Listen and learn from the best in local government

At LeaderGov we want to equip you to lead well.   These enlightening podcasts, from top local government and industry leaders, will help you gain valuable insights into a variety of leadership, management and teamwork topics, so you can lead better. 

You can also listen to LeaderGov Podcasts via Apple, Google Podcast or Spotify. 

Transcript 

Speaker 1 0:01
Yeah, it is wonderful for to see everybody today. Thank you for participating and being here this Friday afternoon. We're excited about our workshop today. It's entitled AI and Local Government. It's obviously a very important topic that we're hearing a lot about in lots of industries all over the world, and we thought we would take a little bit of time today just to explore kind of the early stages of developing an AI plan and strategy for your local agency, and so that's kind of where we're headed today. My name is Bill Stark, and we'd love to hear from you in the chat on Zoom, so go ahead and chat, tell us where you're from, what city, what county, what agency you're with. Most of the folks on this call are in local government, a few aren't, they're perhaps in fields that serve local government. We're glad you're here. We are going to be recording this Zoom. We're going to be sending out a link to it with some additional information, probably Monday or Tuesday, so we will get you the information back later. Just a real quick introduction about LeaderGov: we serve local governments, that's all we do. We serve counties, cities, taxing commissions, special districts, and the way we serve is through leadership development, teamwork development, strategic planning, and retreats. Those are the four main areas that we serve agencies all over the country. We were in Virginia last week doing an offsite retreat for a big public works agency, and we're working on strategic plans all the time and leadership cohort programs, so if any of that stuff is on your radar as a local government, let us know, we'd be very happy to help you make some progress. We really enjoy, we love working with local governments. Well, going to just share a little bit of an introduction before we have our guest join us. When you all registered for this workshop, you were asked a couple of questions, as we normally do. We kind of want to get your beat on where you are with the topic for these workshops, and so just want to share with you what you said. These are the two questions that you answered. So, the first one was, How extensively are you using AI tools, and that was the question. As you can see, the big one, the green green areas, 48% said we're exploring and testing, so really no deployment yet, and where you're utilizing it the most, the blue slice is 10% so we're using it across multiple departments. 10% of you said so. We're clearly the folks on this on this workshop today are clearly in the early stages, the beginning stages of exploring and looking at ways to leverage AI in your organization, so that that kind of helps set our conversation for today. The second question that we asked you all was, what, what's your big concern, like where are you with all of the different roadblocks, hurdles, concerns for AI, and here the answers were a little more distributed. 31% the green box is around security and privacy, so definitely you know local governments, obviously you keep a lot of very sensitive information about people, particularly water utilities, you got addresses, phone numbers, bank accounts, and all sorts of information that you all need to keep private. So that's certainly important. Uncertainty about where to start or which tools to use, that's 24% of you. The red is lack of clear policies and limited staff knowledge, so one of the things is going to come out of this at the very end is we are going to be offering some services for local government agencies to help address some of these things, and so at the at the end of this workshop today, if you feel like you want to explore what that might look like in terms of creating a plan, creating some training, creating some policies, we certainly would love to join with you in that. Alrighty, well, I have a question before we get started, and Tim, I can't see the chat, but I would love you all to just stop and chat and answer to this as we're getting started.

Speaker 1 4:40
So, when you think about a beyond just, you know, fetching you tasks and doing sort of simplistic things like improving your press release, you know, beyond those tasks, which are super simple, right, super easy. What near-term future applications are. You thinking about, or what are you doing in terms of an application that you could share with us? You know, in your public works department, you're using AI to do better mapping and scheduling of the roadways, or maybe you're using AI to prioritize road repairs, or some other.. perhaps it's, it's more proactive policing, you're using statistics better using AI. So, we'd love for you all just to chat, just for a moment. And if you don't have any, you can just say none, that's fine. But we want to hear from you, if any of you are kind of down the road a bit with a specific application where you know AI is going to either improve your internal processes or improve your service delivery, which is obviously we want to put a lot of emphasis on the citizen. Right, how can we improve service delivery to our citizens? So, yeah, I'm seeing

Speaker 2 5:57
a few notes come in here now. A data analysis, future use assist with facility management and maintenance. City bot for answering calls. Yeah, these are jumping all around here. Capital projects using AI to build out RFQs, RFBs, and opinion and costs for multi million dollar projects. Information delivery to constituents looking for applications to enhance analysis of CRA more data organizational workflow optimization spreadsheets and tracking information document management retention destruction of documents review of medical records and police reports leveraging AI to improve speed, consistency, and quality of procurement and project planning documents, so a lot of data-intensive applications, you know, but it sounds like a bunch of folks are kind of maybe at the very front end of considering how to leverage these tools, so

Speaker 1 7:01
yeah, we'll, we'll pull all of that off of the chat at the end of the workshop today, and sort of aggregate that, and let you know in our summary document what a lot of your peers are saying across the country. You know, I saw a commercial recently for community development, where a developer brings in their plans, their site plans to community development, and the local agency, or this tool, analyzes your existing comprehensive plan with the site plan of the developer, and very quickly looks for gaps between what the developer is asking and what your comp plan is stating, and so that's another really good example of certainly a process, because that's going to save hours of time. So, yeah, thank you all for doing that. We just wanted to get an idea of how you're currently using it. Yeah, and if you

Speaker 2 8:01
continue to think of things, just put some wild things out there, put it in the chat, just some ideas you have, maybe you don't even know how AI can solve the problem, but if you're thinking about potentially how you can use AI in your operation, just put it out in the chat for

Speaker 1 8:15
us. Yeah, Tim, by the way, that was for you all. I'm sorry, Tim, Tim Fenbert with Leader Gov, Tim and I are our partners in in Leader Gov. So, sorry I didn't introduce you, Tim.

