use our products. So, or they won't use them. Right. And so we need to continue to encourage the growth and development of the professional sector within city government, city management. So really cool to see all those people in the room and to bring that vendor and for them to be able to see the culture that's been built.

Or they won't use them,

in Texas within the city management community. think you would see the same thing at UTA. You'd see it at UTD. Obviously you'd see it at Texas Tech or even, you know, my undergrad at A or U of H. Like we have some really great MPA programs in Texas. Did I miss one? Sorry, what was that? Yeah. So, but yeah, so cool conference, man. It was, it was a lot of fun. You want to jump into legislation? So it, it is a

or. Let's dive in.

Officially Wednesday April 16th at 10 14th the Texas house has just opened on the floor at 10 a.m. And they are slackers they are debating vouchers today I bring this up for a couple different reasons Educational vouchers is the big political issue in Texas. It's also what will possibly stop almost all other legislation so

slackers man.

We don't know what this fight's gonna be like today. The bill has already passed the Senate. It's gotten through committee in the House. There's, I would say until about three or four weeks ago when polling data started to show that vouchers were losing momentum and losing steam. I think everybody assumed that vouchers were gonna pass and it was gonna get done as an emergency priority and then it was gonna move on. It's taken a little while to get where we are right now. The voucher bill has not changed from the Senate version that we kind of talked about early, early on, or maybe like right before the session started. It is a bill where you would be able to take a kid to private school. They're going to give you $10,000 to take and go there, which is. Interesting because we typically on average fund public schools at about $6,500 per student depending on what district you're in it can ebb and flow. So it would actually give a little bit more money to the private school side. The proponents of the bill would tell you that's because school districts have a tax on INS for facilities and so that's the difference is that you know they're making that money on on bonds for buildings and schools and so forth and so on and so private schools deserve a little bit more money because The other side of the equation would say that's foolish. They should be able to make it on what they make it. It is what it is. So, look.

curious if the average INS rate is 50 % of the O &M rate.

Yeah, I don't know. I doubt it. Right. Because like the INS rate of school districts is usually like 90 cents. It's pretty high. let's let's be or sorry, the M&O rate is usually 90 cents. Let me correct that. And let's be very clear. All M&O taxes for school districts across the state are filtered up through Austin before they come back down as a per student fee. So there's very little control at the local district to do anything with an M&O rate. But there is full control of an INS rate. based on voter approval and bond debt. And pretty much everything a school district does from a debt standpoint has to be taken out to the voters. So the voters have approved all of those INS dollars for the local school district. All that being said, I've been very outspoken. If you follow me on Facebook or LinkedIn or anything like that, I've been very outspoken that this is a fiscally irresponsible bill that we won't be able to afford. We can't educate, you know, 1 % of the population with 10 % of the cost of public ed. like you just, you know, the numbers don't make sense. And so there's a lot of questions about what the cost of this program is going to be. And there's going to be quite a few amendments that get introduced on the house floor is the thought that's going to be there. The biggest amendment that's going to be argued. And I think it's important to note the quorum report, Texas Tribune, there's been lots of articles about this, that the governor was going to meet with the house early this morning. to basically tell them that it would be unconstitutional for them to call an election on vouchers because that's, that is an amendment that has the support of 85 representatives, which would be enough to pass in the house. And the thought process is, that, we shouldn't do this unilaterally. The polling numbers are not great. Let's go ahead and put this on a, on a ballot measure. The interesting thing is, is nationally a voucher ballot measure has never passed. And so obviously proponents of vouchers don't want it to go to the voters. I find this to be like a bad umpire, right? Like an inconsistent umpire because the Texas legislature right now is really pushing for all debt at cities need to go to the voters. Any type of tax increase, regardless of the three and a half percent revenue cap should go to voters. but when we're gonna go spend a billion dollars this year and $11 billion over the next four years with this new program, yeah, we just don't really need to go to the voters like that. And I think it doesn't surprise me, but I just wish that our politics were more consistent. I feel like when you and I were in our undergrad and in graduate school and significantly more politically.

Does that surprise you?

than we probably are today in local campaigns and things like that. I feel like we had a more logical consistency about how we felt, right? I feel like populism injected into local politics has really created a level of inconsistency that would get any MLB umpire fired for balls and strike calls. And so, you know, it doesn't surprise me because of where we are from a national political standpoint.

you

but it kinda is, you know, we are where we are. Does it surprise you?

No. No, the I mean, the only thing that's consistent is the inconsistency. So I mean, it's a it's a very it's a very like poignant argument like, hey, guys, you're saying this over here and you're saying this opposite thing over here, but it doesn't matter. They'll just do what they want to do. But so like bring this back to, I guess, the city side. What like

Okay. Yeah, so I

Why are

think the important...

we talking about this so much other than it's just like the biggest issue that's facing Texas politics right now?

