Transcripts

Transcripts are generated using a combination of speech recognition software and human transcribers and may contain errors. Please check the corresponding audio before quoting in print.

 

 

Coming up on today's episode of Voices in Local Government, the AI Sandbox, less

meeting, less hypothesizing, less bs -ing, more doing with Parth Shah of Polymorphic.

Check out learninglab .icma .org for more on AI and also register for annual

conference, October 25 -29 in Tampa, Florida where AI will be a key theme, conference

.icma .org, boost linked wherever you're listening.

 

Welcome to Voices in Local Government an ICMA podcast with us today is Parth Shah

CEO and co -founder of Polymorphic And he's here to discuss tangible steps for local

governments to improve their AI policies and use Welcome Parth. Hey Joe.

Thanks for having me Parth Martha's MIT and NVIDIA alum, so that right off the top

kind of begs the question, why local government for you, why local government for AI

use in general? And if you could weave into that answer, try and convince the

audience to not only keep listening but keep pushing their organizations to stay on

pace with that kind of what seems like never ending treadmill of AI and technology

in general. Some people might have a little fatigue just feeling like they can never

keep up. Oh, 100%. So you're here to give them the good information, but also

convince them it's worth it. So go ahead. Yeah, no, and I'll start with that piece,

right? There is a lot of conversation right now around AI. There's almost a session

at every conference and hundreds of articles out there. I'll say two things that,

you know, I'll promise we'll have some interesting perspectives here and we'll make

this unique. One, there's a lot of conversation around AI that's happening in local

government with folks who actually don't have a background in AI. And so we're going

to actually dive into what's possible, a lot more technology forward.

The other thing I will say is AI is used in generalities very often. A lot of

folks say, Hey, we should be doing AI, we should be doing AI. And I think it's

frustrating for a lot of city managers, because they don't know specifically what

that means for our organization. So we've done probably the most diverse set of AI

use cases. We've launched some of the most unique ones as well. Have over 200

different departments at the city, county, state level using it. So we've done up to

near the most deployments. So we will focus on stories from our customers and case

studies. And I think that will resonate a lot more beyond just the general, hey,

you should be doing We're gonna we're gonna have some specifics on Here's what folks

have actually done and been very successful with okay, and right before we get to

the specifics The let's cover a little bit about yeah that the policies, which is

maybe sometimes the the first Stumbling block. Yeah, what have you seen for the

local governments? That's worked or maybe not worked as well in terms of actual

policy for staff. Yeah, Yeah, that is a great question. So we had a session

recently, Brett from Sassoon City had a really good perspective on it,

where really the perspective we're seeing from the innovative cities is,

and I understand that a lot of folks want to jump into policy, but it's really

hard to build an effective policy around something that you haven't necessarily tried

out or worked with. And so Brent and Sassoon City mentioned it, and this is my

general recommendation is I think policy comes second. I would say experimentation and

trying some stuff first actually is going to lead to a better policy. The idea is

starting with the sandbox, right? So there is a folks like us, we often will Let

you test before you actually do anything with it. So with our AI chatbots and AI

phone systems And we'll talk about that but in general you can with the right

partner You should be able to test well before putting something Out in the world

public etc. And I think there's value in that because AI is such a new Interface

is a new way of working. I'm sure a lot of your staff is already using chat GPT

or clod in their day to day life. And so instead of starting with the policy and

trying to figure out how to regulate something that we don't have enough repetitions

even using, I would say the policy will come second. It'll be a lot more effective

if you start with the sandbox. And so that's what Brett did a really good job

sharing this perspective on in Sassoon City. A few others have really talked about

this, you know, Randy, Preston, Ryan counties mentioned this where they, they, they

said, we've already have policies like fair use and those content policies that help

us get started. And then we will figure out the AI specific policies once we have

more experience and use cases we've actually developed there.