Speaker 2 8:27
That's okay, Bill. I'm, yeah, I'm here as an AI bot.

Speaker 1 8:32
Exactly. Well, great. Well, let's, let's introduce our speaker. Really excited to have Nadia Hansen with us today. Just, just a quick story, and then I really would like Nadia to introduce herself. We, we met Nadia, and I met Leader Gov met Nadia several years ago, maybe five years ago, when she was the CIO of Clark County, which is Las Vegas, Nevada, as you all probably know, Clark County, very large county in terms of size, employees, and so forth, but we met and provided leadership development training for, I guess, department of about 300 in it, and so really have enjoyed the relationship over the years, and so when we knew that Nadia was gaining a lot of expertise in this area of AI, we thought we've got to have her come and share kind of the state of the art in terms of launching AI at an agency, so at Nadia, we're very happy to have you with us. Good afternoon, and just tell us a little bit about your background, and maybe where your perspective on technology on AI, real quick.

Speaker 3 9:53
Yeah, sure thing. Well, good morning, good afternoon. Thank you for having me, Tim and Bill. It's great to see everyone. Excited for a Friday afternoon. I didn't think so many people would join, but look at this: we have like almost 90 of you. So great. A little bit about me. So, I like Bill mentioned, I used to be the CIO for Clark County. I've been a public servant for 15 plus years. I work for City of Seattle, State of Nevada, City of Henderson. So, I'm a public servant first. I feel like I've been in your spots where you're at currently. I work at Salesforce. I lead go-to-market strategy. I basically cover all of us, some parts of Canada, which, where I'm at right now, in Montreal, and also Europe. So, I get to see from a very global landscape of how local governments are operating, transforming, doing business, just in terms of in this new AI era, other than that, too, along with my W-2 job, I have the privilege of actually teaching at the University of Nevada, Las Vegas, and Nevada State University. I'm from Vegas, so why not? I love to give back to the community, but the fun part about is I get to see everything from a student perspective. I actually teach AI courses to executive MBA students, and kind of how they can transform their businesses within their individual organizations. So, just happy to be here, excited for our conversation, and like Bill and Tim mentioned, please put stuff in the chat, so we can actually engage, because this would be very boring if we speak at you for the next 2025 minutes. So, hopefully, we can keep this very engaging and answer your questions,

Speaker 1 11:22
yeah. And when we get finished, just with some, I would say, overview thoughts, some insights on AI and local government. When we get through with this section of the workshop today, which will be about 20 minutes, 25 minutes, we definitely want to open the conversation up to your questions, and we'll take those via chat, so in about 2025 minutes, we'll transition over to more of a Q and A, and just have a conversation, and then we'll wrap up before the top of the hour. Alrighty, so yeah, what about this? You know, when we were talking, Nadia, it's like, okay, here we go again, we've had a few major shifts in technology over the last, whatever, 30 years, and so tell us about, you know, these transformative shifts, and maybe what's different about the AI transformation,

Speaker 3 12:14
you know, it's funny, so we've had several waves, there were there, so we started with back what in the 2000s it was the internet. Then we kind of moved towards, oh my gosh, is anybody still using a BlackBerry? No, we kind of moved more towards like mobile apps and an iPhone and Androids. Then it was around the cloud era, like, do we take our on-premise data, should we put it on the cloud? Like, that used to be such a huge question, and I think it still is alive in public sector and education, in terms of whether it has to do with sieges data or data privacy, like, should we move to the cloud or not. So that's still around, but where the phase is that I think, in the at least in the last three years, when with the look, AI has been around since 1956 but what made it super relevant and sticky is the instant gratification that we get of getting the right relevant response back, that's what made it very, very sticky. So we've been in this generative AI, generative AI defined as I'm going to ask you a question and maybe you're going to create content for me, create something for you, but where we're moving towards, and with this is where we're at now, which is what in 2026 I'm sure you've heard the word agentic quite a bit. Agentic, the easiest way to understand it, this means autonomous. So, if any of you have sat in a Waymo in San Francisco, that's fully autonomous transportation. You don't have a driver sitting in the driver's seat, it is driving by itself, so think of it when you're thinking about autonomous tasks in government. So you take a little workflow and say, you know what, I want to, as an example, there's a lot of building officials on the call, I want to take a license application and I want to provide guided navigation, or I want to provide a checklist that's going to go back and make sure that before a customer submits a plan or submits for a license or submits for a permit, that is going to go through the checklist and make sure they've submitted everything in advance, so a technician doesn't have to do it. There's so many different things now that you can take simple tasks, especially if they're repeatable, especially if they're manual, that you can automate, but automate in a very clever way, where it's not just a static chat bot, it actually has intelligence tied to it. So I think that's the era that we're in now, which I'm super excited about, because it changes, especially in government, where we have so many mundane spreadsheets and PDF forms, and we ask the same resident 90 different ways to enter the same info in 90 different portals, we don't have to do that anymore, because we have the tools to be able to automate some of that,