Yeah, we're talking about this specifically because it could be such a fight that it basically grinds this, the legislature to a halt. It crowds everything else out. becomes a big argument. And a lot of the local bills, one of the threats the governor has made is that anybody who signs on to this, take it to the voters approach. He is going to veto all of their local bills, right? So any type of, you know, special bill that you're trying to get past, whatever that may be, he's made a pretty

that crowds everything else out.

a blatant threat so far so that his, I would say, chief of staff over the voucher issue actually said we've actually, he was caught on a hot mic saying, maybe we've lost all of our like, gubernatorial decorum. Right? We've gotten to a point now where really, really, really conservative legislators that believe in the will of the people.

Wow.

are going to be penalized because they believe in the will of people. Right. And so the reason we're talking about it is, is because the injection of funding and money that's come into politics, it's about $12 million of money that came into this politically to get this to become the forefront of Texas politics in this session. And two, it's going to have an extreme impact on what happens from here on out. So I want to talk about a few other bills that are kind of on the movement, but understanding that vouchers could really derail a lot. On the back half of that equation, the Democrats have said, if this bill does not get the amendment to take it to the voters, then they will not vote on any constitutional amendments, which is going to really create some issues because there's some really big Republican reforms that are very popular with the voters. that have to go to constitutional amendment, you know, more stringent bail reform measures, criminal penalty measures, a lot of things in the criminal justice side that are required to go through constitutional amendment, go to the voters. They need the votes to get there. Right. And so and you need you need those 85 votes. They can't get there by themselves with just Republicans. And so if the Democrats do hold out on that, there was a letter signed by 50 of them, which is basically going to get them 50 to 51, which is going to get them to the point where the Republicans are not going to be able to get those things on a ballot. Um, so, and the lieutenant governor has popped back and said, we'll just call special legislative sessions, by the way, not his thing. He doesn't get to call those. That's the governor's office. Right. So, um, yeah, so I mean, I think we, you know, in, Texas political theater, we have another historical moment in time. Right. So that's, that's kind of where we are. I think that's why it's important to talk about the voucher issue. Um, it's also important to talk about from, from a state standpoint, we're putting ourselves in a, in a fiscal situation where the state's going to have to start asking for money. I posted this actually to our Zach.

Thanks.

LinkedIn page, but the comptroller posted again, another, I'd say, veil warning. It was a little bit more direct than it has been in the past, but he posted another veil warning about we're basically squandering our surpluses and their fiscal forecast, I'm sure are starting to look bleak just like ours are. And they are saying, you know, we need to be very careful about what we do. So using

Mm-hmm.

the surpluses right now in constitutional amendments to reduce property taxes on school &O taxes and locking yourself into that may not be the smartest thing to do in the world because you can't continually pay for that down the road. And so as the state's revenues continue to get tighter, they're going to start looking for money. And the only place to start looking for money is going to be in local revenues. So it's going to be really interesting. They're going to do, you know, are they going to do some of the similar things that they did to school districts, to cities at some point? basically just roll up funds there as well. with that vision, not trying to scare anybody. just, you know, I'm saying four to six years down the road, there'll be some really tight finances in the state of Texas. And we probably need to start looking at some of those issues. I'm going to roll through a couple of things and talk about significant floor actions that were taken. There are some bills that have passed that I think are important. And so The Mays Middleton bill, SB 414 passed the Senate and has moved on to the House. Specifically, that is the language that requires any debt obligation proposition language to include both the estimated principal and interest amounts. So instead of saying we're spending $80 million, you would have to say over a 30 year period, whatever it is, yeah. So over the 30 year period for the bond proposition.

We're spending a hundred and seventy million.

That passed the Senate hasn't hasn't gotten through the House yet, but it did pass. Also passing the Senate is Senator Kolkost. She her preempts the city's ability to obtain a license or permit of cottage food productions. I just thought this was funny because what cities are in the cottage food business. I would love if it's one of our clients to reach out. I'd love to know why you're in the cottage foods business. So I just thought that was goofy. Here's one. For your dear heart, Senator Hughes SB 673 allows an accessory dwelling unit in a single family zoning or unzoned areas by right and prohibits a city's ability to regulate an ADU in certain ways. This bill applies to all cities. It passed the Senate. It's headed to the House. Chad's actually a big fan of this bill. I'm a little more restrictive in it, but it's going to be really interesting to see where that one lands. So.

So my mom actually texted me about this bill the other day.

huh, yes.

She was like, can you just believe this? So I spent a little while trying to YIMBY -ize my mom, but it didn't work. Yeah, she had some comments about like, there's no room. won't have like if every of all my neighbors put an ADU in there, I won't have any privacy and things like this. And I'm like, well, like my office is basically an ADU. Like it's not that, not that bad. And I was like, also you're getting up there in years. Lots of older people build these things so that they can either move into them and then rent out the house or rent them out in retirement for some extra money. Lots of people build them for their, I mean, they're called mother-in-law suites, right? Lots of people build them so that their, their in-laws or parents can come and live with them as they get older. So like there's lots of good reasons. and good social benefits for this, not to mention it adds housing. But her final comment was, well, where will they park? Because everything comes back down to parking, Anyway, that was a fun little conversation with my mom. I don't do the social media politicking anymore. I learned my lesson about 15 years ago that it doesn't really do any good. So I keep my discussions.

