Exploring AI’s Role in Higher Education

Artificial intelligence is already reshaping classrooms, but will it require a complete reinvention of higher education by 2035? 

That was the question posed to the attendees at the Columbia AI Summit’s workshop, The Columbia Class of 2035: Will We Need to Reinvent Higher Education?, held on March 4 at Columbia University. The interactive session brought together faculty members and participants to discuss how institutions should adapt to AI’s evolving role in teaching and learning.

Columbia University Vice Dean of AI Initiatives Matthew Connelly, Professor of Computer Science and Electrical Engineering departments as well as the Vice Dean of Computing and Artificial Intelligence at Columbia Engineering Vishal Misra, and Executive Director of the Center for Teaching and Learning Amanda Irvin led the discussion. The workshop opened with a live poll, asking the audience to consider whether AI poses a challenge that could undermine higher education or presents an opportunity to innovate teaching methods.

AI Literacy, Ethics, and Core Skills

The conversation emphasized the need for students to develop AI literacy. In a video, Columbia Engineering Dean and Alma Schapiro Professor of Electrical Engineering Shih-Fu Chang stressed that students must understand the full AI ecosystem, including data acquisition, organization, and evaluation of AI-generated results. “How do we have that power, curriculum or teaching and educational program to ensure our students can critically evaluate AI systems, capability and indication,” Chang said, emphasizing that curricula must equip students with the ability to assess AI-generated outputs for accuracy, bias, and ethical concerns.

“AI is data-driven at its core. Students must grasp how to acquire, organize, and assess data effectively,” says Chang. 

Panelists agreed that higher education should focus on developing critical thinking, computer programming, and business model design skills. “As computer scientists and engineers, we are the architects of AI’s transformation in education,” said Misra. “We must ensure that AI enhances learning while maintaining rigorous technical education.”

Customizing learning with AI

The panel also explored AI’s potential to personalize learning experiences. Misra noted that AI tools can bridge gaps in student comprehension by allowing them to ask questions they may be hesitant to raise in class. “AI enables customized learning and provides a new level of knowledge accessibility,” he said, adding that this could fundamentally alter how students engage with the material.

However, panelists acknowledged AI’s challenges, including concerns over bias, privacy, and the risk of students becoming overly reliant on AI-generated content. Connelly shared insights from an ongoing Columbia study on generative AI in undergraduate writing courses. “Students didn't like the idea that they were basically becoming the secretary of AI. They were having to help the AI improve,” he said, underscoring the importance of ethical AI use.

The future of higher education

During the Q&A session, a computer science student asked whether traditional academic majors would still exist in 2035 or if AI would render them obsolete, making way for more interdisciplinary, self-directed learning paths.

Misra responded by acknowledging that while AI may reshape how students learn, structured academic disciplines would still be necessary. “Maybe not the same, but we will have some sort of major level because everyone has different strengths and interests,” he said. “And you want to pursue what you’re good at. I think classrooms without structure would be chaotic. You still need guidance, foundational knowledge, and a framework for learning.”

Looking ahead, Columbia is taking steps to lead AI integration in education. Connelly shared that Columbia’s engineering school is launching a new program to teach students coding with AI assistance, reflecting the university’s commitment to preparing students for an AI-driven future.

Echoing Chang’s insights in his video, the panel concluded with a discussion on AI’s role in fostering collaboration. 

“What is the role of AI in teamwork and human collaboration?” Chang asked. “AI is not just a tool for individuals or companies—it is becoming an indispensable part of teamwork, organizations, and decision-making processes. The key question is: how do we use AI not to hinder or replace collaboration, but to enhance creativity, cooperation, and efficiency between humans and machines?”