Speaker 1 14:49
so, so, Nadia, these, these all these agents are autonomous tools and instructions are just running in the background all the time, they're. Doing their thing right, we don't have to touch them necessarily,

Speaker 3 15:03
if you allow them to. Yes, so obviously you define the guardrails, you define the topic, you say here's where I want you to automate certain things, but here's where I never want you to touch anything, right? So you can now configure that based on your needs, but if you allow it to do autonomous assets, it can, as an example, Bill. So, I have been able to clone myself on a personal level in nine different ways, like I have autonomous tasks that go and respond to my email automatically, in my tone, in my voice. So, we have, we've, we've gotten the tech that can actually do a lot of amazing things. Think of a call center as another example, right? How many times we have our eight to five covered, but what happens between the five to eight when we need the 24/7 services? So, anyway, there's a ton of examples around that.

Speaker 1 15:51
Yeah, yeah, yeah. Well, we'll get into some of those too in just a minute, but I really like the idea that you kind of framed up the last 30 years or so, and the idea that AI is going to change the type of worker, the type of employee that local governments have, and really the nature of our work, and hopefully less, you know, filling out spreadsheets, right, and more thoughtful work, perhaps over time, so that's really good. Yeah, tell us, tell us about, you know, just not, not necessarily the big towns, the New York cities, but where do you see applications? Where do you see energy and time and effort in what areas in local government are people starting to make really serious investments?

Speaker 3 16:41
Yeah, this I just want to point out, this text on this slide is tiny, so don't worry about it. You're going to get a copy of this, but I kind of just want to hit on a few use cases. So, the biggest bang for the buck that I'm seeing when we're talking about true transformation, not just email summaries, not just data analysis, but true transformation form intake all the way up to actually issuing something or providing the result to the customer. I'm seeing that in 311 services, so anything to do with I have a complaint, I have a concern, I have a question. I don't care if I report this on WhatsApp, I don't care if I report this through my mobile app, I don't care if I call somebody and talk to someone, but there's this very omnichannel approach where let's try to meet the customer where they're at, where we can provide this Amazon, this Uber, this Airbnb type of experience to our constituents. A lot of you in the chat, I noticed, mentioned tools, and I think one of the things, if we can backtrack, if I may, a lot of the conversation really needs to start with what's your AI policy and what's the strategic plan. What outcomes are we trying to solve for? The technology part can do what you want it to do, but at the end of the day, it comes down to processes we want to revamp and the people we want to bring along the journey. How you're going to bring the people along. So, we're seeing a lot of use cases around public safety. There's always been predictive analysis. There's always been predictive analysis, and some of the agentic facilities when it comes to transportation, traffic mobility. How do we predict when the traffic is going to be more at 5pm So, how can we route people accordingly? Then it has to do with automating permitting inspections. I can't tell you the number of times where I, as a, so I have my own AI adoption company as well that I do on the side, because I really am passionate about helping people understand the power of AI. But essentially, what if somebody came in and they're looking to get a permit, let's say a construction permit, which is extremely complex to get. There are numerous steps that they have to go through. 90% of the time, they'll bring a plan, a physical plan, to your office that somebody has to go and review. Half the time, that plan is not quite there yet. Maybe they've missed certain steps, maybe they haven't identified that there's hazardous materials on the property. All these little things you can actually now do with AI, but it can do the checklist for you. It can make sure that they've submitted everything and guide the customer half the time. As another example, people don't know what their NAICS code is. So, if you've ever applied for a business license, there's always a question: is like, what's your category of your business? Residents don't know that, nor do they care. Another example, I want to pay my property taxes. Do I care whether I pay to the assessor's office or do I need to know? Do I pay to the treasurer's office? They both collect property taxes, so there's easy examples where you can make the services so accessible to our residents, to visitors, to constituents, where they don't need to know. We can make it very, very keyword basic, where somebody just asks a question, like you do in Chat GPT, like you do in Claude, but imagine this on your website, where it's a services-based approach rather than department-centric approach. I can't tell you, and I work with all kinds of government agencies, not just across the US, but to the entire globe, and. We have this conversation on repeat. Why should anybody know how government functions? They don't, they don't need to know, and nor do people care. I don't even know what public works does as a resident. I don't care what comprehensive planning does, because these are very governmenty terms.

Speaker 3 20:14
So, I would say the easiest use case is make your self-service knowledge base easier to find and access, so if somebody puts in a keyword like I want to complain about something, bring up that portal so they can actually complain, right? So it's easy things like that, but I'll pause here, and we can go on forever about different use cases, but these are some ideas of how it's actually being used to transform operations, not just provide an email summary.