Yeah.

little more isolated, one-on-one. They're still as effective though at changing minds.

Yeah, so I mean, you know, I think that's going to be a highly debated issue, but I mean, I think that's going, that's going down when I say highly debated between city officials, but it's going down the same pathway as being able to, you know, taking away the regulation of building materials, right? I think that's one of those one, one of those issues that could become very interesting. I question whether it's an actual issue or concern like does this pop up a lot of ADUs in areas. Obviously college towns are very against this because it's going to create significant density because that's a really easy market to make money on ADUs. But, know, I, to be fair, I have an ADU in my backyard. I don't rent it, but I do have an ADU in the city that I live in and managed allows for ADUs of related individuals, but not for rental. So, So I think that's going to be an interesting change.

So, since the first ADU reform bill was passed in 2016 in California, over 80,000 ADUs have been permitted. So that's what, 10,000 a year? But that's also a state that has 40 million people in it and a significant housing problem.

Mm-hmm. That is a lot. Yeah. Right. How many people does Texas have? Okay. So yeah, significant housing, but, Texas is being squeezed housing as well. I mean, we sprawl don't get me wrong. And it is easier to develop in Texas than it is in California. But, um, you know, there, are lots of podcasts out there for people to listen about how the ADU stuff has gone on in California. There's lots of conversations to be had there if you're curious. So, but once again, you know, my statement stands true.

30 million?

It just seems like we're a few years behind California. So we don't like to say it, but we do get a lot of ideas from their legislature on what we do in Texas, even though they're different parties. So it's something that we're looking at here in Texas. And I'm actually, I think this is the one that is going to pass. I think it's going to get to the house pretty easily. So I think we're going to see that one come through. The other one is SB 1209 passed the Senate, which eliminates the May uniform election date. So the typical local government, some cities are, or have moved to November already, but the typical local government and school district election date of May, they're looking to eliminate that and move it to other elections. So I think that's pretty interesting. I actually think this is detrimental to, to local campaigns because I think if you're on the November date, most of those campaigns are going to be overshadowed by larger political happenings. So it's going to be very difficult to understand what's going on in your local campaign and the things that are being talked about. Yeah, you're going to get more turnout for sure, but are those people actually going to know what they're voting for? It's going to be a less informed voter in November than it is in May. And I personally think that's a mistake, but hey, to each their own on that

But you should get more turnout though. So get more people voting, but they have less information about what they're voting on.

So talking about things that are moving through committee, one of the ones that I want to talk about is Jeff Leach House Bill 762. And there's also a Senate bill companion limits severance pay for city employees to 20 weeks. So city managers, if you have a contract that is longer than 20 weeks in severance pay, this bill would obviously change that. And so it would limit that 20 week pay. was voted out of the House Intergovernmental Affairs Committee. A little bit of good news on this one, because I think this is a terrible bill. I don't know why we're preempting local officials and their hiring practices. I think it's going to push some people out of the career field. In my opinion, it's probably not a great thing. But there is a Senate bill companion to that one that was left pending in committee, always a good thing. And so maybe

you

Maybe, well, let me double check that. I'm pretty confident I saw a Senate companion that was left. Let me make sure I was right on that.

I'll gladly just edit out this dead air.

Yeah, sorry bud. I'm going to check the bill site real quick just to see where it's at. Sorry. Yeah. The one that was limited was it wasn't a Senate bill. My bad. There was another House bill by Representative Bell, House Bill 4317 that would have just addressed executive employees of a political subdivision. But the leach bill is for both executive employees and independent contractors. Some cities hire their managers as like an independent contractor or contract employee. So one is so one was voted out of the government affairs committee and the other one was left pending. So it's like they're attaching to one horse, but it is not yet gone to the floor or moved over to the Senate. And we're getting pretty close to deadline time. So we'll see if that one makes it any further, but it is interesting that they're going after that. I would love to know the backstory of why that occurs. So yeah, anyways, that is the wrap up, man. Good luck on the edit right there, Chad. Sorry. There was some, a lot of dead air there.

That's fine. We use a new news platform and it does AI cuts. like my, my effort in editing these podcasts is significantly reduced now. So it'll be fine. What's, what's the scuttle on the bill that would like totally change the warehousing for sales tax sourcing? Any news on that? think, I think that's the bill last time you said was pretty much DOA, but I can't remember for sure.

Gotcha. Yeah, so yeah, let me let me go pull that bill number real quick and see where it is because I did not look at that.

