When 3L Ryan Distaso first came to Michigan Law, he was a self-described “AI Luddite” and possessed a healthy skepticism about artificial intelligence. But he was certain that it would occupy an important place in the practice of law in the near future.
During his 2L year, intent on demystifying the technology, Distaso turned to AI Sandbox, a Law School class taught by Patrick Barry, a clinical assistant professor of law and director of digital academic initiatives.
“I didn’t want to be behind the curve in using AI,” says Distaso. “And I thought Professor Barry’s course might help me get up to speed.” The class proved useful, and as a summer associate, he used the technology at his firm as a starting point to research key cases and contours of unfamiliar areas of law.
3L Ryan Distaso[The class] convinced me of AI’s benefits and use cases, while also emphasizing its risks and limitations.
But Distaso still maintains a healthy skepticism and sees AI as neither panacea nor Pandora’s Box, which Barry addressed during the class.
“He did a great job of entertaining both perspectives: skeptics like me and AI optimists,” says Distaso. “He convinced me of AI’s benefits and use cases, while also emphasizing its risks and limitations.”
AI rewards expertise
As Barry leads students through the process of developing AI skills, he says that it is no replacement for the critical thinking skills that are a foundation for their legal education.
“One thing that is beginning to emerge out of the research is that AI tends to reward expertise,” he says. “If you know your subject area, then the power of AI to generate ideas is potentially more useful because you’re able to sort through the nonsense. If you don’t have that ability, then you may be convinced by something that is really just bonkers.”
One question he asks students, as well as himself, when using AI is, “Am I doing less thinking?” If a user is perhaps producing more but doing less creative work, coming up with fewer ideas, or feeling less energized after using the tool, the answer is probably “Yes.”
“I push my students to structure their interaction with AI so that they are not falling into the trap of cognitive offloading,” he says. To reframe perceptions of AI, Barry encourages students to explore the tool as a feedback mechanism that provides responses to their original work and a chance to course correct. He stresses, though, the difference between feedback and advice.
“I get nervous about following the advice of AI,” he says. “One my students said, ‘What I like about ChatGPT feedback is I don’t have to accept it.’”
Using the tools
Barry, who has taught AI courses both in and outside Michigan Law, focuses his sandbox course on the hands-on use of various tools, from large language models like ChatGPT to text-to-image generators.
“The term we use in class is ‘mechanical sympathy.’ I want them to understand the strengths and limitations of these tools.”
Students experiment with AI and can, for example, use it as a study buddy or upload their research to NotebookLM, which then produces a personalized podcast that summarizes legal concepts.
Opportunities for robust class discussions follow the students’ experiments with the tools, as Barry asks them to share their reflections. They also provide feedback on small whiteboards, writing of their experiences on a scale from zero (most negative) to 10 (most positive). Students then discuss their thoughts behind those numbers.
These interactions allow for less top-down instruction, says Barry, with students learning from each other as much as from him as the professor.
“I hope I contribute some expertise, but I’m very open to them coming up with new ideas and suggestions,” he says. “It’s less the pedagogical model that they’re empty vessels, and I’m pouring in information. It’s more of ‘We're exploring together.’ I want to give them principles so they explore in a skillful and responsible way, but I also want them to share with the students and with me what they learned and their reservations.”
Challenges of AI
Lindsey Bressler, a 3L who also took AI Sandbox last year and used the technology in her summer job, says that in addition to hands-on skills, the class also explored the tricky questions that the use of AI raises.
On a practical level, she feels challenged by working and learning in an entirely new manner. Instead of spending hours on research or reading, she might spend more time asking smarter questions to a chatbot. She also has found it useful in jump-starting the writing process.
3L Lindsey BresslerOne of the appeals of going to law school for me was the opportunity to work at the cutting edge of society’s most thorny and interesting problems. AI presents both societal problems and solutions.
“I suffer from writer’s block,” she says. “AI’s ability to break the first step of drafting something—whether that’s a quick email or a more substantive piece—can help spur productivity in ways that were unavailable to me before this technology and make procrastination or fear of getting something ‘good enough’ far less of an obstacle.”
She adds that the biggest challenges of using AI are much more philosophical than practical.
“There are questions of whether or not a chatbot provides accurate information or if it truly saves time in comparison to not using AI, not to mention the questions about AI’s energy usage and ethical questions that come up when people use AI for their mental health,” she says. “One of the appeals of going to law school for me was the opportunity to work at the cutting edge of society’s most thorny and interesting problems. AI presents both societal problems and solutions.”
Continue Reading
At the Crest of the Wave: Artificial Intelligence in the Law Quad
In our Winter 2025 cover story, Law Quadrangle examines how our faculty and students are navigating this new landscape and exploring how AI may be incorporated into legal education and the profession.
