Tag: AI

Monika Stodolska spends two hours a week riding horses. At home, on occasion, she turns to artificial intelligence to help interpret the instructor’s advice. 

For the University of Illinois Urbana-Champaign faculty member, that is a promising use of technology: helping her get more from an activity that takes place far from a screen. 

But the same technology that helps someone become a better rider could also encourage people to surrender the choices, challenges and relationships that make leisure worthwhile. 

That tension drives Stodolska’s work on AI and the future of leisure. A professor in the Department of Recreation, Sport and Tourism in the College of Applied Health Sciences, she sees possibilities for creativity and connection alongside risks that convenience could deepen isolation and inequality. 

Whether the benefits outweigh the harm, her work suggests, will depend on decisions about regulation, corporate accountability and investment in community spaces where people spend time together. 

Stodolska edited a special issue of the journal Leisure Sciences featuring 28 critical commentaries and a closing essay. Published in fall 2026, the collection examines how AI could reshape leisure, education and public spaces. These are critical and theoretical arguments and reflections, rather than reports of new empirical studies. 

“The reality is very uncertain,” she said. “So those papers were very speculative.” 

The uncertainty begins with a deceptively simple question: if AI gives people more free time, will they enjoy it? 

Automating tasks could shorten workdays. Losing a job could also leave someone with hours to fill, but without the income, confidence or sense of purpose that helps make recreation satisfying. 

“Free time is not equivalent to leisure,” Stodolska said. “People who experience unemployment know that firsthand.” 

Meaningful leisure requires more than an empty calendar, she said. People need economic stability, a sense of belonging and accessible opportunities to participate. Parks, community centers and recreation programs matter, although access to them may be limited, particularly for people facing barriers of cost, distance or disability. 

Even for those with the means to enjoy themselves, AI could change how much control they exercise over their free time. 

A chatbot can suggest a vacation itinerary or identify restaurants based on someone’s preferences. That can spare travelers hours of research. But when a system supplies the options, the traveler’s role may shift from discovering possibilities to approving a prepared menu. 

Stodolska’s introduction to the special issue describes scholars’ concerns that increasingly personalized recommendations could narrow experiences while making them feel easier and more enjoyable. 

There is value, she said, in taking the wrong turn, the failed attempt and the surprise discovery. Learning to do something difficult and navigate through obstacles can be part of leisure’s reward. 

“The friction, the mistakes, the trials, the uncertainty is what shapes our character and allows people to grow,” she said. 

An AI system that continually agrees with its user could remove another source of growth: hearing a different perspective. This sycophancy inherent to many AI models, which often surfaces in leisure contexts, may lead to poor moral choices and a perceived license to engage in negative behaviors. Stodolska also worries that the appeal of polished, seemingly flawless results can overshadow the learning involved in producing something oneself. 

Her concern extends to the habits people develop when convenience becomes routine. “It’s very easy to outsource everything to AI,” she said. The risk, she argues, is that people stop practicing the thinking and learning that help them act independently, leaving them less prepared to judge what a system produces. 

Still, she sees ways AI could help people begin activities they might otherwise find intimidating. 

A novice musician, writer or artist could use it for instruction or to explore an initial idea. Her own riding offers an example of using AI to support learning while keeping the physical experience and human instructor central. 

The key, she said, is gradually reducing assistance as skills develop. Educators call that process “dynamic fading.” The aim is developing greater independence. 

The stakes become more complicated when the activity is conversation itself. 

In a paper with Olivia Huang—an RST doctoral student advised by Stodolska—and Sharifa Sultana—an assistant professor in the Siebel School of Computing and Data Science at Illinois—Stodolska examines AI companionship among groups that can face exclusion, including older adults, people with disabilities, LGBTQ+ youth and racial and ethnic minoritized groups. 

The authors caution against romanticizing conventional recreation. A park or community program may promise connection while remaining physically inaccessible, unaffordable or unwelcoming. 

“Not everybody has equal access to recreation opportunities, to natural environments, to parks,” Stodolska said. “And there’s a lot of exclusion and discrimination present in those leisure spaces.” 

An AI companion, by contrast, may be available whenever someone wants to talk. The paper reviews research suggesting such interactions can offer comfort, affirmation and opportunities for self-expression. 

But the authors also describe risks of developing emotional dependence, lost privacy and withdrawal from human relationships. Systems trained on human-generated material can reproduce the biases users hoped to escape. 

Their central distinction is between feeling connected and participating in a reciprocal relationship. A user may experience comfort, but an algorithm has no emotional needs or feelings of its own. The exchange also takes place on a commercial platform whose interests include keeping the user engaged. 

For people already isolated, the authors warn, AI could intensify vulnerability while leaving the causes of exclusion untouched. 

“There’s a risk of amplification also,” Stodolska said. People who are lonely or excluded may turn to AI instead of working through difficulties in face-to-face relationships, she warned. The immediate comfort could make withdrawal more appealing, even as the need for human connection remains. The question is broader than whether an individual chatbot produces a helpful answer or soothing companionship. It includes whose needs shape the product, who profits from intimate disclosures and who bears responsibility when something goes wrong. That makes oversight central to the future they describe. 

The companionship paper calls for companies and governments to build protections into systems from the beginning, include marginalized users in their design and evaluate how regulation could reduce potential harms. It also urges longer-term research to determine whether initial relief from loneliness persists or gives way to dependence. 

The special issue also puts AI’s physical costs in view. Data centers require land, water and energy, and communities may bear those burdens without sharing equally in the benefits. Other essays raise concerns about surveillance and biased information used to manage public spaces. Together, they frame leisure’s future as a question of public priorities, government and corporate control, as well as users’ personal preference. 

The special issue will be turned into a book to be published by Routledge in 2027. 

The potential financial trade-offs trouble Stodolska. Organizations may spend money that is intended to help vulnerable people on customized AI systems in the hope of saving money later, she said. That could draw resources away from the people whom those services exist to support before any promised savings materialize. 

“It may seem like an easy fix to public agencies,” she said. But providers need to ask whether that spending advances their mission. 

She urges recreation providers to always keep in mind their mission: supporting physical and mental health and providing opportunities for building social connections and access to nature. Any potential future use of AI should be judged against those goals. 

Those questions also reach her classroom, where she is preparing the people who will run recreation programs and manage leisure spaces. 

In a paper with RST Teaching Associate Professor Sarah Agate, Stodolska argues that educators should move beyond trying to detect and punish every use of AI. Students need to understand its capabilities, limitations and ethical implications while developing judgments of their own. 

She has redesigned her course assignments accordingly. Some require students to visit the community and assess accessibility of local facilities without the use of AI. Others ask them to use AI to design a recreation program, then improve it using knowledge from lectures and readings. 

“I’m for critical and conscious and very thoughtful use of AI,” she said. 

She is less concerned about recreation jobs disappearing than about jobs in some other industries, though she qualifies that confidence. Human contact and physical experiences remain central to what recreation provides. Her optimism rests partly on what people continue to seek beyond technology: being outdoors, sharing a place and learning to work with others. 

“We love to be in places, specific places,” she said. “And AI cannot replace place-based experiences.” 

For Stodolska, a better leisure future will require protecting opportunities to have those experiences, even as technology makes staying on a screen more convenient. 

The riding lesson still needs a horse, an instructor and a rider willing to learn. 

Related news