We had a couple questions on that recently. I'm going to check the transcripts of our last episode. Actually, it was two episodes ago. Of course, I didn't put descriptions in the show notes. Sorry, guys.

There's two bills. There's basically a bill that is going to codify the comptroller's rule change, right? Because the court held that what the comptroller did was unconstitutional and one section of that rule change. The way that they did it is unconstitutional. So now they're just trying to do it legislatively and clean it all up. That's one bill. Then there is the bill that basically comes out from the city of Coppell and their movement

Well, the way that they did it was unconstitutional,

to basically allow for what they're doing from a business standpoint to continue, which would be to allow for their type of warehouse distribution centers to continue to source the sales tax to the location of distribution instead of the location of delivery, which is the change that the comptroller made back in.

That's the Scooping audio.

January of 2020 at this point? Is that how far back we go?

So what's funny is that they're admitting that these things aren't places of business. Like tacitly.

sort of, or they're trying to fight the cleanup with their own cleanup on aisle four, right? I don't know if they're tacitly admitting that.

Is that. that that that's how I would read it it have all of those lawsuits been either resolved or paused like are any of them still ongoing

They've been left pending yet. No, they've been there. Yeah, they're still ongoing, but they've been left basically in like a pending pause state.

I'm going to keep my mouth shut then because I don't want to get involved in any kind of squalls. But it does seem to me like this is almost tacitly admitting that they're not places of business.

Hahaha You know, look, I'm going to say this again. These cities did this. It's been a long time since we've talked about this, but these cities did this and they went and got taxation letters from the comptroller. Okay. It was a very different world. Correct. So I personally do not understand why we don't have a cleanup bill that just says we are going to leave this group of taxpayers grandfathered.

It was a different world when they did this. We didn't have online sales tax. It was a different world.

They got taxation letters. Cities were built on it. Revenues were built on it. Budgets were built on it. We are going to pause in time the rules for these specific taxpayers due to what they did. Right? Everybody was involved in a process. Everybody knew what was going on. Now, did we know that Amazon was going to be as big as it is now? No. Right. And this, to be fair, this is not Amazon and Coppell is not a client of ours. So we can't see the data nor if anything that I'm saying coming

And even if we could, wouldn't be able to say anything about it.

we would be able to say anything about it. So is this not coming from confidential data? We do have a number of clients that are involved in this lawsuit. But the reality of it is, is I don't think what they did was right. The cities, right? But they have their own fiduciary responsibility to their taxpayers, right? I had a conversation with a city manager the other day who was upset at another city manager over ETJ releases and ETJ reforms. And I said, I know you want to be pissed at that city manager. get it. But they have a fiduciary responsibility. So if you're leaving it open for them to go get that ETJ, right. And it's what's financially best for them. You can't harbor ill will towards that city manager. Right. They have, they, have to ethically do what's best for their community. Period. And I think that's what happened here. I just don't like it as a city manager of a city that would have been impacted by that. Right. You know, then, then you have the other side of the equation, which is there are some communities that legit have business operations in their community that are impacted by these rules where I think we're overreaching. Right. I've said this before, Round Rock, Texas, Dell computer. You cannot, you cannot drive to Round Rock and not see the operations of Dell computer occurring inside of Round Rock. That is a, those are corporate offices. Those are business operations. There's a lot going on there, right? And there's a lot of impact to that city from that employer. there should be some type of revenue stream there. So I think there are polar opposite ends to this equation. There are different lawsuits between all of those different cities as well for different reasons. But I just don't see why we can't use logic and say, okay, let's freeze that $30 million that Coppell claims it is. Cause that's their letter, right? That they gave. Let's freeze that $30 million that they claim it's going to be. And let's say it is what it is. That's the grandfather amount of money and move on. And I just don't think anybody can get there and the legislators themselves can't get there because they represent the political interests, but a little bit fiduciary interest of their markets as well. So if you're a state representative who represents North Dallas cities that are impacted by these distribution facilities, you're not exactly wanting to do anything to help Coppell Texas keep their $30 million. Just as what it is. So Yeah, so mean, it's an interesting issue, but I don't know where it's going. I'm trying to go look at the sales tax bills right now, Chad.

I feel like I would be able to to grandfather in those existing agreements, no matter where I was represented. This is getting really difficult with the dogs. I'm hoping that my oldest dog is like 16. Sometimes I feel like she's on death's door, but she's just like becoming senile and she just bark at everything.

How many dogs do y'all have right now? Well, so violating most city ordinance limits for the number of dogs you could have. I know, I know, but I'm laughing. So.

F5 I don't live in the city, but... Is there a limit on the number of kids you can have?

You violated that one too, my friend. So, Man, this is killing me. Did we not put it in the show notes?

It's a 924. HB 924. I don't see any activity. It was referred to Ways and Means on March 6th.

and it's not been calendared. And we're getting, yeah, yeah, we're getting, we're getting pretty, pretty slow. And then Mays Middleton is the other bill.

