Affiliation:
1. M.D. Millionshchikov Academician Grozny State Petroleum Technical University
Abstract
Objective. Today, significant and in many ways sensational results are being achieved in the field of artificial intelligence systems, and the ChatGPT bot, which is based on the GPT-3 neural network, is called a real revolution in the world of technology.The aim of the study is to analyze and evaluate the application features, advantages and limitations, as well as development factors and reasons for the extraordinary popularity of the neural network dialogue system ChatGPT.Method. A review of domestic and foreign sources, systematization of data, analysis of the architecture and mechanism of action of the neural network was carried out. Result. Functions, opportunities, scopes and risks of using ChatGPT are summarized and evaluated.Conclusions. The main function of ChatGPT – text generation based on given input data – allows to effectively solve a wide range of tasks that have not been automated before, and the quality of the solution is comparable to human work. However, it is important to avoid the risks associated with the possibility of abuse and receiving incorrect and malicious responses from artificial intelligence. This requires control measures, the development and introduction of standards and norms.
Publisher
FSB Educational Establishment of Higher Education Daghestan State Technical University
Subject
Polymers and Plastics,General Environmental Science
Reference28 articles.
1. Grinin I.L. Operation of the text generation model using neural networks as a composite system: modulus analysis the module first. NLP-model: working with text embeddings. Innovacii i investicii, 2020; 7:18-122 (In Russ).
2. Kharlamov A.A., Le Manh Ha. Neural network approaches to the classification of texts based on morphological analysis. Trudy MFTI, 2017; 2(34): 143-150 (In Russ).
3. Bogomolov Yu. A. Overview of neural network models for natural language processing. Nauchnoobrazovatel’nyj zhurnal dlja studentov i prepodavatelej «StudNet», 2020; 4: 203-217 (In Russ).
4. Kruchinin V.V., Kuzovkin V.V. Overview of Existing Methods for Automatic Generation of Tasks with Conditions in Natural Language. Computer tools in education, 2022; 1:85-96. DOI: 10.32603/2071-2340- 2022-1-85-96 (In Russ).
5. Proshina M.V. Modern methods of natural language processing: neural networks. Jekonomika stroitel’stva, 2022; 5: 27-42 (In Russ).