Using Retrieval-Augmented Generative AI to Support and Evaluate Reasoning in Research

Using Retrieval-Augmented Generative AI to Support and Evaluate Reasoning in Research

Mohammadesmail Nikfar1

1) Associate Professor, Ph.D. in Mathematics, Department of Mathematics, School of Hekmat 1, District 4, Tehran, Iran.

IOI Code : AEZA-ACAEBB
Publication : Quarterly Journal of Modern Technologies in Electrical and Computer Engineering, Vol. 5, No. 2 (Summer 2025), pp. 1–22.(https://eljournal.ir/fa/archive.php?rid=30)
Abstract :
    Usage is on the rise, and there is no choice but to adapt to this new reality. Many—though not all—students are using artificial intelligence to complete their assignments and tasks. We must ensure equitable access to AI systems for all students, researchers, and faculty, enabling these technologies to support and enhance their learning. We need to reflect carefully and deliberately on what we want education and learning to look like, especially given that generative AI systems are now so readily accessible. Simply continuing with "business as usual" is not a responsible approach. While chatbots can easily answer questions, generate content summaries, and make the educational process appear more efficient, there is little evidence that such time-saving actually translates into effective learning. It is crucial that we do not allow the use of generative AI systems to bypass or undermine the learning process. The availability of generative and retrieval-augmented AI systems offers new possibilities for what students, researchers, and faculty are expected to know and do. These tools can support the design of more complex tasks—tasks that elicit evidence of "three-dimensional" learning and facilitate individual reasoning. This text provides examples of these possibilities, accompanied by relevant educational materials, AI-driven personas, and important caveats regarding their use.

Keywords : higher education curriculum generative artificial intelligence learning theories student centered learning