Abstract:

This article is a literature review of leading trends in artificial intelligence (AI), with a particular focus on Large Language Models. It is difficult to find anything about news in the literaturę. The Authors have attempted to provide the newest information and definitions, introducing the reader to how they work and how they can be used. The article focuses on a review of available information on the latest technological developments, i.e. self-supervised algorithm. Furthermore, the following article acts as a compendium of knowledge about large language models, data compression in language models, LlamaIndex, self-supervised nodes or neural network poisoning attacks. The authors conducted a thorough analysis of recent developments and advances related to AI topics. They read the newest publications, conference papers, journals and reputable websites to extract and identify key trends. They have attempted to answer the question "What are the potential implications of emerging technologies?". The answer is simple - cyber attacks. Unfortunately, cyber attacks on neural networks, specifically poisoning attacks, are already occurring. Cyber security professionals have quite a task. AI are intelligent machines capable of mimicking humans, and when infected they can be dangerous. GANs are the vaccine here. AI is a fast-growing field of science that has been increasingly successful recently. Autocorrect on the phone, personalised search according to our expectations or ChatGPT is the work of AI, and this is only a substitute for its possibilities. Undoubtedly, the increasing use of AI is making our lives easier, but it also poses new risks.