When artificial intelligence research firm OpenAI hard-launched its generative AI chatbot ChatGPT in 2023, the technology went from an esoteric Silicon Valley product to a mainstream skeleton key for information and content creation − right at our fingertips.
But AI in our daily lives isn't new, even if it feels like it. A 2024 survey found that while nearly all American adults use AI products on a regular basis, 64% of them don't realize it. It's in your Amazon products, your weather apps and social media platforms, which have all been in public use long before ChatGPT.
By making generative AI more accessible, ChatGPT changed the technology game, and most of the world is still trying to learn the rules. But generative AI has already evolved beyond
simply aggregating massive amounts of data and into the next frontier: agentic AI, which can plan and reason to complete multi-step tasks on its own.
"I have students that say that they're not going to use [generative] AI, that they're against it," said David Staley, a history professor at Ohio State University specializing in digital history and technology. "Terrific. How long is that going to be possible? Or will systems be designed or redesigned such that you won't have any choice?"
Earlier in September, both company whistleblowers and established leaders within AI companies sounded the alarm over the risks rapidly evolving AI poses to humanity, and called for increased regulation and a slowdown of AI development.
To try to break down how AI is impacting our everyday lives, The Dispatch interviewed central Ohio experts across different industries and specialties on how they think AI is shaping our day-to-day, not just the wider tech world.
Concerns, hopes, and approaches to AI guidelines differed from expert to expert. But all agree on one thing: AI isn't going anywhere and we have to learn how to live with it.
Andy Lonsberry: AI will help bring manufacturing back to Ohio
Andy Lonsberry and his brother Alex grew Path Robotics out of a love for building things with their dad in their Hudson, Ohio, garage into a company powering welding robots with AI in Columbus.
Lonsberry's background is in mechanical engineering and machine learning. When founding Path Robotics in 2018, he asked dozens of U.S. manufacturers what their "pain point" was. Overwhelmingly, Lonsberry said, the answer was welding. The American Welding Society estimates that the United States needs over 300,000 welders by 2029 to keep up with planned and ongoing construction projects.
It doesn't help that manufacturing jobs steadily dropped for decades as the United States outsourced production overseas for cheaper labor. That hit Ohio, a state with a long history of manufacturing in oil, steel and cars, especially hard. It's estimated that Ohio lost roughly 300,000 manufacturing jobs since 1999, according to data from the United States Bureau of Labor Statistics.
But Lonsberry believes robots powered by AI are the way to bring manufacturing back, not just into Ohio, but the whole country.
"We've talked to customers that are shutting down or were going to shut down their welding businesses here because they couldn't keep it staffed," Lonsberry told The Dispatch. If those businesses go away, their services have to be found somewhere outside the U.S., but Lonsberry noted that for decades, it already has.
His robots, he said, don't take jobs, but create them, or at the very least, keep them. Finding people to do the hard, manual labor of welding isn't easy, hence such a dire shortage, though some industry experts in part attribute nationwide manufacturing job loss to AI.
Lonsberry argues that AI makes robots more adaptive and able to complete more tasks, which can help fill the wide labor gap. However, the robots still need people to build, program and oversee them. Those jobs can more easily be filled as using skilled robots can cut down costs that come with manufacturing overseas. And while there may not be 300,000 more welding jobs on the horizon, there are more roles for those in robotics and other engineering specialties alongside those who understand welding.
"In our industry, there is this huge concern that robots are coming to take our jobs, but we're trying to say that robots are here to keep your jobs and help grow the industry," Lonsberry said.
Detra Price, Ph.D: How AI is changing learning
Detra Price is no stranger to adapting to change. She's executive director of the Center for Digital Learning and Innovation, as well as a professor at Ohio State University’s College of Education and Human Ecology.
But trying to integrate fast-developing technology safely and thoughtfully into any education level has been like trying to jump onto a moving train, she told The Dispatch. AI has been touted as something that will revolutionize education, she said, but there's "just so much we don't know."
Still, the train is moving. The Ohio Department of Education and Workforce required all Ohio public school districts, community schools and STEM schools to adopt a formal policy on classroom AI use by July 1. Ohio State University is implementing AI fluency into all of their undergraduate degrees.
In the limited research on AI impact in education, results are mixed. A Stanford University reportfound some benefits to using AI tools in K-12 education, like teachers being able to create material faster without reducing quality or students having easier access to information. But the report also noted students didn't perform well in subjects they'd improved in using AI tools once those tools were taken away, or using the tools failed to foster "deeper thinking."
Price doesn't believe that AI in every form hurts learning. She's seen it help get to learning outcomes faster by bypassing more tedious work that can demotivate kids.
"If we are tuned into our young people, we know that they want to know that their life matters, that what they're doing matters, it has purpose," Price explained. It's tempting to offload work you don't think matters to generative AI, like an essay on a book you didn't like. But shaping assignments to address problems students care about, or "challenge-based" learning, can help reengage them.
Price previously did a project with students of Indianola Informal K-8 School to identify a challenge a family member had that could be resolved with a form of AI known as an image classifier. Students built app prototypes with help from engineers to detect poison ivy, determine if an avocado was ripe or what candy may have nuts in it via images.