It's just has nothing for a month and a half. He's just announced running for AG.

He is running for attorney general.

at 414. So well, no, that's 414 is the one where you have to state the interest costs as well. I'd be curious. If they're like, would they entertain an amendment that? Sure, like we're going to show you the principle and interest. I get it. That's a cost that we're bearing, right? It's not just the principle that we're paying. But what would be the estimated cost of these projects if you waited to cash finance them? Like if you if you instead of taking out debt because we're taking it out because we don't have the money to cash finance it, instead we're going to save all this money and then you know in 15-20 years we have enough to do this project. Now like what condition is the road in and how much is it going to cost to build it or to or to repair it whatever the case may be. Like could you do a side by side comparison? Because there's like there is a reason that we take out debt for things right is to. bring future spending into current dollars. Like they should also probably require this on any kind of loan.

Right. No, they.

You get a mortgage and ask like it obviously they does tell you like if you look at the amortization schedule it tell you what you paid on it but like still. I get the, like, I understand why they're doing it, but it seems like petty. Woof! Woof! Woof! This episode is really going to test the ability of Riverside to edit.

of this, of the aptitude it's editing. Yeah, sorry, I'm trying to find it, man. I'm not sure. I'm not sure where it got lost.

No, not for you, because my freaking dog's.

Crazy that the number of it's House Bill 134. finally found that the number of bills that were filed and property taxes nuts this year. Let me like 134, House Bill 134.

144.

It has been voted out of committee and it has been sent to calendars. So it is is most likely going to make it to the floor.

considered to the counter.

It did not, however.

This is the one that's codifying the controller rules.

This is the one that is codifying the Comptroller's Rules, that is correct. So it doesn't change anything really for cities that weren't involved or part of the lawsuit issues related to distribution facilities. to give you a little bit more clarity on what the bill does. It basically would codify that a place of business of a retailer does not include a computer server, internet protocol, domain name, website or software. The sale of a taxable item is consummated at the principal business location of a small business permit holder, but its business location in Texas of not more than 20 employees and total combined gross receipts from the sale of tangible personal property and services less than $500,000 in a preceding 12 months. That is a new

you

addition, right? To prevent the, do you remember last session, the lady who testified that owned like the jewelry store and I think it was Flugerville. I think that's what that one was. Each sale of taxable item by a retailer other than small business is constipated at the place of business and the retailer where the retailer receives the order provided the order is placed in person at the place of business of the retailer or the sale of the taxable item not covered in items number two or three above.

Yeah, I think so.

location in the state which the item is shipped or delivered at the possession is taken by the purchaser for the sale not covered by all of those things above the place of business of the retailer where the order is the place of business of the retailer where the order is received or the order is not received in a place of business of the retailer the place of business from which the retailers agent or employee took the order operates.

So it's a little bit different, but not terribly different.

It's a little bit different. looks like they're handling like phone sales correctly there.

It also looks like Not much should change for most things like lawn care. Except for maybe. Pretty big companies made $2,000,000. I don't really know. $2,000,000 in taxable sales. I don't really know if that's like out of reach for most. Probably would be like how many blondes would you have to mow? 13,000 or so at $150. It's probably, probably much.

Right. But. but it does not remove the existing tax code on single business locations. There still is that exemption in the tax code, and this doesn't specifically remove that.

Okay.

So, and then any city that has a 380 agreement that was dependent on sourcing, quote unquote, right? That would be able to continue until December 31st of 2030. it does provide that surprisingly I've talked to a few of these cities though, and a lot of them don't have incentive agreements tied to those users. So yeah, I was surprised by that too.

So five year window.

So one of those communities was at the conference and we had this conversation.

So what would have been the incentive? What was the incentive for the business then to do that?

I think in some cases they didn't need an incentive because the location was so good being centralized, especially in a DFW market. Um, and, and in other cases, I think the incentive has already gone away. It was like a 10 year deal and it's past the 10 year deal.

okay

So, so it is different. Like there, there are more considerations put in this than maybe the comptroller's rule had in there, which to be fair was something that you and I were pretty critical of with the comptroller's rules that it didn't take into account some normal operating business operations. but yeah, that's where we are. That's what it is. So, and that folks is the legislative update. Good luck cutting that out, Chad.

Thank you. I'm looking forward to the editing process.

So let's get into your nerdery now and let's talk. Let's talk about it. Yep.

