By Katherine Beaty
Last fall, I attended a conference, and it was clear that artificial intelligence (AI) is a subject stirring two opposing emotions: curiosity and apprehension. From the start, a skeptical attendee voiced what many were thinking, sharing concerns about whether AI would take away jobs or prove unreliable. This set the stage for the problem my article aims to explore: understanding the reality of AI beyond the hype and concerns.
Every session I attended, every hallway conversation, circled back to AI. Some people were excited, discussing how ChatGPT was helping them write reports or revise policies. Others were anxious. Would AI replace jobs? Was it trustworthy? And somewhere in between were the myths, my favorite being that vast numbers of people are using ChatGPT as a kind of digital therapist.
That’s why a new study by Chatterji et al., published through the National Bureau of Economic Research (NBER), caught my attention. Titled “How People Use ChatGPT,” it’s the first large-scale, data-driven look at what people are doing with this technology.
So, what are people doing with ChatGPT?
The research team analyzed billions of anonymized messages from late 2022 through mid-2025, using automated tools that categorized messages by topic and intent. Here’s what they found:
• More than 700 million people use ChatGPT each week, or about one in ten adults on the planet.
• Most use is not for work. In fact, by 2025, more than 70% of messages were for non-work purposes.
• The top three use cases were writing, practical guidance, and information-seeking, which make up 80% of all activity.
• Programming is actually a small share.
• Perhaps most interestingly, users report four times more positive experiences than negative ones.
In short, ChatGPT isn’t used for therapy or for sharing deeply personal confessions. It’s being used for thinking — drafting, clarifying, explaining, and making better decisions.
At the same conference, someone shared a story that perfectly illustrates this. They had just helped plan their brother’s wedding and got sticker shock from the cost to hire a DJ. So, they turned to ChatGPT to build a playlist. In a few minutes, they had a 5-hour playlist of crowd-pleasing hits from the 1980s to today. The result? A packed dance floor and a family grateful for the saved expense.
That’s the kind of use case this study captures: not therapy, not science fiction, but practical problem-solving.
Why this matters for parking and transportation
In our industry, the top use cases of writing, guidance, and planning directly address some of our most pressing challenges. For example, clear and well-crafted signage is crucial for avoiding confusion and ensuring smooth operations. AI tools can enhance our ability to draft and revise essential communication materials such as signage, memos, reports, policies, and customer notices.
Moreover, providing practical guidance in explaining rules, procedures, and updates to the public is a common hurdle, particularly during policy rollouts. ChatGPT can help simplify complex policy language, making it more accessible and understandable for everyone involved.
Planning is another critical area where AI can amplify our efforts. It helps us develop clearer ideas, summarize proposals, and ensure effective communication that bridges convenience and enforcement, as well as technology and human service.
That’s precisely where AI tools like ChatGPT can help, not by replacing expertise, but by amplifying it. The study describes ChatGPT as a “judgment support” system, a tool that enables people to plan, write, and reason more effectively.
In parking and transportation, judgment is everything. We’re constantly balancing needs: convenience and enforcement, technology and human service, space and time. Having a tool that can help clarify an idea, summarize a proposal, or even rewrite a confusing customer email is less about automation and more about communication.
Let’s clear up some fears
When I talk to colleagues, I hear a lot of concerns that sound familiar from other industries: “Will AI take over my job?” “Can I trust what it says?” “Won’t it make us lazy thinkers?”
The truth is more straightforward and less scary. As this study suggests, AI isn’t replacing thinking; it’s changing where we start. Instead of fearing how AI might dominate, consider where it can collaborate: Which part of your workflow most needs a thinking partner? This question not only shifts focus from anxiety to curiosity but also respects your ability to experiment and innovate.
Used thoughtfully, AI can save staff time, reduce repetitive writing, and even enhance consistency in customer communication. Imagine using it to:
• Draft internal training guides or FAQs
• Simplify complicated policy language
• Brainstorm new service models or signage
• Translate notices for multilingual customers
Those are the kinds of “doing” tasks that the study shows are already common among professionals worldwide.
From hype to practical help
What this research really shows is that AI isn’t happening to us; it’s something happening with us. The line between tech and operations is blurring, and understanding how people use these tools helps us adapt more effectively, not just faster.
In our industry, that could mean:
• Using AI to analyze feedback and spot patterns in complaints or usage
• Generating more explicit customer messages or social media posts
• Exploring predictive tools that help plan for events or seasonal demand
The key is to experiment. Start small, stay curious, and treat AI like a new colleague, one who works quickly, never sleeps, but still needs direction. To ensure that these experiments yield meaningful insights, consider implementing a system of “weekly AI huddles.” This would encourage teams to gather and share the outcomes of their AI experiments, creating a loop of organizational learning that enables insights to compound across departments. This way, experimentation becomes not just a practice, but a part of our culture.
The takeaway
The NBER paper gives us something we’ve been missing: objective evidence that AI use is practical, creative, and overwhelmingly human-centered.
For those of us in parking and transportation, that’s an encouraging message. It means AI isn’t a threat waiting around the corner; it’s a toolkit waiting to be opened.
So, the next time someone tells you, “AI is coming for our jobs,” you can smile and say: “Actually, it’s coming for our endless stack of parking appeals and maybe even our DJ bills, too.”
KATHERINE BEATY is the CEO and president of Beaty Solutions. She can be reached at [email protected].
