When ChatGPT launched in 2022, Clay Shirky — then NYU’s vice provost for education technologies — suddenly found himself guiding the university’s response to artificial intelligence. Since then, he’s spoken with hundreds of faculty members and students, aiming to understand how AI is changing academia.

Before joining the provost’s office, Shirky was an Interactive Media Arts professor at NYU Shanghai and served as the site’s chief information officer, helping faculty adapt to teaching in China. He is currently an associate professor at the Arthur L. Carter Journalism Institute and the Interactive Telecommunications Program, and has authored several books about technology and societal change.

In an interview with WSN, Shirky spoke about what he’s hearing from the NYU community, the university’s AI partnerships and why classrooms should preserve space for human connection.

This interview has been edited for length and clarity.

WSN: There’s a lot of discourse around AI moving faster than anyone can keep up with. What’s your view on that?

Shirky: I actually don’t like the narrative that change is now happening faster than ever. I lived through the web revolution, and although there was rhetoric of, ‘Oh my gosh, change is coming so fast,’ it took two decades to fully seep in. Right now, the most important paper on AI is called ‘AI as a Normal Technology.’ What it says is that behind the hype, this is just another powerful new technology diffusing through the ordinary sectors of society in ordinary ways. We should stop treating it as this unprecedented meteor strike and instead think, ‘How do we integrate new technologies when they show up?’ My first job is assuring people that there is time to deal with this. We need to deal with it expeditiously, but we don’t need to panic.

In an August New York Times op-ed, Shirky advised universities to take a “medieval turn” and transition from take-home assignments to in-class assessments, forcing students to learn without offloading work to AI. In an earlier op-ed, he said that despite the helpfulness of AI, some students are feeling a “growing sense of sadness” about their dependence on the technology. He told WSN that AI poses two main problems: a trade-off between productivity and learning and a tendency to excessively validate users.

WSN: Can you elaborate on your concerns with AI?

Shirky: If what we care about is human experience, these tools are very much a double-edged sword. Students, and some faculty, can get so focused on the output that we forget the only reason we’re asking students to do stuff is so they’ll have the experience of doing the work. That’s our general problem. The more specific thing I’m worried about is flattery. The tools say, ‘That’s such a good idea,’ ‘That’s such a smart question,’ ‘No one’s ever thought of that before’ —  just endless obsequious glazing. Your calculator is not telling you how good you are at math while you’re typing in seven times eight.

Despite his reservations, Shirky believes it’s also important to make AI financially accessible for students. NYU offers access to external services such as Microsoft Copilot, Zoom AI Companion, Adobe Firefly and, most notably, Google Gemini and NotebookLM. Shirky said that the university continues to prioritize partnering with companies that protect student data under the Family Educational Rights and Privacy Act. 

WSN: What drove NYU’s decision to partner with Big Tech companies?

Shirky: We were in a world where rich kids had better AI than poor kids. If you could spend $20 a month, you could get really high quality AI, but if you didn’t, you got the freemium model and constant upselling. Given the ubiquity of adoption of AI tools by students, we were very worried about equity, so we wanted to say, ‘Here’s a standard set of tools that everyone has access to.’ On the defensive side, for a long time, no one would write us a FERPA-compliant contract. It was only when it became clear that higher education was one of the principal sources of new users that those companies started to turn their attention to us. Google ended up being the only one that was willing to write us a compliant contract. 

A year after ChatGPT’s launch, Shirky asked students to confidentially share their experiences with AI in a WSN guest essay. Since then, Shirky has worked with Associate Vice Provost De Angela Duff to gather feedback across NYU’s global campuses through a student council. 

WSN: What have you learned from student feedback on AI? 

Shirky: The principal thing we’ve learned is that students don’t trust us. When we say it’s important to acknowledge and disclose AI use to your faculty member, that’s proven to be something students are only moderately willing to do. There seem to be two reasons for that. One, students don’t trust that they won’t get in trouble if they’re using tools in ways the faculty member doesn’t like, even if we promise in advance that won’t happen. The second is that students know taking shortcuts reduces the opportunity for learning, but they’re cross-pressured with time, extracurricular commitments and anxiety about grades and GPA.

WSN: What concerns you most about that? 

Shirky: The single biggest issue here is not cheating, it’s learning loss. When the people who assign the projects are also the people who grade them, we are both coaches and referees. The more students regard us as referees rather than coaches, the harder it is to have that conversation. This problem is more cultural than policy-oriented, but it’s one of the things that keeps me worrying.

Last month, President Linda Mills announced plans for “device-free environments” across the university. The move mirrors a policy Shirky implemented in 2014, when he banned laptops, tablets and phones in his classes, arguing that “multi-tasking is cognitively exhausting” and that devices create “second-hand smoke” for other students.

WSN: What role do you think technology should play in education?

Shirky: What we need to preserve in the classroom is the ability for human beings to hear each other. People often want to treat the university as a filling station where students drive in, fill up with as much knowledge as they think they need and then drive off. Frankly, in many classrooms, the most important things that happen are student-to-student conversations, where students are wrestling with ideas together. When I was in the classroom, I would assign the final paper to be due in week 12, and every student had to answer questions about what they wrote instead of presenting their paper. Those are some of the greatest conversations I’ve ever had the privilege of listening to. 

WSN: As an administrator, what would you like to hear from students?

Shirky: We want to hear from you. We want to know what you’re seeing, what you’re doing, what you’re thinking, what’s the difference between what you’re doing and what your friends are doing. The secret of higher ed is that the relationship between the university and its students is like the relationship between a river and its water. In the short term, the river tells the water where to go, and in the long term, the water tells the river where to go. We don’t have any illusion that we can change how AI works at NYU without involving the student perspective.

Contact Krish Dev at [email protected].