- Okay, and when you say sandbox, can that be department or even maybe a smaller

team? It's not necessarily company -wide, so they can test it. They meaning maybe an

individual, maybe a small team, on uses that will actually help in day -to -day work.

And I think in a lot of cases, you said people kind of are maybe already using

chat, GPT or others to help summarize or take notes or even record meetings and

just spit out the notes instead of having to take them manually. But what

specifically, can you give us an example of, okay, they had the sandbox, here's what

they did with it? I love that question around sandboxes, because I think it really

opens up a level of comfort and ease when using these

AI technologies. And so you mentioned And I think there's two great ways to do it.

I think of it as breath versus depth. And so one is you start with a specific

department. So in our case, you know, we're doing the AI concierge, right? So more

than just a chatbot, we'll do the chatbot, the phone system, SMS, et cetera. And

one easy way to start that is actually picking a single department. So sometimes

we'll start with human services or planning and development. And what's great with

that is you actually get to see the full suite, see how it works on the phone

system, iron everything out with one department, and often folks like to pick the

department that might be seeing the pain point most acutely, or in some cases, it's

a department that can more easily deploy it. They might be the smaller or more

nimble team, and that's great because all the lessons and learnings from that fold

really well as you move organization wide. Add that to, of course,

you know, our experience having done probably some of the most extensive deployments

here, we're able to bring those from other cities, other counties we've worked with

and bring that over. So that's generally why the chatbots and all that go better

with us. Can you, sorry to throw up, but can you, it was interesting you said the

department that may need it the most or has the biggest pain point. It's a little

surprising to me that that's not just kind of automatically the IT team because

they're maybe naturally a look toward look to To be in charge of it for the CIOs

out there or the CTOs whatever the title might be Can you convince them to let

another department? Kind of go first or be in charge of that sandbox Yeah,

so I think you may still be in charge of it and still managing it. But I think

from a use case standpoint, oftentimes, IT is not as connected to a natural

department that's facing the pain point, right? So AI is a great tool in my view,

but it still matters what that use case or challenges that you're trying to solve.

So in our case, it's a lot of overwhelming phone calls. Customer service is very

hard because constituents have a hard time finding answers on the website or they're

dealing with a a legacy chatbot that doesn't work. So they can't find the answers,

they keep picking up the phone calling, right? So that's a use case that we start

with. So IG's involved, but an example, I was just talking to a county in Florida

today about their planning and development, and they get 150 calls into one of their

small divisions within that department a day, and they only have a few folks who

are actually handling the front desk, they actually don't, they had to rotate who

hops on that. Right. And then you get, you get whole times and then the, the menu

bots that just kind of send you in circles and in the, in the press one for,

yeah, the press one for this division, press two for that. And so it's really

scoped down. If you think about it, it's a division within a department. So it's,

it's, it's an easy, the group that's involved in This is five,

six people, so the training's easier. It's also a valid proof point still.

So you get the benefit of both worlds, which is it's easier to roll out. It's

gonna take two weeks to get the chatbot up and running, the phone system takes

another three, four weeks after that because it's such a small group. The training

is really easy. The surface area is very small, but it's still valuable because What

they're doing isn't necessarily unique from the other divisions that department rolling

out or the county as a whole doing it and so I really like that kind of sandbox

rollout or

almost initial like phase one phase two because You're doing the full depth.

You're just controlling You're scoping it down to a certain area. The other way I've

seen it on the sandbox, which I also like, is starting with, if it's the larger

organization, that's also fine, but starting with a scoped version of the product. So

you go wider, but it's shallower in that. And so oftentimes we'll see that as well,

where people will say, this is great. We could use this AI customer service

platform, AI CRM, across the board, but let's start with the chatbot, because that

can go on the website and we'll get to the phone systems later all of that and so

that's the other way of doing those sandboxes which I love where now everyone gets

a little bit exposure and then they get comfortable with AI that way in the easiest

