Keyword: «neural network»

There are several definitions of the term "data". Depending on the context, this word can be interpreted differently, but in any case the point is the same: data is an information, that can be transferred one way or another. First of all IT data is a certain information, that can be recorded on some material media for the following transmission or processing. Meanwhile it is not necessary that information to reaches the other person. The point is that it should be possible. The purpose of this article is researching the development tendency of data transmission: how people started doing that, what information they did transmit, which ways they did it, what was the distance and other properties. How will this information be valuable? Basing on acquired knowledge we can conclude and make assumptions about the future evolution of data transmission and to learn about the current development trend in this area.

The phenomenon of synchronization in graph networks of coupled oscillators is discussed. To describe interactions between oscillators (neurons), a finite-dimensional first-order Kuramoto model without inertia is used. By means of parallel computing technology, a computational experiment has been carried out to simulate the time evolution of such networks for the number of oscillators n from 4 to 40 and for various values of the coupling coefficient K, 0≤K≤30. The influence of chosen coupling topology (including random topologies) on the amount of time required for the network to reach the complete synchronization state is considered.

In recent decades, developments in the field of artificial intelligence have led to significant changes in education. Specialists taught the computer to independently perform complex tasks using algorithms built on the basis of self-learning. However, how can these technologies be useful in education? And how are they applied in this area already today? In this article we tried to find answers to these questions.

ChatGPT (Generative Pre-trained Transformer) is artificial intelligence that can answer any user's question in the manner of a constructed text. It is already widespread and popular with most students. ChatGPT is able to write dissertations, pass medical tests or even make court decisions. However, it is worth noting that the appearance of this chatbot shocked the academic and university world: there was a scandal in one of the Russian universities in February 2023 because of a graduate qualification work written by ChatGPT neural network. The relevance of this research is due to the fact that this chatbot simultaneously attracts and worries the world pedagogical community by raising some ethical questions. Can ChatGPT clarify complex concepts and structure ideas in addition to creating smart mentors and teaching assistants? Is there any risk to create unoriginal, inaccurate or misleading content, as well as decrease creativity and skills? Will this tool be able to support pedagogical practice and successful studies, or will it endanger the fundamental education? What are the ways to reduce the risks by using ChatGPT in education? The authors of this research answered these questions. The scientific merit of the work is to consider trends in the development of artificial intelligence technologies in education, as well as possible scenarios for the application and impact of ChatGPT. Today, there are no clear rules in educational institutions regarding the use of artificial intelligence for educational purposes. The academic novelty of the study is that the authors identified and described the negative and positive aspects of using ChatGPT in education, and also formulated certain conditions to use painlessly the text generator in training. The methodological basis of the study is the analysis of Russian and foreign scientific literature on the development of artificial intelligence, in general, and the use of ChatGPT, in particular, as well as the synthesis and generalization of the results obtained.

The relevance of the research is caused by the close attention of students to the resources of the neural network for creating texts. It is extremely problematic for a teacher to give tasks related to creative work on writing texts within the framework of linguistic courses, since nowadays the neural network begins to replace a person in this type of activity. Therefore, there is a need to use the generated text in the educational process so that it would help, not harm, and contribute to the development of text analysis skills and the writing of original products. The purpose of the article is to propose a system of exercises that allow productive use of generated texts in the framework of linguistic courses. In conditions when a teacher cannot control students' access to programs that create text, there is a need to turn to the productive use of neural network capabilities in the educational process. As the literature review has shown, methodologists and scientists evaluate this potential differently. On the one hand, the use of a neural network causes negative feedback related to violations of ethical norms, copyright and a decrease in students' motivation to write their own texts. On the other hand, we cannot stop progress and it is no longer possible to prohibit students from accessing text generators, which means that teachers should, keeping up with the times, use new educational resources. The article presents the results of the conducted pedagogical experiment aimed at identifying the linguodidactic potential of those monological text products that the neural network creates. The authors collected 96 such texts, analyzed their linguistic side, genre and stylistic diversity, and content. A thorough analysis allowed us to propose five types of exercises that make it possible to productively organize work with texts created by artificial intelligence: 1) exercises aimed at analyzing the correspondence of texts to the chosen topic and the level of its revealing; 2) exercises related to the formulation of personal position and its argumentation; 3) exercises aimed at correcting language and speech errors in generated texts; 4) analysis of the genre correspondence of texts to the topic formulated in the request; 5) creative exercises aimed at writing one's own original texts. The authors came to the conclusion that text generators can be useful in the educational process, provided that they are used in a timely and methodically competent manner. This is what determines the practical significance of the work.