Okay, let's get into my notary now real quick because we're already at 52 minutes. We're at 52 minutes live. I have no idea how long we'll be into the podcast once it's edited. if it's like 40 minutes, then just bear with me here. Okay, so last week, you know, you talked about the UNT conference. It was all about AI, which is kind of crazy because like just a couple of years ago, we never would have expected to have an MPA conference. focused on AI, but here we are. So as part of the process of like prepping for the talk I gave, like I did a lot of demos in my talk because the topic that I had was like AI 101. So like, does this stuff work and why does it work? And so I didn't want it to just be like dry, you know, me talking in front of people, but so I wanted to have like interactive things. But the process of building all those things kind of got the creative juices flowing. And so we obviously have Gavin and our data science sort of like team that's growing over time. That's much more. well suited to do these types of things, particularly on the machine learning side. But I just want to report to you, Patrick, I did successfully build my first neural network over the weekend. And it was pretty cool. It's not going to be like a final product, because I've already got lots of different ideas about how to make it better. But it does work. And it's pretty neat how it all works together.

This is for the cuisine. Is this what this one is? Okay.

Yes. What's funny about this is that I think there will be other applications beyond the initial purpose, but the initial purpose was can we build some kind of machine learning model? Because so right now we when we import data, we are tracking the next payment to determine if that is anomalous in some way. But we're not looking yet at the broader payment history of the taxpayer to see if it's anomalous relative to its peers. And so that's kind of where that's the next step is let's expand it beyond the next payment coming in and look at the bigger history. And to do that, we really need to build models for different types of businesses. We can't just have like one global model because there's so many various variances or variables. So I wanted to start with restaurants. And, you know, as I'm kind of like working on the feature list for the things that we're going to be modeling, one of the things was cuisine. Right. Cause different restaurants based on their cuisine are going to have probably have different, different patterns that we want to isolate. And so, so this first iteration, I pulled like 1400 recipes from all recipes.com and I classified them into a cuisine. And then we extracted all of the ingredients and built this like vocabulary of ingredients, which we then, we, coded those. We used a model to code the semantics behind those ingredients so that you could do a similarity search on like Greek olives and you're going to get like a feta cheese and other things that are like in the Greek cuisine, which is really kind of cool. So then the neural network itself takes all of this and creates a new model on top of that that classifies the cuisine based on the ingredients. So not only can you take a list of ingredients and find the cuisine, You can take one cuisine and find similar cuisines. So like if you're looking at Japanese, it's going to pick up Korean and Vietnamese and Chinese and things like that as like the closest counterparts, right? So it's pretty cool. The biggest problem with it is that the way that it's built is I hard coded the cuisines into the model, right? So like I could have easily just

Okay.

built a little graph saying these cuisines are similar and these cuisines are similar. when a business gets tagged as this cuisine, like these are ones that are kind of similar. And the idea here obviously is at least one of the ideas is that if you are say trying to model a development and you know that you're going to have this type of taxpayer, well, we can't give you data on how much that taxpayer is generating across the state because it's confidential. But we can find other instances like within that cuisine, right? And then even filter down to price point and atmosphere and things like that. So like we could, this is kind of overkill if we're going to be hard coding the cuisine. So the next step is to actually take menus from actual restaurants and create a model that will learn from those menus and build its own cuisine classifications. So I'm excited about that next step. And then, and then taking that same concept and applying it to retail stores so that we can find stores that are similar, especially for like smaller stores where, it's not not a national chain. But if we can pick out a few of the items that they're going to sell, then we can find other stores that are similar to it across not only the state but across the country.

Right. Okay, so to the economic developer, to the city official that's working on budget or finance or things like that, what, what is the real life expectation of what this is going to bring to their world when we're able to do this?

Well, I think that that first use case I mentioned is probably like the initial reason why we considered something like this, right? I want to have just a suite of tools running all at all the time to enrich the data that we get from whatever, you know, Comptroller or CDT FA or wherever we're getting this data. We want it to be supplemented with additional data points so that we can connect things together.

Right, right.

and provide you with a richer just experience with your analysis. But the initial use case was, hey, because we get this all the time, hey, I'm getting a, you know, recruiting this particular business, like how much are they going to generate? Well, like, depending on the data that we have, we can maybe give you an estimate and a range. But there are a lot of cases where we either don't have enough instances of that one taxpayer, or maybe we don't have any, right, because it's something local or whatever the case may be.

if it's 2000 square foot of Italian cuisine, we would, we would now be able to give you an estimate of that regardless of who the brand was. Yep. Yep.

Is it Sicilian? Is it northern Italian? Is it Tuscan? We can start to get really granular in terms of those estimates.

And to kind of give this a lot of our clients, I don't know, do we even show this on the public side? So if you go in our system and you look at like Walmart, I know on our side, this may be a beta thing that may not be visual to everybody, but I know on our side, we can see in the Walmart taxpayer area, there's on the right-hand side, there's an area that says also located near or something like that. I need to look this up specifically, but, it would.

frequently located near.

