How AI helps Lake County, Florida’s Levar Cooper research, plan and break down barriers

Implementing an AI program can be daunting for any organization. But add in the constraints of working for a small county government and it can feel nearly impossible to keep up with the wave of new technology.

But Levar Cooper, director of communications for Lake County, Florida, has led the charge for not only the communications department but the entire county government to incorporate AI into its workflow.

Using his background as a web developer combined with his skills as a communicator, Cooper has helped his county implement create an AI working group and start using these tools that will define the workplace for the foreseeable future.

Here’s how he does it.

Answers have been edited for brevity and clarity.

How did you start your AI working group?

As we were using it across teams, we were looking at the things that we could do going forward, and when using it, we got a chance to kind of see what the risks were, what it’s good at.

On an organizational level, we didn’t have a policy. So thankfully, being a member of the executive management team, I was able to bring that to the team. And I think when you’re having that conversation, they’re more focused on business, the numbers, the day-to-day operations, and you’re talking about AI, which still sounds a lot like sci-fi. So, it took working with IT, because they’re going to have a pivotal part in this, because AI isn’t just the chat bots in the ChatGPT, there’s so much more that it can do, but a policy is necessary.

So, we worked with (IT’s) consultant. And funny thing is, I used a lot of those materials along with what I understand about the organization to kind of align things, even to the point where I uploaded some of those materials to ChatGPT. Gave it some additional context, added the strategic goals and let it work as our consultant of sorts.

Did you train a special GPT for that?

I did. Essentially, it was taking those materials, those best practices, along with some of the presentations that I developed for the executive team outlining the risks, outlining what an implementation would look like for Lake County. But then also making sure that the GPT was equipped with the organizational goals, like the mission and where it’s trying to go and some of the organizational makeup. And I let that be the consultant. Now, was it perfect? Absolutely not. But it did make something attainable that probably would have taken several months – and it still took a couple of months – but you see the thought that went into everything.

How did you go about training your team, including your executive team?

I try to make things accessible. One of the things I see in this space is people really want to be super technical. And let’s use all the terms: neural networks, and large language models and all. That’s fine. I want to understand it. It’s important that I understand it. But for the person who’s just on the other side trying to use it, and for the executive team who’s trying to understand what this thing is and how we can use it, it’s really not beneficial.

You mentioned that there were objections. Can you share with me some of those objections and how you overcame them?

I would call it more resistance … maybe hesitation in terms of wanting to adopt the technology.

I’m in government. If you’re in Amazon, and you say, ‘Hey, we want to do AI,’ and everyone’s excited. In government, it’s like, ‘Wait, wait, wait. We have to make sure there’s a policy in place.’ In one conversation, it was, ‘Until we get things together, can we just stop people from using AI?’ And I don’t think that’s the right way, either.

So, what we’re really encouraging is sort of an adaptive adoption of AI. We don’t have to revolutionize everything today. Let’s find some use cases. Let’s build some policy around those use cases, and as we continue to expand our use of it, we can further expand the policy based on really thoughtful and careful implementation, and then also working with different technology as it advances.

Tell me some of the things that you’re most excited using AI for, specifically in your comms department.

Our web services department, they’re using it for code generation. And one of the things that we are also mindful of is what we’re giving to ChatGPT. This is still a public platform, so nothing sensitive, nothing that would put us at risk or anything like that. But functions and things like that, it’s really good for. And then (Microsoft) Visual Studio has some built-in tools that will review code.

We have a very small team. Having that aspect to improve the quality of what we’re doing is helpful for creative services for ideation, just coming up with ideas, giving it brand information and then coming up with some concepts to kind of get the creative juices flowing. We also use it for image editing — if you have a vertical image and you need to make it a horizontal image, it can fill in some of that background space. We don’t use it for actually creating images, but for the communications side of things and content, it’s been amazing for us.

Obviously drafting content, and I emphasize draft, because it’s not ready to go. But it’s almost like getting a promotion, in the sense that, I’m not just a writer. Now I can go right to editor. I still have to do all that. I gotta get the talking points. I gotta get all the facts, understand who my audience is, but I can give all of that to ChatGPT and it gives me a draft and I can work from there versus starting from a blank page. That can be helpful for news releases. That could be helpful for even speechwriting, although there are some concerns around that.

Another reason why I believe communicators should be involved with this transformation is the socio-technical risks. Yes, the technology works, but what does it actually mean to your customer? If you wrote a speech after a tragic event, and it was this heartfelt thing, and it hit all the points, and someone runs that through an AI checker, and they find out that it’s written by AI –  what does that mean? How do the public feel about that? And I would bet they’re not going to like it very much.

Other than implementation, the thing that’s been surprising that I don’t hear as many people talking about is the research and the planning. It’s been tremendous for that. ChatGPT has a custom GPT that’s partnered with Dimensions, which is all this research, that’s another thing that we use. We use it for stakeholder meetings. We record meetings, we take the transcripts, put them in ChatGPT and give summaries, sometimes outlining requirements. Quantitative data analysis, and especially qualitative data analysis, which can be really resource intensive.

What would you say to communicators who maybe are earlier in their AI journey than you? What advice would you give to them?

I would say continue to go forward if you haven’t. For anyone who hasn’t used it yet, my advice is to go to ChatGPT, because it’s really accessible. Type “hello,” and then, and then go from there. I really would say that’s a starting point, as opposed to a training. Because when you’re doing a LinkedIn learning or something like that and you don’t have the context, it’s like they’re speaking another language. For those who are further along, I go back to having those conversations. If you know someone else who is working in your space, someone who’s doing similar activities with AI, ask, ‘Hey, how are you using it?’ Because it’s something that we’re all going to be learning about for years to come, because it’s continually changing.