Optimize UC Berkeley's Machine Learning Course for the AI Age

Narges Norouzi and Joseph Gonzalez, faculty instructors for the CS 189: Introduction to Machine Learning course at UC Berkeley, are optimizing the curriculum to prepare students for rapid changes in the tech industry. Over 400 undergraduate students majoring in computer science and data science are currently enrolled in the course, which meets twice weekly. The CS 189 course update incorporates the latest advancements in machine learning and student feedback.

Berkeley has long been recognized as one of the top computer science programs in the nation. With the AI boom, students are utilizing existing tools and developing new ones while learning foundational areas in computer vision and neural network architecture, keeping up with the latest research and industry advances. Faculty like Norouzi and Gonzalez are rethinking how to best train the next generation of computer scientists.

Gonzalez, an associate professor in the Department of Electrical Engineering and Computer Sciences, emphasized that instead of starting from the ground up, the course begins by asking students how they can use machine learning in their daily lives and apply existing models to new problems. This reflects an approach that highlights practical use of AI.

The updated course continues to emphasize rigorous math and coding skills but also includes a focus on real-world problem-solving throughout the curriculum. Students tackle substantial machine learning problems, such as reproducing large language model ranking algorithms, which allows them to engage with chatbots and evaluate different models.

Norouzi expects educators and students in STEM and other disciplines to increasingly adopt AI tools to support teaching and learning. She referenced a recent Berkeley AI symposium that gathered faculty to discuss the use of generative AI in the classroom. Computing education is uniquely positioned to decide how to use AI tools and how much to revise the curriculum.

Norouzi also led the development of a tool called Askademia, which won the best paper award at the International Conference on Artificial Intelligence in Education. Gonzalez noted the need for continuous curriculum updates and flexibility in a rapidly changing field. Berkeley’s goal is not just to prepare students for the present but to help them learn to adapt to new challenges.