to roll out piece which for us it is the chat bar right it takes two weeks it's

very easy to roll out we're easily able to replace some of those legacy ones that

that folks have struggled with and then they're able to move into doing the phone

system is more complicated takes a little bit more you know comfort and testing and

now everyone's developed that muscle across the organization doing that on a much

simpler product which is sits on the website so those are the two different ways I

see these sandbox or rollouts going really well is either wide exposure or really

deep exposure in a smaller division or department. Okay. And then once that example

of the chat bot, maybe just helping the customer service or even the zoning team

respond to those inquiries. Once the users get to play around with it and hopefully

the tool helps Take the volume of their workload down, but they're still kind of

controlling it then

Is the next step to then getting getting more Formal or written policies down or is

it still is there something in the middle? So what's phase three? I guess that's

what I'm asking Yeah, I mean, so phase three is where you start to get into just

the continuous

Deployment right you're moving to an age now where "Hey, we've done this." I think

that's when you start in parallel working on those policies if you need them. We've

seen a lot of folks who have actually then found that our existing policies with a

few amendments actually kind of cover it already. - They cover it. - And now at this

point, what's great is you've kind of developed the muscle with a successful

deployment and AI is gonna pop up everywhere, right? But what I would encourage

folks to to think of as it's a tool, right? It's the same way. It's like the

internet, right? Before the internet, everything was done on pen and paper, right?

Before we had software and technology

that even ran on premise. So I guess even before the internet, how did you do

finance and internal tracking, right? There was no such thing as an ERP, right? And

then we have ERPs, we have permitting software, So,

so that was a big change, right? That was a different way of doing work than the

pen and paper based way. And AI is going to do the same thing, right? It's just

the next phase of that. So you're going to see it everywhere. So phase three is

more, hey, we've had this success. Most folks are having that success with us,

right? But there's other technology out there that's great and and use sandbox that

now it's like, we understand how to roll this out, we understand how it works,

we're comfortable with it and we see its value right and so that's what phase three

starts to look like you start to see it in more departments you start to see more

use cases you start to see deeper use cases that are that are really driving even

bigger outcomes so that's that's really the path that goes down in phase two phase

three what I would say is is a continuous deployment phase of it right we now know

how to do it we're gonna keep adding it we're gonna keep on the lookout we've

developed that muscle and now we understand we can use this for a variety of use

cases and more more will keep coming okay can you address some of the common

challenges I don't necessarily want to say failures but what have you seen with

maybe common mistakes, even as organizations try to follow that path you just laid

out, what are some common mistakes and then the second part obviously is how to

either avoid them or mitigate the downside if it's already happened or is about to

happen. So one that is very personal to me is as I mentioned, my background is in

this world of AI, right? I did electrical engineering, computer science at MIT. I've

led the machine intelligence community there for a few years. I've spent a lot of

time around AI a lot of our team has. What's concerning to me is I'm seeing a lot

of folks branding stuff around AI or claiming it's AI without the background around

it. And that's where we're seeing a lot of failures. So not to, you know, obviously

not naming any specific cities or counties or situations, but it's really unfortunate

because folks will not do the due diligence on, "Hey, are there technologists on

this team? Do they really understand this AI technology well?" And then the

deployment, as you'd imagine, would go poorly, right? If I asked, "Hey, I want to

get a home build," and someone has your experience, you know, doing construction

building, those buildings, A year later, all of a sudden, we see these organizations

that now shut off to all AI, and I think that's very unfortunate. So that is,

I think, something we've seen go wrong from a diligence standpoint. As much as you

may have worked with this organization in the past, they may have the history. I

mean, there's a different bar, I would say, and you want to have that domain

expertise. Right? They may work with a lot of organizations and it may be really

effective for XYZ technology, but just because they add AI doesn't mean That's an

area of expertise for them. And so I think you have to be be really careful about

that. We've seen that In other industries in this industries, but something as simple

as even, hey, like people try to integrate The chatbot that doesn't really have an

AI background in it. And all of a sudden, people turn off to all chatbots, right?