Speaker 1 20:41
Yeah, yeah, surely, in, you know, every department there are lots of things being considered, lots of things being worked on to create smoother process, and I like the idea that you're really asking the question, How do we make that citizen experience better, and focusing on that on the service delivery aspect of local government. There are a lot of internal improvements. How do we produce a financial report faster? Right, that's a process improvement, and that's that's good, that's nice. Finance folks love that, but looking for ways to really transform the way the citizen interacts with local government, certainly is a great place to start. So, thank you for that. Yeah, let's transition to security and privacy governance for just a minute. That was the number one thing in our survey that people were concerned about. And what do you, what are you seeing out there in terms of a focus on this area?

Speaker 3 21:44
The doctor chat just put it best in the chat, honestly. Like anybody who is working towards allowing people to test and play in a safe environment, that's been working out quite well, but you need a policy to be able to accommodate that. So, as an example, some of you had put in the chat that you're using Chat GPT. My question would be, are you using the free version or are you using an enterprise version? Because there's a big difference. If you're using a free version of anything, you are the product, so your data is going to be used to train these large language models that can produce the results you're looking for.

Speaker 1 22:21
Okay, say that again. Hold on. If I'm using the free version of Chat GPT, whatever I enter, whatever I'm doing creates the learning that's in the global database. Is that right?

Speaker 3 22:33
That is correct. So, anybody at some point it can come up. So, let's say if you're putting customer information, if you're putting sensitive information like PCI data, health and human services related data, HIPAA data, sieges data, I don't care what kind of data you're putting in that has any kind of compliance or regulatory measures. If you're putting it in any kind of a free version, you are the product, so your data is going to be out there for the world to use, and it's used to train the model to give better results. So this is what I number one is what. When you are evaluating different vendors, the security, the trust, and the privacy question needs to come first. Do you use my data for training your model? So that's number one. Number two is, what is the back end look like? Like, where does my data go? Where is it stored, and how is it utilized? Is it just. is my data in a trusted realm, but only I get to use it, and I get to produce the relevant results that I want from it, or is it being used somewhere else? The second thing I'll point out is, if you don't have a data governance strategy, and what I mean by data governance is, as an example, we're talking about, I'll take a health and human services example. You have sometimes nine or 10 different agencies that you're working with. It could be a nonprofit, it could be a grant system, it could be different providers, it could be your family services, it could be child services, it could be social services. Sometimes that are captured in multiple systems. If you don't have a data governance strategy that says this is the data we're going to pull from each system. Here's how it's going to be utilized. Who are the stakeholders who's going to be touching that? It doesn't matter how good of a technology you use, you're not going to have things in place. So one, have an AI policy of what we're going to utilize, how we're going to be using it. So you set some guardrails, and then two have an accompanying data governance framework in place, so your agency feels okay. We have addressed the data security aspect of it, and we also know where our data is going to reside and go. Those are extremely important before you launch into any of these other journeys. Then, when we start talking about use cases.

Speaker 1 24:41
Yeah, so that's sort of step one. Ground level step one is to get those things in place. Tell us about the policies and the plans that you see out there. How are people creating plans and policies around AI?

Speaker 3 24:59
I. Um, I had the privilege of attending the US Mayors conference last week. It was in Long Beach, California, and I'll give you an example of that. We were all sitting in a room together, and some of the examples that came up were wonderful. So, like, as an example, Boston, Seattle, San Jose, New York City, Colorado - they all have some kind of an AI policy already put in place, you're welcome to go and take a look at how they've drafted it. You don't have to start anything from scratch. You can actually borrow, but then you customize it to what your agency needs are. So that's one. Two, a lot of mayors actually mentioned how they are bringing AI into government is by allowing employees within the organization by answering the question, what's in it for me. So, it's great that you have a plan, it's great you have a policy, but if your people are not allowed to use any of the tools, and they're just allowed to use, maybe let's just, as an example, co-pilot, there's going to be very limited use, you're not transforming anything, so it's very important to allow for the safe space for people to one try out different tools, and then say, hey, now that you've used it for any kind of personal adoption, it doesn't even have to be work adoption, and bring your people into the fold, so they can feel the change themselves, and then you roll it out to the organization, but allow for these safe forums, so that it comes straight up from leadership, it's not going to be bottom up. One of the things I wanted to touch on is I am seeing, so if you don't have an AI policy in place, if you don't think people are using AI, you're very.. because everybody is leveraging what we call shadow AI. They all have some kind of a tool, whether you allow them or not, because all these are web browser versions now, that you don't have to download anything, like you can just use it on your phone at this point. So people are using it because they want to accelerate in the manual job, or in the manual steps they're trying to do, and make it a little bit more efficient. So just kind of keep those guardrails in mind that if you don't have a policy, it doesn't mean that people are not using it. They probably still are.

