In a review of Georgia Tech’s OMSCS CS6750 (Human Computer Interaction), Eugene Yan highlights a core lesson for product builders: “You are not your user.” The course pushes back against the instinct to rely on personal intuition when designing interfaces, especially for systems meant to serve diverse and nontechnical audiences.
The review frames HCI as a practical discipline for anyone shipping software, including machine learning practitioners. Yan points to the value of structured techniques such as usability evaluation, personas, and iterative testing. These methods force teams to validate assumptions about how people will actually use a product, rather than assuming the user behaves like the engineer.
For AI and ML work, the lesson is especially relevant. A model may perform well on benchmarks, but its real-world impact depends on how humans interact with it—how they interpret its outputs, trust its recommendations, and recover from its mistakes. The source focuses on course content rather than specific ML techniques, but its emphasis on empathy and empirical user research speaks directly to building intelligent systems that people understand and rely on.