They say, well, we don't get accurate answers out of it. It's nice.

It's a cool tool, seems fun, but it doesn't work. And that's not necessarily true,

right? It depends on what you're using matters. Let me put it that way.

And just as a user, not necessarily even on the local government, but it could be

be a healthcare situation or any type of website. I'm pretty skeptical of the bots

just because they've been around for a while and it's more often a waste of time.

But lately, I've noticed at least it does a decent job of getting me to the right

real person quicker in the right department and I can just explain with one

paragraph what I'm trying to do and it can get me there. But I still ultimately

want to talk to the person. I've found it pretty rare to get every answer I need

just from the bot. Honestly, at this point, I feel like I have a better chance

Googling it and trying to get to the right, because I'm certainly not going to poke

through these clunky websites either, but if you Google it the right way, you can

maybe find the answer quicker. I think that's a great point, and that's exactly in

some ways.

- It's a funny challenge for us, is we're having to show folks that that can be

different. Because chatbots is the same term, but they act very differently.

So back in the day, the way chatbots used to work is, it was actually closer to

search, right? It was just a fancy way of doing search. So you had to pre -program

all the responses in. What it's not doing is actually an intelligent,

Hey, here's the knowledge base. I'll look at your site, you know, we have

proprietary algorithms for, for how we work with like municipal codes. And so no one

has to program responses and we're looking at your content and we're actually

understanding your content, learning on that. So when someone asks a question, it

answers it. And so I think we'll see too in like healthcare and other industries,

like these, these bots are going to get better over time. They're going to start

doing stuff like we're doing. But I think it's not to be understated. There is a

historical frustration with those experiences.

And so oftentimes we're seeing people shut off to that. And I would encourage

people, you know,

you can, you will feel the difference. I wouldn't shut off to experience because

there was probably a reason that you were interested in that solution, right? Search

probably wasn't working. People were getting frustrated. A lot of phone calls coming

in. So the problem's still there Just because this old solution set didn't work

Technologies changed a lot. And so it's a tough thing, right? It's it's fool me

once, you know Kind of situation for lunch and one you but people don't want to be

fooled twice And and so that's that's just something that we with and I would say

hey if that's something that's happened to you and you're a listener like do reach

out I think we have a different approach but George W Bush excuse me about can't

be fooled a third time can't be fooled a third time that's right pull me pull me

again but you're not necessarily gonna be if it's actually different and improved so

are there other not necessarily in getting it wrong or maybe going for a solution

that claims to be AI but AI but isn't, are there any other unintended consequences

the audience should be aware of that they don't even know about yet or someone like

you maybe have seen just because of the nature of what you're doing day in, day

out, but I think everyone as you said might already have that experience of

frustration with a particular solution that didn't help and even as you said maybe

testing it or sandbox is a good way to preempt that. But what other potential

unintended, excuse me, unintended consequences can come up? I mean,

I think another big one that that I've seen, and it's it's it's actually a very,

this is going to sound like a funny way of framing it, but I think it's helped a

lot of folks in a positive way rethink what The outcomes the key outcomes of the

organization are and and how we can use our time to achieve that And I think

that's that's something that takes some time and thinking to grapple with so what

what I mean by that a Lot of stuff that we do in this space Especially in

government. It's it's not an easy It's not an easy World to work in right it's

often a thankless job right people you hear a lot when people complain You don't

hear a lot of thank yous when things are going well. And so there's all these

mandates that these these folks have. And what we've found is it really does help

to have an understanding of what the outcome we need is. But equally importantly,

what the internal process or how we're currently doing it is. So we talk a lot

about cost of inaction, just as much as the cost of the action. Because I think

what people often overlook when they're looking at AI or something new is they view

it as something new, right? They view that as, oh, we're doing some change, right?

