Harnessing AI for Good

Machines don’t know right from wrong. So how do we make sure the choices they make uphold what matters most to us?

USC researchers study how AI systems weigh decisions to inform the design of technologies where human wellbeing is programmed from the start.

Somewhere in the noise of thought and feeling, there’s a rhythm. AI picks it up, helping experts harness signals that shape perception, memory, and behavior. Research at USC Dornsife brings us closer to understanding the brain’s most intricate work: being human.

USC Dornsife researchers create and work with intelligent systems to fine-tune predictions, reduce friction, and uncover smarter ways to solve complex problems. It’s not just about improving health today — it’s reimagining what’s possible.

Simulating events across time. Weaving together scattered clues. Creating room for serendipity. AI sparks human ingenuity, reshaping the questions researchers can ask and powering the tools they create to answer them.

The basic research conducted at leading universities like USC isn’t often featured in the headlines. It’s over our heads and its relevance might not be obvious. But this research has been the leading driver of innovation in America for more than a century.

Follow the story of two brilliant USC faculty whose work connects across decades — and the basic research that was essential to AI-powered breakthroughs that will help scientists improve health in ways we never thought possible.

USC Dornsife structural and computational biologist Helen Berman’s seminal work in expanding the international Protein Data Bank (PDB) — an effort she co-founded in 1971 — serves as a foundation for AI systems that transform our understanding of human biology and accelerate new treatments for disease.

Think of proteins as tiny molecular machines that do almost everything in our bodies, from helping us digest food to fighting infections. Each protein is like a string of beads that folds into a specific 3D shape to function properly — crucial information for developing medicines.

The challenge? Proteins fold in billions of ways, making prediction through traditional computing nearly impossible. Scientists once estimated it would take longer than the age of the universe to calculate all possible configurations for even a simple protein!

This is where the Protein Data Bank revolutionized the field. By establishing the first open-access repository of experimentally determined 3D protein structures, Berman created the foundation for highly curated training datasets that would enable machine learning systems to identify protein folding patterns. Today, AI built on the PDB such as AlphaFold and DeepMind help scientists identify precise molecular structures and targets that could unlock treatments for Alzheimer’s, cancer, and many other health problems.

With complete cell models, we can better understand processes such as the way insulin production functions in diabetes or how neurons secrete mood-affecting hormones. The USC Cell Modeling Initiative (CMI) builds on the lessons learned from creating the Protein Data Bank to map cellular architectures and functions. Using AI and virtual reality, CMI aims to accelerate our understanding of disease mechanisms through a paradigm shift in visualization. The initiative is co-led by biochemist Kate White and Helen Berman, as well as USC computer scientist Carl Kesselman and USC production designer Alex McDowell. Together, they create immersive experiences that enable scientists to explore the inner workings of cells in ways that were previously impossible.

At USC Dornsife, undergraduate students can explore AI from all angles. Dozens of courses across the college — from our globally renowned School of Philosophy, our top-ranked Department of Quantitative and Computational Biology or our specialized Undergraduate Neuroscience program — ensure undergraduates gain a broad and rigorous foundation across areas that influence emerging technology. Whether they’re questioning the ethics of machine reasoning or completing an internship in Silicon Valley, they’ll think critically about how AI reshapes our world.

Selected Courses Featuring AI:

QBIO 465: Artificial Intelligence in Biology and Medicine
Study Al techniques including traditional machine learning and advanced deep learning methods for genomics, system biology, data integration, drug discovery and medical imaging.

PHIL/ENGR 265g: Ethics, Technology, and Value
Introduces and explores philosophical and ethical questions and theories related to technology, including biomedical technologies, internet applications and artificial intelligence.

CORE 499 Special Topic: Simulation & Society
Students use the technologies of extended reality (XR) and AI-powered simulation as a lens through which to explore the shifting nature of human values, interactions, and attitudes in the early 21st century.

Many of our PhD and master’s programs recognize and prioritize the rising demand for AI fluency across every sector. Students work alongside leading experts who redefine their fields and provide advanced training that anticipates the future. Grounded in the depth, versatility, and specialized skills that organizations are relying on to stay ahead of the curve, they’ll graduate with a competitive edge.