Speaker 1 27:03
Yeah, Nadia, I was going to ask you about that, because when we, when we say the word plan, you know, in local government we have a comprehensive plan, transportation plan, strategic plan, all these different plans, which are really awesome, because they provide a pathway for us, and I think AI is no different, you know, you mentioned that the shadow users, I think if an agency can have a plan that says, you know, what we're going to first focus on a few low-hanging fruit process improvements internally, that's part of our plan. We're going to roll this out to our leaders this way, that's a plan. We're going to teach them these tools, that's part of the plan. We're going to prioritize our citizen-facing application improvements, ABCD. That's part of the plan. So we do, we agree wholeheartedly that having a thoughtful, progressive plan, and how to educate the employees, equip them with ideas and thinking, put the guard rails up, put the policies up, and then have a prioritized list of the areas that we want to focus on as a local government agency. Maybe start small and build up to the more complex, but, but that really is a plan, that's that's a set of goals, a set of priorities, a set of objectives that your local government agency should have. Otherwise, it's kind of the wild west, right? We're all kind of doing our own thing, we're sort of off with different tools doing different things, and the organization doesn't learn together through the process. And so I really like this idea that you brought up of not just a policy that's part of it, but a larger plan that the agency can can work through over a year or two period,

Speaker 2 28:49
and Nadia is the problem being amplified by the fact that our all the software that we're using, even in our departments now, has AI components that we either turn off or turn on it just seems like we're being inundated by these opportunities, some of them built into our software,

Speaker 3 29:08
correct. And I think one is what type of software, and not every software can do everything you're looking to do. So I think sometimes I hear a lot of people say, oh, I'm using Copilot, but it doesn't create PowerPoint presentations for me, or maybe I created an agent, which is, if you don't know ages, they're basically your digital buddies in the flow of work. These could be very autonomous if you want them to. So, it really, it's less this conversation is not about what tool should we use to do what things. It really starts with what's the outcome that we're trying to solve for. Have we benchmarked what our existing KPIs are like. Well, if we're, if we're issuing out a permit that takes us 30 days, are we trying to now issue a permit in three days? That's a KPI, right? So, it's like, where are we at now? Let's baseline it, but then when we implement AI, how do we know that it's actually working? So, I have so many conversations with with each. Seeds, where they're like, hey, we tried a pilot for three months, and now we decided that we're not going to roll this into production, we're going to try another tool. Well, that's great. So, you're just in tool land, so all these initiatives are just going to be static for the longest time, because you actually don't know what you're trying to, you didn't have a baseline for what you're trying to solve for.

Speaker 1 30:17
Yeah, you know, I think that's part of this idea of the plan, because it's identifying where are we today with our service level and where do we want to be. What's our, what's our goal, and so thinking with the end in mind, but that's a part of a planning process, and I'll say to even Nadia Leader Gov, we're a training company, right. We're asking ourselves, how can we leverage AI to train people better, and for instance, we're thinking about, like, nobody likes appraisals, nobody likes giving negative feedback, nobody likes giving credit, nobody likes holding people accountable, right. So these are skills that people in our business, and that leadership business, we know these are skills that people aren't very good at, and they don't like doing them. So we're thinking about building a program where a local government leader can actually record a conversation with a certain profile of a person, a new employee who doesn't, who didn't do some something they were asked to do, and now I have to hold them accountable, and so we're thinking about building a program where it's like a practice program, right, I get to practice doing appraisals, and until I get, get the tone right, get the message right, or I practice, get to practice giving feedback, so that when I do give it live to a human, I've had a chance to go through and try it, try it a few different ways, and maybe even get a grade from the system, grading me on some things I left out, or some things I should have put into my conversation, so even in our business, which is training, we're asking this question: How can we make our, our customers, the local governments better, and how can we ensure that they really have the skills to do what what they want to do? So, lot of opportunities. Yeah, and if you all have the need for a plan, let us know. This is something that we would love to collaborate with you on. So, how about this, Nadia? Just wrap up here in the next few minutes, and then we'll transition over to some questions here in just a moment. Best practices.

Speaker 3 32:38
Oh my goodness, so I have so many of these, I just picked four, but I'd love to hear from others in the audience too. So, number one, I came from public service, like I mentioned, the biggest bang for the buck I see is I still see job descriptions that are maybe 20 years old that are just not forward facing for the new generation. So, in government, we, what, serve five different generations. We have traditionalist baby boomers, Gen Xs, pretty soon Gen Zs. Guess what? I have an 18 year old. My 18 year old has zero desire to work in government because she's.. I don't know what I don't understand, like what kind of a job. What does a programmer analyst who mean? So a lot of our job titles and job descriptions are not quite there yet, so I would say, if you really want to become an AI-ready government, take a look at some of the how the jobs are shifting and what the job descriptions could be like, so you're trying to attract the right talent into your organization, that's one, two, is please invest in upskilling your people, I cannot tell you how many mayors, elected officials I recently have spoken to who are trying to invest in their people first, and every time I say don't do one of those where you have them mandatory watch a video, mandatory go through like 10,000 PowerPoint slides, because that's just extremely boring. The only way you're going to get people to upskill is if you answer what's in it for me. So I always start with personal adoption. Just set up a forum, set up a training condition where people, somebody comes in, like leader gov, and basically, like, okay, you know what, let's try to solve this one use case together. It's hands on keyboard exercise. We're going to build an agent together, as an example. Oh my gosh, we're going to take this one use case that you have so much trouble with that it takes you nine hours. Let's see if we can like bring it down to one, but you want to work on something tactical, which is going to be part of your training plan, so everybody has their hands on keyboard. Once you have that, then you can start addressing little things. It's like, okay, you know what, we are going to tackle, like you mentioned, Bill, what are some of the pain points that our residents have? Oh my gosh, they have no idea where to go search for information. It's like very basic things that people are looking for. I either want to ask somebody about a question, how do I about something? Let's take one process and make that super accessible for somebody using a. Which can provide relevant information, but not just a static, frequently asked question that we may have not updated in the last 10 years. So, there's a lot of different examples like that, where honestly, people investment will go a long way, because a lot of employees, frankly, are scared. It's like, well, will this replace my job? No, this is all about augmentation. How can we make your job way easier, way better, way faster, so you can work on more complex things that actually require your brain, because the manual repetitive tasks may not require your brain anymore. These are just things that AI can do for you. So, let's try to list out those things up front as part of it, and then see. Okay, let's build it together, but do it with your teams as part of that, so it's not in isolation. And I love all of comments that Dr. Chadwick is putting in the chat. It's like different departments adopt different tools. Yes, they will do that, because we don't have a standard across the organization. Every department wants to do their own thing, especially if it's an elected official department.