It's a new path we're going down. What I would encourage folks to think about is

instead of thinking about it as like a, I don't know, like I'm thinking of a

highway and it's an off ramp, what you're really at every given point is a fork in

the road. Like doing what you've always been doing is a choice as well, right?

Inaction is a decision. And if you don't know the cost of inaction or how we're

currently doing it, I think it leads people to severely underestimate the value of

some of the new stuff that's coming out. So a lot of folks, I mentioned that

conversation I was having with this, this development department. So what is the

current problem, Right AI sound it's a tool in technology, but I always mentioned

it's all about the problem. We're trying to address and

For them it's they don't have a front desk and they're getting a hundred and fifty

calls a day Right, so staff is having to spend time going out of their way to

answer calls. We're just slowing down Reviews on their permits slowing down all these

other things that they need to be doing and so there is a cost today, right?

Obviously, it's new but choosing not to do AI for them

That's something that I would say is a big lesson, is be looking inward, right?

And I think if you haven't mapped out those processes, sometimes that ends up being

a key learning, which is we severely underestimate the cost of inaction.

How does that go for the frontline workers then? Do you have any, maybe not

necessarily long term, but do you have a sample size of, or any data showing not

just okay we've saved x amount of time people answering the phone but this staff

use the tool and it freed up x percentage of their time to then accomplish or

maybe speed up the average permit. Yeah no I think in this like this specific case

I'm talking about right they it's unique in that or it's not that unique,

it's very common, right? Most departments don't have their own dedicated 311, right?

If you think about most small, mid -sized cities, even larger counties, individual

departments, the brunt of handling those phone calls falls down to folks where it's

not even in their job description. I think this is, and this, what ends up

happening, right? Everyone knows this, retention in government right now is not Very

high hiring is very very difficult next five to ten years about 52 % of staff is

within retirement so That's what I encourage folks to think about a lot and what

we're seeing is if you can save that 70 to 80 % of calls that are general Q &A

Stash requests or they call the wrong department and you just need to route it

elsewhere The retention goes up People are more fulfilled in their jobs and it's

they're able to go back to the original job description that they signed up for

Right no one no one no one puts on the job description Hey, you're gonna answer

the same question over and over again for two hours a day and that's gonna the

phone's gonna be ringing You're gonna be mid permit review and then you're gonna get

a call and you're gonna get taken out of flow You're gonna go back and before you

can even get back into flow again You're gonna get another call right no one no

one's gonna advertise a job that way but that is the truth right and so this is

where turnover happens and people don't think about it that way they look at it oh

why is retention so low or why is you know maybe you know obviously there's there's

a complex question right there's there's a lot of deeper conversations around benefits

and things like that but I think an equally important part is Hey,

is the job itself something that could be engaging and exciting for not just one

year, but for four or five years, 10 years?

And I think that's the hard part with a lot of city managers who we're working

with who have been progressive. Obviously, they see it as technology, but in some

senses, it's kind of a retainment tool, right? It's a way to invest in my existing

people So they can do the job they want to do instead of doing the the laundry

and dishes we call it of what you have the grunt work Yeah, that's why I keep

joking it when is the AI gonna? Automatically resize these images that I want for

emails or whatever else That's I mean chat GBT is good for summary and even these

recordings. I get pretty good show notes Automatically, but there's still some just

grunt grunt work tasks that I shake my head and say,

okay, well, where is it? There probably is some tool out there, but like you said,

it's not always as easy for each department to get some specific thing just for

what they need. And I think the example of the chat bot and the customer service

is the obvious one because that is such a common pain point with residents. That's

right. Once that's all in place, let's say it has gone well and there's kind of an

established use case where you've taken that big, messy, complicated topic of AI and

you've kind of simplified it to here's what we're using it for and here's the time

it saves and the outcomes that are improved.

