Computer Science is About Shaping the Future, Not Just Writing Code

If you care about shaping the future of medicine, climate, art, policy, business or justice, Computer Science already has a place for you. You don’t have to arrive fluent in code. What you do need is a willingness to wrestle with complex ideas and a desire to understand how intelligent systems are transforming the world. 

For many students, especially those coming from outside of STEM, the first CS course can feel like learning a foreign language, which is natural — and temporary.  

Programming is often the entry point because students need a shared language. “We don’t have the English major’s advantage of students coming in already speaking our language,” acknowledged Kristin Stephens‑Martinez, associate professor of the practice of Computer Science and a recipient of the 2025 Trinity Undergraduate Teaching Award. Early courses necessarily focus on syntax and mechanics, but once students move beyond their first one or two classes, the field opens into something much broader. Programming remains a core tool, but today’s CS students are also learning how to work responsibly in a world increasingly shaped by AI. 

“Computer science education has to meet students where they are and where the field is going.” – Brandon Fain, assistant professor of the practice in Computer Science

Artificial intelligence now influences how we diagnose disease, allocate resources, create music, interpret language and decide what information audiences see. These systems aren’t just running on code, they’re being built by human choices. Computer science teaches students how to interrogate those choices, make better ones, and defend them with rigor.  

At Duke, this vision of computer science takes shape in a curriculum that treats AI not as optional, but as a core foundation. The result is an evolving field that invites students from across campus to treat AI not as optional, but as a part of the coursework environment.

One examples of this evolution is CS 216: Everything Data, a course taught by Stephen-Martinez. On paper, it might look like a data-oriented elective, but in practice, it’s a microcosm of modern computer science. 

“In the last three semesters I’ve taught CS 216, about half of them were non‑CS majors or double majors, with enrollments ranging from 77 to over 110 students. The non-CS majors ranged from Public Policy to Biology to Psychology,” Stephens‑Martinez said. That diversity is intentional and beneficial. The course brings together students from across campus to grapple with real-life data and the uncertainties that come with it. “I want students to come to the course bringing their own interests and contexts, to investigate questions they are genuinely invested in.” 

The emphasis is on asking good questions, evaluating evidence and making defensible claims. “Programming is actually very little emphasized in this class,” she said. “Instead, it’s about thinking critically, like a data scientist, and using programming as one of your tools to analyze data. Just like an English major is not just learning how to read and write English, a computer science major is not just learning programming.” 

This is also where AI enters the picture, not as a buzzword, but as an accelerator. “Layering AI into courses can help us get at those non‑programming CS concepts faster,” Stephens‑Martinez said, which allows students to focus sooner on reasoning, ethics and impact rather than mechanics alone. 

That vision resonates strongly with Brandon Fain, assistant professor of the practice in Computer Science and a 2025 finalist for the university-wide Judith Deckers Prize for Teaching Excellence. He describes his central goal as an educator as “preparing students to thrive in a world increasingly mediated by algorithms and artificial intelligence by cultivating technical expertise alongside creativity and societal consciousness.”  

“Layering AI into courses can help us get at non‑programming CS concepts faster.” – Kristin Stephens‑Martinez, associate professor of the practice of Computer Science

Fain sees courses like CS 171: Learning how to Learn with AI, taught by Stephens-Martinez, as a natural entry point for students interested in the AI & Humanity Constellation — part of the new Arts & Sciences Curriculum — a cluster of coursework for first-year students that explores the complex relationship between AI and human society. The Constellation also includes an ethical inquiry course from Fain concerning fairness, justice and power in modern applications of algorithmic systems and AI.  

Fain also created CS 372: Introduction to Applied Machine Learning, now enrolling more than 100 students per semester, which won a Trinity instructional award based on student evaluations. Rather than positioning machine learning as an abstract, math-heavy specialization, Fain’s aim is to use CS 372 to introduce students to modern deep learning by augmenting traditional math-focused courses. Through hands-on applications in computer vision, large language models, generative models and reinforcement learning, the course culminates in a final project that emphasizes creativity and real-world relevance. 

“Computer science education has to meet students where they are and where the field is going,” Fain explained. That forward-looking approach is visible across the department. Duke now offers a very popular AI/ML concentration within the CS major, with 45 students in the Class of 2025 and 87 declared for 2026. In just two semesters, undergraduate AI/ML/data science courses attracted 636 enrollments, while graduate offerings added another 624, spanning topics from natural language processing to generative models and the theory of deep learning.  

It’s also visible in who studies CS. Among Duke undergraduates graduating in 2025, 278 students — about 15.5% of the entire graduating class — pursued CS as a secondary major, minor, or interdisciplinary pathway, often alongside interests in everything from public policy to biology to the arts. CS is not replacing other disciplines; it’s interweaving with them. 

That interdisciplinarity is now being formalized in a proposed new CS curriculum, set to take effect as early as Fall 2026. The vision rests on three foundational pillars: theory and algorithms; computer systems; and artificial intelligence, machine learning and data science. 

This proposal aims to make Duke one of the only top CS programs where AI learning is required for all students. New core courses include CS 270: Math for AI, which teaches linear algebra and multivariate calculus through computational and AI-driven examples, and the new undergraduate CS 375: Introduction to Natural Processing, launching in Fall 2026 with Bhuwan Dhingra, assistant professor of Computer Science — the department's expert on the topic. Other core AI classes include CS 370: Introduction to Artificial Intelligence, CS 371:Elements of Machine Learning, plus the scaled-up CS 372 ensuring every graduate understands not only how AI works, but how it is applied, evaluated and questioned. 

This kind of thinking is central to an evolving discipline, reinventing itself. Computer science is not frozen in “learning to code.” Programming remains an essential tool, but at the heart of CS lie questions about abstraction, scale, algorithms, systems, models, data, and the ways people can use computer technology to think, reason and adapt in a rapidly changing technological world.