Yeah. Frequently located near and it for like an economic developer to give like five or six other users that we see around Walmart's within our system and other areas, right? Or throughout the state, it's kind of a globalized push that, it was like one of the initial things that we did. Now what you're doing with neural networks is significantly more sophisticated than what that is. Right. but like for an economic developer, I just kind of want to give them a, like we've talked about doing these like economic development roadmap plans based on Okay, hey, like I had this conversation with the city of Willow Park a couple of years ago where it was like, Hey, we just need to know like, what should we go after? And I'm like, well, it's actually kind of a cool question. Like, can we take Willow Park in their development stage and then go look at like 20 other cities that are in different development stages, but similar demographics and similar future demographics, and then build a model. Obviously we could do that. Like there's, there's no question that we could do that, but What you're talking about is, building a neural network that would eventually be able to more, like create that. What, did you use, provide, like nourishment or enrichment enrichment. Sorry. Like we could just almost, I wouldn't say automatically, but like the machine learning that's going on behind the AI that's going on behind the scenes and always running in our system, like where we're moving in the future.

enrichment, yeah.

we would be able to constantly enrich and to be honest, provide a lot more reason to you, for you to be in the system more often. Right. and so,

Yeah, like, this restaurant you're looking at is in the top 20 % of restaurants that are similar to it. Right? Like, we've been building tools to tell you like how you're how you do with Amazon revenue relative to other cities in our platform. Like, what percentile are you? That's great for Amazon. And it kind of gives you an idea of like, how much, you know, per capita do my residents spend?

Right.

versus other cities across the state. But that's just that one business. And so if you're looking at like Joe's Pizza and Pasta, of which there are thousands of individual Joe's, I don't understand why like Joe, people like named Joe like to make pizza and pasta restaurants. But you know, there's one on almost every corner. They're not like, it's not a chain. It's just a restaurant. But if you could look at it's like, how is this one restaurant doing relative to its actual peers, not just to other instances of that taxpayer ID. That can really give you an idea and also track that over time. like restaurants or webs or stores or sales taxpayers and businesses that are either improving or struggling, right? Like retention and things like that are those are good use cases also.

Right. Yeah, absolutely. mean, I think this is a I think we always wanted to get here, right? And I've, but I think we're getting here a little faster than, I expected, right? I think if you, you know, four years ago, I, you know, I would always be like, we're going to get here one day. And I had no idea when that day was. And like, now I actually can see visually like that day is, in focus, right? Like we, know where that's coming. The next thing that I eventually want to get to is, which I've said this to folks and I know I've said this to you, but.

soon.

I eventually want to get to a place where you have a little bit more of like a free flowing language interaction with the system, right? Where you can speak to the software, you can speak to Zac in natural language and you can say things like, how is my Walmart performing compared to other Walmarts in the state? Right? Should I be concerned about any businesses in my city? And there are so many things that have been enriched in the backend through these these models that you're working on that it would be able to answer those questions, right? And we've, you know, as part of your presentation, we had a ton of ideas that flowed as part of that presentation. But, you know, one of those, one of those things was, you know, we got a great idea for how to build some of this. Cause the biggest concern that we've always had is, right. And I'm just going to say this out loud on the podcast. The biggest concern we had is the use of AI. is sending confidential data to an AI source, right? Now, whether that's open AI, you we would never consider sending something to DeepSeek, you know, like, yeah, you, you see what I'm saying, correct. but talk about that. Talk about from a city perspective, the ability to kind of sandbox systems and, and what that means from a security standpoint for us and for those cities.

You seek so good source, but yeah, so you'd have to either find a provider or run it yourself. Yeah. So. So an interesting point that was brought up at the roundtable was. Each chat that you're engaging in with one of these language models, that's a record that you're going to have to retain. Or at least it's going to have to go through your retention analysis and you have to decide if it needs to be retained, right? And so, what do we do there in terms of using a commercial language model? like Claude or chat GPT or whatever, can we instead host our own language model internally? Crank up a server, get a couple of GPUs and run an open source model. Facebook has a really popular open source model called Llama. Lots of people are putting out open source models. You can use them. Some of them are good, some of them aren't as good. But then you can control it, right? problem with that is you need to have at least a little bit of understanding of how to run that server. also, that server or is that model that you're using going to be out of date in two weeks or a month, right? so, yeah. So how do you, how are you constantly following and updating versus just using a commercial platform and you know, they just, they do that themselves.

Right. That's the bigger issue, right? Is that piece.

So there's this balance between those two things. You obviously can control more of the retention. Also, you have better privacy if it's your model. so, so yeah, one of the things that came out of that is that I think we have now figured out our roadmap for. for utilizing a language model in a sensitive and secure environment where we can generate content now based on confidential data and we are confident that there is no problem with retention by a third party or leaking this information out into future training sets or things like that. So that's exciting because there's a lot of stuff that we can do now already, but there are

Yeah, I think.

several things that we really want to do that we just haven't been able to yet because we cannot send confidential data to like, you know, Anthropic or OpenAI or whoever.