What are the practical takeaways for the local government leaders, whether it's the

city manager or their deputy or the IT director or these department heads, what's

the cleanest way to then actually get the policy updated factoring in all of those

tangible practical use cases, but also things like legal and compliance and budget

because it seems like there still might be a lot of chefs in that kitchen. So how,

what's the most efficient way to then, and I think you mentioned earlier, you don't

necessarily need a new policy from scratch. Sometimes it could be as simple as

updating what you have. What's the, what are the actual steps to get that done

cleanly? So, and I, and this is, I'm going to shout out an organization we're very

involved with here. So Gov AI Coalition, you know, we've uploaded our AI fact sheet

for our chatbot and AI AI voice there, we published it fully transparently. So we've

been working with them for some time, but they've done a great job of, they

actually have template policies you can just use. And I encourage folks,

there's no, they're literally meant to be copied, right? It is not a case of, hey,

plagiarism or anything. This is literally they are templates meant to be used by the

government.

San Jose has done a ton of work around this right they built out this robust

policy They have a large legal team, etc. If you're a smaller mid -sized city even

a larger County

There's a very rigorous process that's been taken and you can go onto their website

and actually just download that policy So I would really recommend in terms of the

fastest way to do that start there, right? There's already people who have done it.

there's no reason we need to reinvent the wheel, right? This is the whole value of

governments collaborating. And there may be some differences and changes you want to

bring in. I'll say I would start there. And the second piece I would add is just,

you know, throwing a policy document out there doesn't mean anything in and of

itself, right? I would really focus more on the practices and the training. So Micah

Goudine does a great job of this. He has, I think, a whole AI and local government

training course, but he talks about this at length where you could even have the

best policy in the world, and it doesn't matter if it's just a document that sits

there, right? What makes anything effective is how does your organization interact and

think about AI? And that's something that, you know, Micah, there's a ton of folks

who are doing a great job with this. We've We've often been called into, hey,

present to our council, not even on our products, we just do it, you know, pro

bono on our education side, we'll present to councils every once in a while, and

hey, talk to us about how we should think about AI, right?

And I think that is the second piece of it, right? See, if I were, and once

again, I can't pretend know all the all the ins and outs of what it takes around

some of these policy developments. But I would really rely on what these larger

organizations have already successfully shown and then make sure I focus more of my

time actually on encouraging us to use technology, trainings, sandboxes that we could

rely on because that's there's no substitute for practical use.

It's just like, it's just like anything we learned in school, right? Like you can

read the textbook as many times as you want, but if you're going to learn chemistry

or anything like that, easiest way to do is the experiment, right? Is to actually

do the stuff, hey, get your hands on. So that's really what I would say.

Policies are out there. Use them and spend most of your time actually using it.

That's the whole point is the policy is there I encourage you to use those things.

Yep. And ICMA is a member with GOV AI Coalition, and I actually interviewed that

team about a year ago. So I'll link that episode on this one if listeners want to

go back, and we'll also link to that website where these policies are. And as Parth

said, they're literally there to be shared and reused, so don't be shy.

And also, there's just a network of people there I believe individuals can sign up

so it doesn't necessarily have to be a full local government organization and that's

where you can start interacting right with your peers to learn more and again you

don't you don't have to be the IT person or even the city manager just department

heads or team leaders that want to figure out how to how these tools can help

their mission get involved with Gavea Coalition talk to people like Parth ICMA has

sessions and the learning lab has some AI focused courses and we'll definitely have

some more tracks at annual conference in Tampa this October. All right, Parth, last

question. I know it's a cliche thing, but you got to tell us, Matrix Terminator or

Other, if you have a movie or a TV show that's your favorite. And it doesn't have

to be Doomsday. The AI is going to Take over and wipe us all out. It can be

positive. Yeah So what's what's your go -to? Maybe a popular one and maybe like a

Lesser known gem that people might still be to find on their streamers and to

clarify specifically an AI related movie Yeah, or in that vein.

