Subjective Intelligence: A Framework for Generative AI in STEM Education

Generative artificial intelligence (GenAI) is increasingly transforming science and engineering education through prompt-based interactions. While promising to transform how students learn engineering, GenAI’s increasing presence raises concerns about misinformation, bias, academic integrity, and inequity in learning environments, especially in the absence of clear guidelines for fair and appropriate access and use. This position paper advances a conceptual framework for the use of GenAI in science and engineering through the lens of students’ identities and subjectivities, subjective intelligence, including students’ varied linguistic resources as well as gender and cultural identities. Our subjective intelligence framework investigates the emerging role of GenAI in shaping socio-academic engagement and pedagogical practices in STEM higher education contexts while examining its implications for equity and ethics. Our work draws on our first-hand experiences from an engineering undergraduate course, a graduate STEM seminar, and an engineering design task to illustrate how this framework can foster innovative STEM education. The framework comprises three core tenants: (1) cognitive and moral development towards ethical engagement in data practices, (2) identification and interrogation of potential human biases, and (3) multilingual/multidialectal support for design considerations. Across cases, the framework enables inclusive and reflective teaching strategies, while also surfacing new tensions and possibilities around GenAI’s limitations and misuses.