*Well, not everything, but some basic concepts that the author of this lecture has noticed are problems that still puzzle the IT community, such as the following statements:

AI will steal my job
AI can do anything
AI is always right
AI is real intelligence
AI is crap; it cannot be used for any real-world implementation
You have probably heard these (or some of these) statements many times, or you may have even used them yourself.

In this lecture, we will investigate hallucinations by using offline LLM models and explain the reasons behind them. The AI/ML correlation will be defined, including one simple concept called attention, which changed the AI landscape several years ago and gave birth to modern LLM models. We will explain attention's importance in NLP (Natural Language Processing).

Since we will start with a hallucinations demo, the culmination will be creating a useful model from the one that was hallucinating in the opening demo by modifying prompts and understanding the limits of the model itself. We will also explain the power behind KAGs and RAGs after defining basic concepts.