The gold standard for cognitive science is transdisciplinary cognitive modeling of human behavior evaluated by quantitative comparisons with experiments involving human participants. However, this restrictive standard excludes much work in the history of cognitive science, and I argue sticking to this strict definition would impede future work in the field. In this paper, I examine conceptions of cognitive science and apply them to a breakdown of different kinds of cognitive modeling and AI, demonstrating that many different kinds of AI research have contributions to make to cognitive science. Even when a system cannot be evaluated by direct quantitative comparisons, rigorous methods exist to evaluate even novel qualitative work’s contribution.