And to be fair, because I love to give people a glimpse into like how we think and come up with ideas and how we, you know, like the way our entrepreneurial minds work. We're sitting in this, you're on this panel, right? I'm just a, you know, a peon in the back who's, yeah, by the way, being made fun of because Chad generated pictures in his presentation of me on the constant. But we're just, but it was me of,

You're just the heckler giving me a hard time. Just two. And on one of them, you were, yes, the guava dude.

holding a guava. This relates all the way back. I'm called the guava dude internally folks, because I didn't recognize a guava.

It wasn't actually a guava. It was a lemon. The joke was you were like, what fruit is that? And I said, guava dude. And then Allison said, you're the guava dude from now on. And I'm doing my damnedest to make it stick.

Okay. Yes. For now on. Yeah. It's trying. It's probably, I think it's kind of stuck at this point, but you know, it'll work. holding my guava that I don't have in my hand, but anyways, so, but I want to talk about this because we have been, I mean, what six months now. I mean, it's, it's been a while. Like we have been talking about it. How do we do this? We know what we need to do. We know the things that we need to build. There's so much innovation that's happening, but we kind of know like in our mind, like wireframed what we need to do. Right. but we didn't know how to do it mainly because of the confidential data. And we couldn't, you know, we know that we weren't going to go build our own system, right? So we're sitting in this room. There's this round table that you guys are having. It's awesome. This panel discussion, y'all are doing a great job. The CTO, right? Of Lewisville, Texas, CIO of Lewisville, Texas is on this panel. And he starts talking about something similar, but not the same. And he kind of starts talking about how they had started playing with a model and put it on their own server and sandboxed it.

CIO.

and what they had done with it. And without knowing it, you and I had the same thought at basically the same time. And so I'm driving home from Denton, you're driving home from Denton. I pick up the phone and call you and I say what my thought was, is this possible for us to do this from the idea that he was kind of giving it, you I kind of took what he said and I was like, I wonder if we could stretch it to here. And I call you and I ask you that question and you're like, dude, I had that same thought. Like in that moment, same time, same place, same thought. and like we're a week, we're like five days out from that. And now we know like we can do it, which is cool. It's like really, really cool on how fast things move.

It's also cool because we rarely have the same thoughts. Like we coalesce to the same place, but we really start at the same place.

Right. That is true. mean, that's a really, really like true statement. Very fair. Right. Would you say that you would you say that I typically reach I typically in a reach guy and you typically are the bring me back down the planet Earth guy on that?

I think there are times I think when we both fill that role.

Okay, in some areas you're extremely creative, right?

Thank you. Mostly, mostly you're probably more of the Reacher. But I think that's probably because there is this. I'm sorry, I'm him bark. I think part of that's probably because when you don't have as deep a grasp of what the tech can do, then you're not limited by that understanding, right? So like I work in this stuff a lot more frequently, so I have a little bit better understanding of what we can actually do with it. Whereas since you don't, you're not limited by that constraint. And so you also are a lot more in the weeds with our clients, and they're asking you random questions, and you're like, I wonder if we can do this. Whereas a lot of my interaction is customer support, sometimes random questions about tax law and things like that that come through the support channel. But I'm more on the building the product side and you're more on like, let's answer questions and see if there needs to be some kind of consulting thing that comes out of it.

right? Well, taking that full circle, which is why I love that AFI conference, right? Because I'm going, you're looking at what cities are doing that is quote unquote creative now and thinking about, but how can we push the limit 10 years further? Right? Like what are we looking at way past there? I, I use the word Reacher. I would say I'm more of a, I probably push a lot too. I did that in management as well to the point where I almost broke people to be fair, but, So I can be fun to work for, I can also be kind of hard to work for because sometimes I have like very high expectations. And a lot of times I'll sell things that I don't know that we can really do. Right. But then we end up doing them. And so.

times. Yeah, we just finished an analysis that was very mentally taxing because the data we had, it was just difficult to achieve what we were trying to achieve. And so we had to do a lot of thinking about how this is going to work. So I appreciate that from you.

Right. But it's, it's, it's cool though, because I think when it comes to this, when it comes to this growth within our company, when it comes to what we're going to be able to deliver, I like what you said again, the enrichment that we're going to be able to provide. Yeah. I think, I think we're both in that, on that same page is like, this is going to be something that's really a game changer for us. And I'm pretty excited about that. So yeah, I know we've been on this for a long time, so we probably need to wrap up. Did you want to add anything else to that? Okay.

No I do not.

So I'm gonna let you go edit all the dog barks out. Some of them are for sure gonna be in there. but yeah, so that was helpful. Yeah. But yeah, so anyways, I appreciate you guys listening to us today. If you've got some free time, I hope you are actually listening on your computer like I will be to the house floor debate. It should be fun. There's nothing better than Texas legislative politics in Texas. It's the wild, wild west sometimes.

For the most part, didn't bark while I was talking.

So take some time and enjoy that. Other than that, that's all I got Chad. See you man.

See you next time.

All right.