Central Asian Research Index

SCIENCE and SOCIETY, 2026, № 6-3, с. 30-32

OLIY TAʼLIMDA TALABALAR MUSTAQIL IShINI BAHOLAShDA SUNʼIY INTELLEKT IMKONIYATLARI

Sultanbek Nazerbaevich Kalbaev

AJINIYOZ NOMIDAGI NUKUS DAVLAT PEDAGOGIKA INSTITUTI

Аннотация · Abstract

The article analyses the possibilities of artificial intelligence (AI) in assessing students’ independent work from a mathematical standpoint: notation is introduced, seven postulates are stated formally, and a sequence of theorems is derived with proofs. It is shown that the question “is AI useful” is ill-posed; the correct question is where the binding constraint of learning through independent work lies and whether the cost structure of AI relaxes it without making its own errors expensive. From the postulates it is proved that the binding constraint is the feedback bottleneck, that the natural domain of AI is the formative function, that reliability and validity are orthogonal, that the quality of AI assessment is bounded above by rubric precision, and that the unit of assessment necessarily shifts from product to process.

Сведения · Details