Generative Artificial Intelligence and Teaching According to Aptitude in Private Undergraduate College English: A Production-Oriented Approach Framework

Generative artificial intelligence (GenAI) is reshaping how language is taught and learned, yet its potential to support teaching according to each learner's aptitude—a long-stated ideal in Chinese higher education—remains under-theorized, especially in resource-constrained private undergraduate institutions. This conceptual paper proposes a framework that embeds GenAI within the Production-Oriented Approach (POA) to enable differentiated, aptitude-responsive college English instruction at scale. Drawing on POA's three-stage cycle (driving, enabling, assessing) and on established accounts of differentiated instruction and individual differences in second language acquisition, the paper unpacks six implementation pathways: AI-assisted diagnostic profiling, adaptive enabling materials, culturally situated micro-content, differentiated production tasks, automated personalized feedback, and teacher mediation with integrity safeguards. Rather than treating GenAI as a substitute for pedagogical judgment, the framework positions it as an engine that extends the teacher's capacity to respond to learner readiness, interest, and learning profile. The paper clarifies the conditions under which such a model is likely to succeed, flags risks of over-reliance and inequity, and offers propositions for future empirical testing. It is offered as a theoretical contribution, not a report of an empirical study.