Speaker 3 35:59
They don't need to go and ask a central body to be able to do what they want to do, so there's a lot of different things that come with it, but I think if we really had to boil it down to please take a look at forming some kind of a data governance committee, an AI committee include different stakeholders, you have your legal included, your procurement included, HR is included, it is included, securities included, and the different business representatives as part of that, so everybody has a voice on the table. And then you develop the policy, but you can't do this in isolation, because that is just not going to remain long term.

Speaker 1 36:34
Thank you. That's good. I like that policy, a plan, training, upskilling a committee looking at those opportunities to streamline certain services for citizens, even the job description idea, pretty pretty basic best practice. Some, some ground floor things that we can do to take a next step. I, you know, it's exciting to hear in the chat what so many people are doing. It really is, but I wonder, how many agencies actually have a plan for what they want to do, as opposed to just doing things, which is fun in and of itself. So, really, would challenge you from an agency perspective. Do you have a strategy for AI? All right, so we're going to shift over. Let's see, it's almost 40 after the hour. We're going to shift over and for about 10 minutes take some questions from the audience, so if you all want to go ahead and post any questions in the chat, you're welcome to, and we'll have Nadia share some of her perspectives, what she's seeing around the country in terms of AI. So, Tim, what are you thinking? What do you question

Speaker 2 38:01
for Nadia? Yeah, we're waiting for some things to come in, but hey, Nadia, I know that you mentioned that you're teaching at UNLV and University of Neda and Executive MBA students. How are you? Have you thought about how you might be able to transition some of that training to local governments, and what's kind of an approach that you might want to take to that?

Speaker 3 38:21
You know, so it's funny, a lot of the conversations around piecemealing, one they always talk about, how do I get my leader and how do I bring people on this adoption journey, because my leader doesn't quite get it, and I don't know how to convince them. I think we're kind of beyond the convincing part. It again, it comes back to education, so I always start with, you know, what. Why can't we host a art of the possible day? And this is, oh my gosh, if we had to reimagine what our city government, what our county government would look like in the next five years, let's come up with all kinds of creative ideas. These could be crazy ideas, like, oh my gosh, we can have, like, an avatar which answers questions for people, which is a live bot that sounds human, but it's grounded in our data. That's one example. Those kinds of open-minded sessions, which is very much focused, community, community-centric, you know, like a focus group that you set up that allows people to be like, oh, you know what, I can actually voice my opinion, and it's going to be okay. So, a lot of these conversations are really more around adoption and less around technology, so I always come back to people process policy. It's really that kind of a segue. Vazim actually put a great comment in the chat, and he's like, I'm an elected official from the Caribbean, and we have an issue of AI being adopted, and I'm taking the initiative to develop AI chat bots, but we still have the issues of digitizing documents, so you know it's a great question, and I want to say AI adoption again needs to start at a personal level, like I would say, Wazim, have we provided an environment where you maybe do like an AI training day in a second. Everybody's going to come with one thing that they absolutely hate to do, absolutely hate to do, and let's see if we can automate that, and if people can bring that into a very hands-on setting where it's meant to be a little bit more of a conversation that allows them actually to feel a bit safer, to be able to adopt, I don't know how you guys feel about like having an AI day, I call it AI day, or maybe it's even an art of the possible day, where people bring in ideas to the table, and then you kind of have a priority list that's set up. Okay, we're going to start with this one first. What can we do to automate this process? Because AI is not the answer to everything. It will be the answer to some things where it makes most sense, because you have to remember the cost of tokens when you're asking and prompting and using it. So, you have to kind of keep the costs in mind as well, but kind of bring a partner who can kind of help you phase it out in an approachable way, where it'll be in bite-sized chunks for people, so they don't feel, oh my gosh, like you're trying to change everything that I do in my day to day that I've been doing for the last 20 years. And yes, Lydia made a great point about change management practices. It's all about or change management. This is a.. I feel like in the era that we're in, this is less about technology, it's really more about organizational change management. How do you bring this wave where people still feel comfortable about it, and they're going to be willing to adopt, so you can move forward.

Speaker 4 41:27
Yeah, yeah. What I was going to

Speaker 2 41:30
ask, real quick, there was a comment about forming task force, and I think probably for a lot of people on the call that might be their first step is to form a task force. How do you sometimes come along a task force, how could you, Nadia, support a task force that's created? Because the most folks might know a lot about AI, but some people in the task force may not know very little.

