- June 16, 2025
-
-
Loading
Loading
The city of Fort Myers had a problem.
It happened in mid-May. The city’s utility billing system, for an unknown reason, began rejecting transactions for water bill payments. Residents were getting upset and, to be blunt, it was costing the city revenue.
Diana Centeno, Fort Myers’ information technology services (ITS) development operations manager, probed the issue. She dug into the code that controls the communication between the city’s webpage, the payment provider and the financial system. But she hadn’t written the code — that person had retired two years earlier — and couldn’t reverse engineer it.
So, after a frustrating couple of days of dead ends, Centeno decided to try something else. She put the code into ChatGPT — an artificial intelligence program — and typed in what the problem was.
Within minutes she had an answer and a step toward a solution. There was a setting controlling the token length and she needed to look at the token provider.
“She reprogrammed it and everything started working,” says Richard Calkins, the director of the city’s Information Technology Services (ITS) department. “And this happened in a matter of about half a day.”
“That was one of the things where I'm just like, ‘Yep, I'm sold on it.’”
The “it” Calkins refers to is ChatGPT, an artificial intelligence program researchers at the University of Central Arkansas call a highly capable chatbot that “uses machine learning algorithms to process and analyze large amounts of data to generate responses to user inquiries.”
Artificial Intelligence, which took a big leap two years ago with the introduction of large language models led by OpenAI’s ChatGPT, has been around for some years now. Think Siri on an iPhone or Amazon’s Alexa. And for several years, companies, nonprofits and governments of all sizes and varieties have been considering and contemplating if, how and when to use these tools — against challenges that include learning curves and ethical concerns. The City of Fort Myers, with a fast-growing population of about 97.000, is one of the entities embracing the AI revolution.
The city has been experimenting with ChatGPT for a year and just this month agreed to fund a program that will bring the use of the technology to a wider group of employees.
The program allows for the purchase of 150 enterprise licenses for a total cost of about $72,000.
(According to ChatGPT, Enterprise is its business-focused version designed for organizations that want to integrate AI into its systems. It has enhanced security, scalability and administrative control.)
Calkins says that the ChatGPT will be used as a digital assistant of sorts to help staff augment existing processes and workflows. It will also help with research requirements, content writing and coding.
“It’s primarily just going to be another tool in our toolbox that we use to be more productive with our day to day workloads,” he says.
Calkins makes it sound like it will just be a simple helper, but the reality is that ChatGPT will allow employees to do days of complicated work in a short period of time and dig deeper than they’d imagined.
Look at what happened with water bills. A byzantine problem that required a study of minute coding language was mostly solved with a simple search and a good question.
The idea behind expanding the use of ChatGPT is that you can take that computing power and put it to use elsewhere to find solutions.
For example, the city’s police department could input its crime statistics into the program to create a predictive model. Officers, let’s say, working the third shift next Tuesday would know there is a probability they will see a specific crime in their zone at a specific time.
With this in mind, the officers can be prepared, knowing what to look for, where they should be.
“That heavy analysis and heavy calculation stuff is really the wheelhouse of an AI,” says Calkins.
He says a user can input huge document and data sets into the program and have them summarized in matter of minutes. A job that would take a human days to complete can be done with AI in no time.
And, since the system was created to learn, the users can create best practices and get help with complicated documentation as well as build on research to find problems not previously thought of.
“That’s productivity,” he says.
While Calkins lauds the productivity, vast advancements in AI technology have also led to fears it will permanently transform the world’s workforce — and replace people with robots. Others point out that organizations and industries not prepared will vanish.
In a May 19 column, Jim VandeHei, CEO of online news organization Axios, wrote that there is “a stunning lack of preparedness for a technology that could hit every person, every job, every company over the next year or so.”
“America is facing the biggest, fastest, most consequential technological shift in history — at the very moment people have lost faith in the big institutions,” he wrote.
“Making matters worse, most of us feel exhausted before contemplating superhuman intelligence — which is often so unimaginable or scary that it's easier to ignore than engage. Many are jamming their heads in the sand instead of exploring this new frontier.”
(A February report from the Pew Research Center found 55% of workers rarely or never use AI chatbots at work, while 29% haven’t heard of the technology at all.)
Calkins, meanwhile, says, the idea is not simply to use ChatGPT as a tool for employees, as valuable as that is, but to begin using AI to make the administrative side of city government more efficient.
The plan is for the city to begin experimenting next with creating internal chat systems based on its existing data to create digital assistants within various departments.
The development team, for example, could feed ChatGPT (or another AI technology) the city’s policies and procedures, charter and employee handbook to create a virtual human resources assistant that can help employees with questions and issues.
Which gets us back to the subject about whether the technology will replace the people.
Calkins says, no. For one, he says, humans remain needed to input the information and to edit what ChatGPT puts out.
That’s important to note because as powerful as the tool is, it can be unreliable.
In May, two prominent national newspapers published a supplement giving readers advice on what to read over the summer. While the authors existed, the books didn’t. The human involved in producing the list later apologized for not double checking the AI’s work.
Plus, as in the issue with the Fort Myers water bills, Centeno got advice on how to solve the problem but had to do the work herself.
“We do not want this to replace any staff,” Calkins says. “We want this to augment our staff, to make the people we do have be more productive now.
“Maybe what that does is take some pressure off of our hiring, but we're not going to use it to actually replace any positions.”
A group of 30, including Centano and Calkins, has been beta testing the system in a team environment for the past year.
It has used that time to bounce ideas off one another, discuss what works well and what doesn’t and study potential pitfalls. Calkins calls it “going back to that human in the loop part.”
Those 30 will now be rolled into the enterprise environment, which allows for wider use and protections.
Anyone else interested will have to fill out a form which needs to be signed off by their department director. Those chosen will go through internal training and have access to resources on AI they must first review.
That includes knowing that despite all the upside ChatGPT and AI are not all-knowing.
“What we want our users to understand is this technology, while it's been around for several years, is still in development,” Calkins says. “It's very powerful, and it's getting better every day, but you still have to understand that its intent, its goal, is to give you an answer. So we want to make sure it's giving you the right answer.”
