I actually was on a committee to do this for a university. Ultimately, some believed it would take resources away from computer science and compete with the pro

I actually was on a committee to do this for a university. Ultimately, some believed it would take resources away from computer science and compete with the program, so it was a year of work without anything to show for it. I'm still a proponent of undergraduate AI degrees, but this proposal doesn't suffice. Deployment, testing, dataset collecting, statistics, more math, better organization of electives (computer vision, robotics, NLP as applied electives), bias mitigation, ethics, and more need a role. The state and university guidelines also provide a lot of constraints. I do think that a research or applied project is essential, which is captured in this proposal. I'd definitely remove operating systems and compilers. I took both courses at the graduate level and have been working in basic and applied AI research for 15+ years, and those courses haven't helped me with that.