Speaker 3 41:50
Some of the best AI committees I've seen include public-private partnerships, so sure you will have all the public government government agencies, maybe you'll have education as part of it, maybe we include the university as part of it, because at some point you want to build a talent pipeline as well. So, the more I mean, I would say don't have like way too many representatives, you can never get to an answer, but you want some perspective on what is happening in from a landscape perspective, how fast things are moving, but keep it centered toward what is the end goal that we're trying to achieve, so be very, very specific on who you want to pull in, but I've seen the best ones be where you've included legal, where you've included procurement, where you've included HR, and bring a private and sector entity who is going to be a little bit non-biased, it's not a technology conversation, again, not a tool conversation, but more out. Okay, what is going to be the best practice based on where we're headed? And I, when I say future, I'm not talking five years from now, I'm literally talking one year from now. Things are going to look dramatically different. I mean, just in the last 18 months, look how fast and far we've come. We're all talking about agents. I mean, I work for Salesforce. We don't even send emails anymore. Everything is is done on Slack, so things are moving extremely fast in the tech realm of it, but we have to bring people slowly in that journey as well.

Speaker 1 43:11
I know I liked what Dr. Bacon had said about the comment here, and we'll put this in the summary as well. Indian River County, Florida, he just stated his policy, their AI policy, which is up in the chat if you want to look at it. Ray Cruz, our friend in California, says you mentioned the need to change job descriptions for AI related to positions, but government has, however, always been unable to pay the bright tech employees. Is contracting the only realistic option, since we cannot compete with private sector for tech talent. Wow, that's a good one, Ray.

Speaker 3 43:47
It's a great question, Ray. So, I'll give you an example. Yesterday, Anthropic, which is Claude, if you've ever used Claude, one of the biggest, largest research labs, it's comparative to Open AI's Chat GPT, they released a job called, I forget the exact name, but I think it's called Claude for Beginners, and they are hiring people at $85,000 a year, where they're going to train them on their AI model, and they're going to place them into nonprofit agencies. Now, coming from government myself, I think the salary ranges are just around that, so if you created a position, you don't have to pay them a zillion, gazillion dollars, because you're going to partner with somebody who already has, who's paying for engineers from a private sector side of the house. What you want is somebody like an AI automation engineer person. So, when I say revamp, think of you don't have to pay them a lot, but make it to where they would understand what the job is going to be. So, if you're going to be attracting younger talent from universities, they will understand an AI automation career. You can pay them $85,000 exactly the same as what these large research companies are paying, and you get them to come. And this is their only job that they're going to focus on, but what I meant by updating the job descriptions, I still see a lot of like mainframe jobs out there, I still see like these cryptic IT solutions analyst, it's like what does that mean, so I think just try to be very specific within the same job descriptions, and I know some of you are unionized, so you have to go through that process, but start that now, but include some language of, hey, we need artificial intelligence. We don't need 10 years of experience, because to be frank, nobody has it. So just put in, like, hey, we need somebody who is open to learning, who has done, who maybe build some automation, personal automation on your own to come into the organization and help us build it, make it very simple for people to understand, so it's not in government speak.

Speaker 2 45:46
Hey Nadia, I know you're going to tell us about how we can connect with you on LinkedIn, and maybe get in your, your, your database for some mailings, but back to this committee. If I formed a committee in my agency, and how could you, I'm gonna put you on the spot. How could you support us? Could you come alongside us and do some training? Or

Speaker 3 46:06
absolutely part of our

Speaker 2 46:08
governance conversation.

Speaker 3 46:11
I do this all day long with many. I only work with government organizations, because that's all I know. I come from government, so happy to come in, partner with you, sit and walk through, do a discovery, be like, okay, here's where we're at right now, here's where we want to be happy to help with that.

Speaker 1 46:27
Yeah, I like what Andrea Claus just suggested, something in the chat. I've been recommending to recommending short free basic courses to my board and employees to get them more familiar with AI. Part of the challenge is the fear of the unknown, so yeah, everything that really you've been talking about, Nadia, is this idea of getting the conversations out of the open, talking about them, talking about use cases, keeping it simple, making it accessible, so people can touch it, feel it, experience it, and all of a sudden the fear level comes down, so at the end of the day it's, it's some sort of a plan to do all of that, some sort of concerted plan to do all of that, because without a plan we tend to scatter and go to our own corners. Unfortunately,

Speaker 3 47:19
I had an agency, I'll give you an example. I had an agency which had a very tough employee group who was completely closed to anything AI. So, what we did was we actually put together a workshop for them, which was like, hey, you know what, why don't we, Susie, as an example, Susie loves to travel. Susie, you're going to be going to three cities. How about we plan what the itinerary is going to look like by just asking AI to build something for you. I'm talking about these kind of relevant examples that change people's lives and makes it a lot easier, so Susie doesn't have to go and figure out,

