Generative Artificial Intelligence as a Tool for the Development and Evaluation Student’s Thinking

The paper considers scenarios and prospects of transformation of educational activities in connection with the widespread use of generative artificial intelligence (GPT). The paper systematizes the practical experience of using GPT in education and suggests the possibility of its application for the development and evaluation of students’ thinking. The case of using neural network YandexGPT in teaching students the basics of entrepreneurship in terms of generating innovative ideas and developing product concepts for specific markets is considered. A pedagogical design is proposed and a comparative experimental study is conducted to determine the possibilities and limitations of using neural networks for students to quickly immerse themselves in the topic and generate innovative ideas. To interpret the results of the experiment we used the method of expert evaluations, semantic analysis of prompts and neural network responses to student queries and statistical methods of hypothesis testing, in particular the non-parametric Kruskal-Wallis criterion. It is shown that students have different levels of practical skills in using GPT and lack of critical thinking for solving practical problems. Nevertheless, the use of GPT allowed student teams under time constraints to achieve results and present their concept of an innovative product for a specific market. According to the results of the experiment it is concluded that for more effective use of GPT language models in educational activities it is necessary to pre-train students in the basics of prompt-engineering and to develop new pedagogical methods of GPT application in various fields of knowledge.