I mean technology doesn't have to be specific specific AI, but In the in that

realm, what's your go -to? Let's see. So my two favorites. I don't think they're

necessarily related to AI per se on this one I mean my favorite movies go well

hunting obviously love that one think a rival does a really interesting job around

Technology and things like that. I would say That's a good question

Actually one that comes to mind which is a movie I absolutely love is an older one

Iron Giant a long time ago was this movie that came out right it's about this this

robot that i want to spoil it but i think it shows a very it's just like large

like at first you think it's going to be like this scary you know robot and it

goes to show what that spoiling is it's a very heartwarming you know charming movie

and i think it just gives a great perspective on Sometimes we make things out that

are new to be scary. We make these things out that we don't know the unknown can

sometimes feel scary to us But I also think sometimes the unknown can wow us and

it can be magical and that's how we've got things that are new right it really

Everything we look around us at some point was unknown and new and I do think iron

giant really captures that wonder and shares this idea that just because something's

made of steel and metal doesn't mean that it can have a heart right and so I

that's one of my favorites I would say I like that I like the positive outlook on

it I think that ties back in the scary unknown even when the leaders or again the

technical people are kind of leading the charge I think there's some not just

responsibility but it's in the best interest for those types to communicate well with

the staff that ultimately are going to be interacting and using these tools and

again hopefully doing their job better. That's right. And not just so it's one more

thing to learn, one more thing to do and eventually this thing's going to replace

me anyways so why am I doing this at all? That's a great point. Yeah I guess this

is like a final almost like call to action or note, I really love that piece,

which is, I know our audience here is mostly city managers and I would say a lot

of folks who are at the city management level, sometimes you're so high level,

right, you're not in the day to day of these departments and it's, it can be

challenging, right, because you don't necessarily feel sometimes the tangible, your job

description is everything, right, you don't feel necessarily the individual things that

are going on and I would say that one of the most valuable things you can do is

you set the tone for the organization and we talked about this at Polymorphic.

I tell folks I will be more upset if we had an idea to try something and we

didn't try it then we then if we tried something and didn't work because that's a

different mindset, right? If you know that, hey, we're going to be an organization

that is going to try things, improve, and I have support from city management to do

that, I think that goes a very, very long way. So I think that's a great point,

Joe. Yeah. And let me as the worker, let me get back to doing my expertise,

technical or not, and help help remove some of that red tape or grunt work,

as we talked about earlier, to let me focus on the stuff that I'm skilled at, and

the bots, the AI, whatever you want to call it, can help with the rest. All right,

well, Parth Shah, thanks for your time and expertise in sharing tangible steps for

local governments to improve their AI policies, and more importantly, the use cases

themselves, which is what we really wanted to get to today. So it's tangible and

not just kind of hypothetical talk around AI so thanks again for your time and the

audience check out the learning lab for more on AI and save the dates for this

October in Tampa where we'll have a few AI sessions. Perfect. Thanks

for having us.

 

 

 

In this episode of Voices in Local Government, Joe Supervielle speaks with AI technologist, Parth Shah, about how local governments can move past AI talk or hype and start experimenting and implementing it to be more productive and produce better outcomes for the residents they serve.
 

Key Takeaways for local government AI:

• Why policy might not be the best starting point... start doing!
• Get in the figurative sandbox and to play and test AI with small teams for real outcomes.
• Official policy documents and templates can be found (and copied!) via the GovAI Coalition.
   

Featured Guest:

Parth Shah – CEO and Co-Founder, Polimorphic
 

Voices in Local Government Podcast Hosts

Joe Supervielle and Angelica Wedell

Resources

ICMA Annual Conference, October 25-29 in Tampa. 

Multiple AI trainings on the ICMA Learning Lab.

AI policy, templates, and more tools from the GovAI Coalition 

Voices in Local Gov Episode: GovAI Coalition -  Your Voice in Shaping the Future of AI