Speaker 1 47:54
yeah,

Speaker 3 47:54
by putting in eight hours a day to what cities am I going to travel to, and what food am I going to eat, and what am I going to wear, right? So all those little things come into play will make ask good question, and I want to come back to that. So I talked about tokens. I don't know how many of you know what tokens are. So if any of you have been playing with, let's say, Chad GPT or Claude, any of those open free tools, even the free version, the number of questions that you ask, which is called a prompt, the number of questions that you ask at some point. I don't know if you've hit a wall, they're like, "Hey, you're out of tokens at this point, you have to come back the next day because you've asked x amount of questions, and we only give you this much for free. Those are called credits or tokens. Now, depending on how much usage you have with a lot of organizations, if you don't have, if you're not budgeting for these tokens, so let's say if you have 100 people in your department and all of them are using AI, you have to budget for that, because at some point, depending on the number of questions people are going to be asking, especially if we open it to the resident community, you need to understand what that budget is going to look like, so that is part of the strategic planning session. Like moving forward, I always tell you're going to have a token budget at some point that is going to be a capital line item in your capital budget. Like, how, what is going, what is that going to look like? So software is going to look inherently very different within the next two years, so start thinking about those things.

Speaker 1 49:24
Yeah, Nadia, I'm building an app for a nonprofit that I work with. It's a, it's a user app, it's a consumer app. Right now, it's a web app, and I've got this tool, Manus, which is part of Meta, and so Manus is building this app for me, and I just tell it what to do. Okay, build me this, and have it talk to this, and output this, and I just type in what I want, and it's amazing how fast I run through 4000 credits. And then, and then Manis wants $60 and then Manus wants $100 So, you're exactly right, it's hard. To know the number of questions you're going to have when you're building something from scratch, it's really kind of a mystery, but it's, it's very economical. I mean, it's so cost-effective and so fast to get these apps built, but running through those credits is in the budgeting part is something to keep in mind. I've learned the hard way. We're going to wrap up here in just a second, and I wanted to just kind of engage you all in a couple of things. If you would like to connect with Nadia, she would love for you to be a part of her LinkedIn community. If you want to scan that on the left, or you can just go to LinkedIn and search for Nadia Hansen on the right. You can scan it as well, or go to Nadia hansen.com/forward/community She offers a lot of newsletter content and best practices, and sort of ongoing dialog about the AI industry, particularly from a governmental perspective, and so please connect with her, link to her on LinkedIn, and you'll get signed up for a lot of her really cool educational stuff. A newsletter, I think, Nadia, is that right? Monthly, a quarterly. Yes, I have

Speaker 3 51:16
actually a weekly newsletter that I write, but it's more on personal adoption, like here's how I use Claude, and I've built, like, nine apps on it, and I share around data and security and privacy, and all those good things. So

Speaker 1 51:27
nice

Speaker 3 51:27
love for you to be a part of it.

Speaker 1 51:29
Yeah, yeah, I really encourage you to do that. And if you are at an agency and you're scratching your head, saying, where do we start? What do we do? How do we train our people? How do we get adoption? How do we break down the walls of fear and resistance? And you would like some help, and want to just have a discovery phone call, where we can just get on the horn for 30 minutes and talk through where you are and what you think you want to do. We'd be very happy to have that conversation with you. You can put your name or email in the chat right now, or you can email me directly, or you can DM me inside Zoom, and I'll get a message, and Tim and I will follow up with you,

Speaker 2 52:11
and Bill, not to scare people away, it's not just you and I, it's actually Nadia as part of our team as well.

Speaker 1 52:16
No, completely,

Speaker 2 52:19
nobody wants to talk to me about AI.

Speaker 1 52:21
Yeah, exactly. So we're very happy to help you take that next step in organizing your approach, whether it's through the training or policy, any of those aspects. We'd be happy to help you kind of get things off the ground, so to speak. And one other sort of just housekeeping thing, you all know that we, we facilitate these free workshops about every four to six weeks. Our next one is July 10, and it's going to be around the topic of culture with seasonal employees, so think of it part-time, think of interns, even elections, part-time volunteers, we're going to be talking about how to, how to set the culture so that those seasonal employees, interns, part-time employees really want to come back and maybe work for you full time, and how to bring them into your culture and care for them, even though they're a part-time employee, and make them feel special, and make them feel like they belong, and what they're doing is important. So, we got a really, really special guest that's going to be facilitating with us next month, and we're excited to get into that topic. So, you can register, I think, starting tomorrow, you'll be able to register for our next workshop, so if you know public works folks, city managers, county managers, parks folks, we'll be talking about parks a good bit, that's a whole nother different topic next month, Friday, july 10. So any closing thought as we wrap up here in a couple minutes? Nadia, you're love to love to hear from you, and yeah, if you all want to stay on a few minutes after the top of the hour, we'd be happy to chat with you. Anything you'd sort of wrap up with Nadia?

Speaker 3 54:10
Nah, thank you so much for having me. It was great to see so many. I feel like my public service peers, thank you for your comments in the chat. I feel like I learned so much from you, and where people are at, but like Bill mentioned, please connect with me. Happy to share more strategies. I'll be part of Bill and Tim's discovery calls if needed, and I would love to kind of don't be scared. I know this is a big forum, so sometimes people don't really open up as easily, so please DM us and ask us questions, and happy to help along the journey.

Speaker 1 54:41
Yeah, thank you. I hope you have a great weekend, and we'll see you in July. Take care. Good luck to you, and stay on if you want to chat for a minute.

Speaker 3 54:50
Thank you, everyone.