Penerapan Generative AI untuk Meningkatkan Produktivitas Pembelajaran Berbasis Komputer di Perguruan Tinggi

Abstract

This study examined produktivitas belajar, dukungan penulisan, pencarian ide, umpan balik, dan pengelolaan tugas berbasis komputer through an integrative literature review of recent research on artificial intelligence and computer-based products. The review addressed a research gap concerning the tendency of existing discussions to separate technological capability from the conditions required for meaningful, verifiable, and responsible use. Relevant journal articles, conference papers, and authoritative guidance published primarily between 2022 and 2026 were analysed thematically. The analysis identified recurring patterns of application, reported benefits, implementation risks, and requirements for human oversight. The synthesis indicated that AI applications were most valuable when they supported an existing human activity rather than replacing the activity entirely. The literature consistently reported benefits related to efficiency, rapid feedback, ideation, personalisation, automation, and access to computational assistance, while also identifying risks involving inaccurate outputs, overreliance, privacy, academic integrity, unequal access, and insufficient verification. The review further showed that successful implementation depended on users’ domain knowledge, AI literacy, task design, and institutional governance. The study therefore proposed a human-centred implementation framework in which AI generation was followed by verification, reflection, and accountable use. The findings contributed a structured basis for designing or evaluating computer products that incorporate AI, particularly in educational and academic environments. Future empirical studies should test the proposed principles using comparative and longitudinal designs and should examine whether productivity gains remain beneficial when AI support is reduced or removed.