Features

Professor Patrick Barry Wants His Students to Become “Conspicuously Good” at AI

By Patrick Barry

Patrick Barry teaches a class at Michigan Law. — Patrick Barry is a clinical assistant professor of law and the director of digital academic initiatives at Michigan Law.

I have an ambitious goal for the students who take the courses on artificial intelligence I teach. By the end of the term, I want them to become “conspicuously good” at AI.

I want them to become conspicuously good at using AI. I want them to become conspicuously good at understanding the risks and limitations of AI. And I want them to become conspicuously good at adapting to whatever new AI tools and trends develop in the coming months, years, and decades.

AI Frankfurters

I take the phrase “conspicuously good” from a compliment that US Supreme Court Justice Felix Frankfurter received early in his career, when he was a 24-year-old prosecutor in New York. His boss at the time, Henry Stimson—who later served in the cabinet of four presidents—described Frankfurter’s work as “conspicuously good.” What Stimson meant is that Frankfurter’s work stood out in a positive way. It attracted laudatory notice and attention.

I want people to say something similar about the work my students produce, especially when it comes to AI. I want them to be the person in their office, their industry, or just their group of friends and family who earns a reputation for being AI savvy—someone whom other folks turn to when they need help with an AI-related issue or want to learn more about the latest AI innovation.

Imagine what that kind of status and expertise might do for their job prospects and professional ambitions. Imagine what it might do for their professional development, as well as their entrepreneurial endeavors—regardless of whether those endeavors are commercial, cultural, or wonderfully idiosyncratic.

I love seeing how, by the end of our time together, [my students’] relationship with AI transforms in several positive ways.

Professor Patrick Barry

More importantly, imagine what being AI savvy might do for their confidence and own sense that they’re equipped to navigate a fast-changing future. As Mhairi Aitken, a senior ethics fellow at the Alan Turing Institute, observed during a lecture on AI she gave at the Royal Institution in London, it used to be exciting when there was a new AI breakthrough every two to three months–now, it seems, there’s one every two to three hours.

Helping students, lawyers, and other advocates adjust to (and then excel in) that kind of fiercely dynamic environment is one of my favorite things about the AI courses and workshops I teach. The participants come from all over the world—India, Japan, Brazil, Nigeria, Mexico, Canada, Australia, France, and Saudi Arabia, not to mention every US state. I love seeing how, by the end of our time together, their relationship with AI transforms in several positive ways.

They become a lot more comfortable with AI. They become a lot more creative with AI. They also, importantly, develop a critical eye for spotting what the Princeton computer scientists Arvind Narayanan and Sayash Kapoor call “AI Snake Oil”: those bogus, over-hyped AI products that “don’t work, and probably never will.”

Part of being conspicuously good at AI, I tell my students, involves knowing how to avoid what is conspicuously bad.

Patrick Barry is a clinical assistant professor of law and the director of digital academic initiatives at Michigan Law. His four-part series, AI For Lawyers and Other Advocates, launched in November 2024 on Michigan Online and Coursera. The courses are free for U-M students, alumni, faculty, and staff.