While figuring out how to wield AI is important, there's a line. AI should not be used to determine discipline for a student or decide whether they should advance a grade or be held back, Price noted. That's a conversation between parents and teachers. There are also concerns aboutknown racial bias in AI hurting students if used these ways.
"If you value ethical and transparent use of AI or technology in general and data privacy, then you will say 'because I value this, here are all the guardrails I'm going to have in teaching and learning in my school district," Price said. "...So what do you actually value? How does that show up in the way you're using the tool?"
Dave Billiter: AI could revolutionize cancer diagnoses
While health privacy with or without AI woven into it is a growing concern, Dave Billiter is asking whether sensitive data, collected responsibly, could save lives.
Billiter is CEO of Virchu Sciences, a Columbus-based biotech company bringing AI software, oncologists and other researchers together to diagnose cancer earlier and more precisely. The company compiles huge amounts of medical imaging data, like CT scans or MRIs, to "train" AI to recognize the signs of cancer.
Large language models, he explained, can analyze hundreds of thousands of documents or images faster than any human. Because of that, AI can be a tool for medical professionals to compare those scans and better diagnose cancer in tandem with their own expertise.
At the same time that U.S. cancer rates are rising, there aren't enough pathologists in the workforce. In Ohio alone, rates of cancer under age 50 have increased as much as 20% in recent years, according to Ohio State University's James Cancer Hospital and Solove Research Institute.
The federal Health Resources and Services Administration projects a 7% decrease in pathologists in the U.S., while the need for them will increase 16% by 2037. In rural areas, only 20% of the actual need for pathologists will be met, according to the administration's projections.
Billiter noted that the data being used to train their AI software is secure, and a patient can't be identified by their scans.
"We're focused on cancer because it's hard, and we've only scratched the surface," Billiter said. But this same analysis can also apply to drug development.
If AI has access to millions of data points about existing drugs, their ingredients and uses, it could help build new formulas. This could speed up drug development, speeding access and reducing costs for the average patient.
"If you're someone where any disease or ailment touches you, and AI can aid in better treatment, the right treatment, the faster treatment, cheaper treatment, those are the things that we care about," he said.
Kelly Merrill Jr., Ph.D: In health care, AI is a start, but never an end
Kelly Merrill Jr., is an assistant professor of health communication and technology at the University of Cincinnati. His is a rare title, but in this new AI landscape, he predicts he'll soon have plenty of peers.
AI and health care are becoming increasingly intertwined, including in mental health. At least one in six American adults use AI for mental health help, according to data tracked by KFF Health News. That ranges from asking ChatGPT to help them find a therapist or even using the chatbot as a therapist itself.
There are myriad reasons someone may choose to ask AI for mental health help, Merrill said. It's accessible at a time when hundreds of thousands of Americans have lost their health insurance and as health care remains too expensive for many. And there's still a stigma around seeking mental health care.
But no matter how much data AI has, it's only getting some pieces of a puzzle while a medical professional is trained to see the whole. AI also tends to be "largely confirmatory" to the person using it, Merrill said, and could make people resistant to seek outside help if they're getting the validation they desire.
But Merrill doesn't advocate for cutting AI off completely. He encourages his students to use different AI tools until they "break" them. By doing so, they figure out how these tools work and can better understand AI's limitations, and therefore what it should and should not be used for.
If used correctly, AI can be a good starting point in someone's mental health journey, Merrill emphasized. General feelings of loneliness and depression, when discussed with a chatbot, could lead someone to helpful resources. But the end point should always be a medical professional.
Dennis Hirsch: AI and privacy rights
Dennis Hirsch is currently a professor at OSU teaching law, policy and governance of AI and is the director of the Center for Responsible AI and Governance (CRAIG). But he started his career in environmental law. Pollution is to the environment as privacy issues are to the information landscape, he said, recalling a lecture that fascinated him enough for a total career pivot.
"The pollution equivalent in AI governance is privacy invasion," he explained. "It's harmful bias, it's manipulation and it's what we call AI slop." These things are created through collected data and are "polluting the internet... hurting our digital environment."
AI, Hirsch said, is capable of good work. But it can't do that work if it's not trustworthy. A 2023 Pew Research Studyfound that 81% of adults are concerned over how companies use their data, and a 2026 study found that nearly 80% of adults who refused to use chat bots cited privacy concerns. The fear is valid. A report from Stanford University found that AI companies became less transparent from 2024 to 2025.
Right now, privacy protections for individuals are scarce. One thing Hirsch said people can do is not share deeply personal data with generative AI like ChatGPT. This limits the profile it create of you, which it may share without your consent.
Regulators have historically viewed individual consent, like saying yes to all cookies before visiting a website, as the "default." However, Hirsch doesn't believe the onus should ultimately fall on individuals to protect themselves.
"Individuals cannot understand how their data is being used, and so they cannot protect themselves in a meaningful way," Hirsch said. So, the future of privacy in AI is up to companies if they care about their customers.
At CRAIG, Hirsch and other academics work alongside private companies to help them create responsible AI governance in how customer data is used. If they show themselves trustworthy, it's good for their business − an incentive to protect customers.
"Responsible AI is successful AI," Hirsch said.
Business and consumer issues reporter Samantha Hendrickson can be reached at shendrickson@dispatch.com
This article originally appeared on The Columbus Dispatch: AI is going to change all our lives. Ohio experts explain how
