Collaborative Research: EducateAI: CUE-T: Designing Artificial Intelligence Curricula for All Undergrads

Artificial Intelligence (AI) has permeated almost every discipline and is shifting the landscape of the world in which we live. AI knowledge, skills and dispositions are essential not only for those computer science graduates pursuing a computing related career, but for students in other disciplines as well. Providing universal training of AI skills to a diverse populations will determine whether the next generation workforce of the US will remain competitive in the globe economy. There are several barriers to overcome to address the issue. For most computer science majors, AI courses are not required, even though most students take AI courses. For non-computer science or related majors, such as humanities, fine arts, etc, the main hurdle is that current AI courses require significant programming skills and prevent students without programming training to take them. Smaller regional colleges or minority-serving institutions have a relatively smaller faculty, more often focusing on covering existing courses with little time to develop new AI courses for students. Two-year community college faces a different problem that the curriculum is filled with required courses with little room to add new courses and little incentive for (often economically disadvantaged) students to pay for courses that are not strictly required for a degree or certificate. The goal of this proposal is to design an AI certificate that can appeal to a diverse population of students so that every student interested in AI can have an opportunity to take AI courses, and can adapt to institutions of different scales, including community colleges, regional universities and minority-serving institutions. We identified five principles in the design of the certificate: a zero-background entry point, a non-programming option, a surety of ethics, a discipline specific application, and an invitation to go deeper. These pieces combine in ways that allow for technical competency regardless of the student’s background, ensure an ethical and social good focus, and still do not require significant focus on programming. In particular, the project consists of four planned activities: • Through collaborative efforts, we will develop an AI certificate that will be accessible to widest groups of undergraduate students, including those who are from underrepresented and underserved groups. • The certificate will first be offered at the University of Kentucky and will be evaluated by the team of educators to make improvements. • The certificate or part of the curriculum will be improved and adopted by other universities/colleges, either as a separate course or a module in existing courses. • The offerings in these universities/colleges will be evaluated about their effectiveness, and revised version of the certificate, curriculum and modules will be developed for further distribution. Intellectual Merit This project explores novel design of an AI certificate that attracts to both computer science students and students from other majors who have little programming background. The effectiveness of teaching AI concepts and skills without programming to non-majors will be evaluated and improvements will be identified. It will investigate different adoption approaches, including adopting the whole certificate, adopting some courses, and adopting modules into existing courses, to meet the requirements of different institutions. It will also study how to teach students to combine basic AI skills to applications to different domains. The project will reveal new ways to provide effective AI education to a widest possible population of undergraduate students. Broader Impacts The project will provide opportunity for every undergraduate student to receive AI education, no matter what majors they are and whether they have previous programming skills. The collaboration among a community college, a regional university, a minority-serving university and a research university will evaluate the approach in different scenarios and potentially benefit students from a wide variety of higher education institutions. The majority of PI/Co-PIs on this project are female and may encourage female students to pursue a technical career. Graduate Teaching/Research assistants will be trained in both pedagogy and evaluation. The developed curricula will be made public and shared with the community.